diff --git "a/checkpoint-9000/trainer_state.json" "b/checkpoint-9000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-9000/trainer_state.json"
@@ -0,0 +1,63033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 14.37699680511182,
+  "eval_steps": 500,
+  "global_step": 9000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001597444089456869,
+      "grad_norm": 65.0,
+      "learning_rate": 0.0005,
+      "loss": 4.0939,
+      "step": 1
+    },
+    {
+      "epoch": 0.003194888178913738,
+      "grad_norm": 193.0,
+      "learning_rate": 0.0005,
+      "loss": 44.689,
+      "step": 2
+    },
+    {
+      "epoch": 0.004792332268370607,
+      "grad_norm": 11.9375,
+      "learning_rate": 0.0005,
+      "loss": 2.4985,
+      "step": 3
+    },
+    {
+      "epoch": 0.006389776357827476,
+      "grad_norm": 135.0,
+      "learning_rate": 0.0005,
+      "loss": 18.185,
+      "step": 4
+    },
+    {
+      "epoch": 0.007987220447284345,
+      "grad_norm": 10.375,
+      "learning_rate": 0.0005,
+      "loss": 2.5044,
+      "step": 5
+    },
+    {
+      "epoch": 0.009584664536741214,
+      "grad_norm": 18.75,
+      "learning_rate": 0.0005,
+      "loss": 3.1121,
+      "step": 6
+    },
+    {
+      "epoch": 0.011182108626198083,
+      "grad_norm": 6.1875,
+      "learning_rate": 0.0005,
+      "loss": 2.3853,
+      "step": 7
+    },
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 13.75,
+      "learning_rate": 0.0005,
+      "loss": 2.6798,
+      "step": 8
+    },
+    {
+      "epoch": 0.01437699680511182,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0005,
+      "loss": 2.0754,
+      "step": 9
+    },
+    {
+      "epoch": 0.01597444089456869,
+      "grad_norm": 9.0,
+      "learning_rate": 0.0005,
+      "loss": 2.4117,
+      "step": 10
+    },
+    {
+      "epoch": 0.01757188498402556,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0005,
+      "loss": 2.0883,
+      "step": 11
+    },
+    {
+      "epoch": 0.019169329073482427,
+      "grad_norm": 5.84375,
+      "learning_rate": 0.0005,
+      "loss": 2.2847,
+      "step": 12
+    },
+    {
+      "epoch": 0.020766773162939296,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0005,
+      "loss": 2.0143,
+      "step": 13
+    },
+    {
+      "epoch": 0.022364217252396165,
+      "grad_norm": 8.8125,
+      "learning_rate": 0.0005,
+      "loss": 2.2207,
+      "step": 14
+    },
+    {
+      "epoch": 0.023961661341853034,
+      "grad_norm": 9.25,
+      "learning_rate": 0.0005,
+      "loss": 2.2669,
+      "step": 15
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0005,
+      "loss": 2.026,
+      "step": 16
+    },
+    {
+      "epoch": 0.027156549520766772,
+      "grad_norm": 6.875,
+      "learning_rate": 0.0005,
+      "loss": 2.1755,
+      "step": 17
+    },
+    {
+      "epoch": 0.02875399361022364,
+      "grad_norm": 7.9375,
+      "learning_rate": 0.0005,
+      "loss": 2.245,
+      "step": 18
+    },
+    {
+      "epoch": 0.03035143769968051,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.0005,
+      "loss": 2.085,
+      "step": 19
+    },
+    {
+      "epoch": 0.03194888178913738,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0005,
+      "loss": 2.0103,
+      "step": 20
+    },
+    {
+      "epoch": 0.03354632587859425,
+      "grad_norm": 5.1875,
+      "learning_rate": 0.0005,
+      "loss": 2.05,
+      "step": 21
+    },
+    {
+      "epoch": 0.03514376996805112,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.9547,
+      "step": 22
+    },
+    {
+      "epoch": 0.036741214057507986,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.9833,
+      "step": 23
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0005,
+      "loss": 2.0097,
+      "step": 24
+    },
+    {
+      "epoch": 0.039936102236421724,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0005,
+      "loss": 1.9292,
+      "step": 25
+    },
+    {
+      "epoch": 0.04153354632587859,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0005,
+      "loss": 1.9597,
+      "step": 26
+    },
+    {
+      "epoch": 0.04313099041533546,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0005,
+      "loss": 1.9812,
+      "step": 27
+    },
+    {
+      "epoch": 0.04472843450479233,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0005,
+      "loss": 1.9219,
+      "step": 28
+    },
+    {
+      "epoch": 0.0463258785942492,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.9522,
+      "step": 29
+    },
+    {
+      "epoch": 0.04792332268370607,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.9688,
+      "step": 30
+    },
+    {
+      "epoch": 0.04952076677316294,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.9314,
+      "step": 31
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0005,
+      "loss": 1.9033,
+      "step": 32
+    },
+    {
+      "epoch": 0.052715654952076675,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.8972,
+      "step": 33
+    },
+    {
+      "epoch": 0.054313099041533544,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.8782,
+      "step": 34
+    },
+    {
+      "epoch": 0.05591054313099041,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0005,
+      "loss": 1.8881,
+      "step": 35
+    },
+    {
+      "epoch": 0.05750798722044728,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0005,
+      "loss": 1.8715,
+      "step": 36
+    },
+    {
+      "epoch": 0.05910543130990415,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.8595,
+      "step": 37
+    },
+    {
+      "epoch": 0.06070287539936102,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0005,
+      "loss": 1.8555,
+      "step": 38
+    },
+    {
+      "epoch": 0.06230031948881789,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0005,
+      "loss": 1.8432,
+      "step": 39
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.8471,
+      "step": 40
+    },
+    {
+      "epoch": 0.06549520766773163,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.8343,
+      "step": 41
+    },
+    {
+      "epoch": 0.0670926517571885,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0005,
+      "loss": 1.8101,
+      "step": 42
+    },
+    {
+      "epoch": 0.06869009584664537,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.7895,
+      "step": 43
+    },
+    {
+      "epoch": 0.07028753993610223,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.7941,
+      "step": 44
+    },
+    {
+      "epoch": 0.07188498402555911,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0005,
+      "loss": 1.7748,
+      "step": 45
+    },
+    {
+      "epoch": 0.07348242811501597,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0005,
+      "loss": 1.7795,
+      "step": 46
+    },
+    {
+      "epoch": 0.07507987220447285,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.7323,
+      "step": 47
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0005,
+      "loss": 1.7048,
+      "step": 48
+    },
+    {
+      "epoch": 0.07827476038338659,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.6633,
+      "step": 49
+    },
+    {
+      "epoch": 0.07987220447284345,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0005,
+      "loss": 1.6381,
+      "step": 50
+    },
+    {
+      "epoch": 0.08146964856230032,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.6256,
+      "step": 51
+    },
+    {
+      "epoch": 0.08306709265175719,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.6109,
+      "step": 52
+    },
+    {
+      "epoch": 0.08466453674121406,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.649,
+      "step": 53
+    },
+    {
+      "epoch": 0.08626198083067092,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0005,
+      "loss": 1.5915,
+      "step": 54
+    },
+    {
+      "epoch": 0.0878594249201278,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.5747,
+      "step": 55
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.5402,
+      "step": 56
+    },
+    {
+      "epoch": 0.09105431309904154,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.5304,
+      "step": 57
+    },
+    {
+      "epoch": 0.0926517571884984,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0005,
+      "loss": 1.5405,
+      "step": 58
+    },
+    {
+      "epoch": 0.09424920127795527,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0005,
+      "loss": 1.5307,
+      "step": 59
+    },
+    {
+      "epoch": 0.09584664536741214,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0005,
+      "loss": 1.4952,
+      "step": 60
+    },
+    {
+      "epoch": 0.09744408945686901,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.5089,
+      "step": 61
+    },
+    {
+      "epoch": 0.09904153354632587,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0005,
+      "loss": 1.4917,
+      "step": 62
+    },
+    {
+      "epoch": 0.10063897763578275,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.4577,
+      "step": 63
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.4759,
+      "step": 64
+    },
+    {
+      "epoch": 0.10383386581469649,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.421,
+      "step": 65
+    },
+    {
+      "epoch": 0.10543130990415335,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0005,
+      "loss": 1.4349,
+      "step": 66
+    },
+    {
+      "epoch": 0.10702875399361023,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0005,
+      "loss": 1.4464,
+      "step": 67
+    },
+    {
+      "epoch": 0.10862619808306709,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.4142,
+      "step": 68
+    },
+    {
+      "epoch": 0.11022364217252396,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.413,
+      "step": 69
+    },
+    {
+      "epoch": 0.11182108626198083,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0005,
+      "loss": 1.4292,
+      "step": 70
+    },
+    {
+      "epoch": 0.1134185303514377,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0005,
+      "loss": 1.4128,
+      "step": 71
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.3825,
+      "step": 72
+    },
+    {
+      "epoch": 0.11661341853035144,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.3928,
+      "step": 73
+    },
+    {
+      "epoch": 0.1182108626198083,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.3785,
+      "step": 74
+    },
+    {
+      "epoch": 0.11980830670926518,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3725,
+      "step": 75
+    },
+    {
+      "epoch": 0.12140575079872204,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.3653,
+      "step": 76
+    },
+    {
+      "epoch": 0.12300319488817892,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0005,
+      "loss": 1.3924,
+      "step": 77
+    },
+    {
+      "epoch": 0.12460063897763578,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.4068,
+      "step": 78
+    },
+    {
+      "epoch": 0.12619808306709265,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0005,
+      "loss": 1.3757,
+      "step": 79
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3729,
+      "step": 80
+    },
+    {
+      "epoch": 0.12939297124600638,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3941,
+      "step": 81
+    },
+    {
+      "epoch": 0.13099041533546327,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0005,
+      "loss": 1.3857,
+      "step": 82
+    },
+    {
+      "epoch": 0.13258785942492013,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0005,
+      "loss": 1.3822,
+      "step": 83
+    },
+    {
+      "epoch": 0.134185303514377,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.3937,
+      "step": 84
+    },
+    {
+      "epoch": 0.13578274760383385,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.3824,
+      "step": 85
+    },
+    {
+      "epoch": 0.13738019169329074,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.3654,
+      "step": 86
+    },
+    {
+      "epoch": 0.1389776357827476,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.3623,
+      "step": 87
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.3661,
+      "step": 88
+    },
+    {
+      "epoch": 0.14217252396166133,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3393,
+      "step": 89
+    },
+    {
+      "epoch": 0.14376996805111822,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.359,
+      "step": 90
+    },
+    {
+      "epoch": 0.14536741214057508,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0005,
+      "loss": 1.3434,
+      "step": 91
+    },
+    {
+      "epoch": 0.14696485623003194,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.376,
+      "step": 92
+    },
+    {
+      "epoch": 0.1485623003194888,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.358,
+      "step": 93
+    },
+    {
+      "epoch": 0.1501597444089457,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3711,
+      "step": 94
+    },
+    {
+      "epoch": 0.15175718849840256,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3362,
+      "step": 95
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.3429,
+      "step": 96
+    },
+    {
+      "epoch": 0.15495207667731628,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.352,
+      "step": 97
+    },
+    {
+      "epoch": 0.15654952076677317,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.3373,
+      "step": 98
+    },
+    {
+      "epoch": 0.15814696485623003,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3479,
+      "step": 99
+    },
+    {
+      "epoch": 0.1597444089456869,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.3259,
+      "step": 100
+    },
+    {
+      "epoch": 0.16134185303514376,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.3432,
+      "step": 101
+    },
+    {
+      "epoch": 0.16293929712460065,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.3332,
+      "step": 102
+    },
+    {
+      "epoch": 0.1645367412140575,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3245,
+      "step": 103
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.3316,
+      "step": 104
+    },
+    {
+      "epoch": 0.16773162939297126,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.3295,
+      "step": 105
+    },
+    {
+      "epoch": 0.16932907348242812,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3369,
+      "step": 106
+    },
+    {
+      "epoch": 0.17092651757188498,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.3271,
+      "step": 107
+    },
+    {
+      "epoch": 0.17252396166134185,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3213,
+      "step": 108
+    },
+    {
+      "epoch": 0.17412140575079874,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.321,
+      "step": 109
+    },
+    {
+      "epoch": 0.1757188498402556,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.3253,
+      "step": 110
+    },
+    {
+      "epoch": 0.17731629392971246,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.3094,
+      "step": 111
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.3168,
+      "step": 112
+    },
+    {
+      "epoch": 0.1805111821086262,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2921,
+      "step": 113
+    },
+    {
+      "epoch": 0.18210862619808307,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3098,
+      "step": 114
+    },
+    {
+      "epoch": 0.18370607028753994,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0005,
+      "loss": 1.3256,
+      "step": 115
+    },
+    {
+      "epoch": 0.1853035143769968,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.308,
+      "step": 116
+    },
+    {
+      "epoch": 0.1869009584664537,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.3305,
+      "step": 117
+    },
+    {
+      "epoch": 0.18849840255591055,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3202,
+      "step": 118
+    },
+    {
+      "epoch": 0.1900958466453674,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.318,
+      "step": 119
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3197,
+      "step": 120
+    },
+    {
+      "epoch": 0.19329073482428116,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.3213,
+      "step": 121
+    },
+    {
+      "epoch": 0.19488817891373802,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0005,
+      "loss": 1.3155,
+      "step": 122
+    },
+    {
+      "epoch": 0.1964856230031949,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3185,
+      "step": 123
+    },
+    {
+      "epoch": 0.19808306709265175,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.3372,
+      "step": 124
+    },
+    {
+      "epoch": 0.19968051118210864,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.3209,
+      "step": 125
+    },
+    {
+      "epoch": 0.2012779552715655,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.3032,
+      "step": 126
+    },
+    {
+      "epoch": 0.20287539936102236,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.3103,
+      "step": 127
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.3057,
+      "step": 128
+    },
+    {
+      "epoch": 0.20607028753993611,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.3093,
+      "step": 129
+    },
+    {
+      "epoch": 0.20766773162939298,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.303,
+      "step": 130
+    },
+    {
+      "epoch": 0.20926517571884984,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.3118,
+      "step": 131
+    },
+    {
+      "epoch": 0.2108626198083067,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.3077,
+      "step": 132
+    },
+    {
+      "epoch": 0.2124600638977636,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.3087,
+      "step": 133
+    },
+    {
+      "epoch": 0.21405750798722045,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2919,
+      "step": 134
+    },
+    {
+      "epoch": 0.21565495207667731,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.3037,
+      "step": 135
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.3094,
+      "step": 136
+    },
+    {
+      "epoch": 0.21884984025559107,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.3026,
+      "step": 137
+    },
+    {
+      "epoch": 0.22044728434504793,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0005,
+      "loss": 1.3295,
+      "step": 138
+    },
+    {
+      "epoch": 0.2220447284345048,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3431,
+      "step": 139
+    },
+    {
+      "epoch": 0.22364217252396165,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0005,
+      "loss": 1.3114,
+      "step": 140
+    },
+    {
+      "epoch": 0.22523961661341854,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0005,
+      "loss": 1.3115,
+      "step": 141
+    },
+    {
+      "epoch": 0.2268370607028754,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3255,
+      "step": 142
+    },
+    {
+      "epoch": 0.22843450479233227,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2961,
+      "step": 143
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.3081,
+      "step": 144
+    },
+    {
+      "epoch": 0.23162939297124602,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3114,
+      "step": 145
+    },
+    {
+      "epoch": 0.23322683706070288,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3076,
+      "step": 146
+    },
+    {
+      "epoch": 0.23482428115015974,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0005,
+      "loss": 1.3302,
+      "step": 147
+    },
+    {
+      "epoch": 0.2364217252396166,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0005,
+      "loss": 1.31,
+      "step": 148
+    },
+    {
+      "epoch": 0.2380191693290735,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0005,
+      "loss": 1.3392,
+      "step": 149
+    },
+    {
+      "epoch": 0.23961661341853036,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3512,
+      "step": 150
+    },
+    {
+      "epoch": 0.24121405750798722,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.3118,
+      "step": 151
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0005,
+      "loss": 1.3106,
+      "step": 152
+    },
+    {
+      "epoch": 0.24440894568690097,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0005,
+      "loss": 1.3129,
+      "step": 153
+    },
+    {
+      "epoch": 0.24600638977635783,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3034,
+      "step": 154
+    },
+    {
+      "epoch": 0.2476038338658147,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0005,
+      "loss": 1.3245,
+      "step": 155
+    },
+    {
+      "epoch": 0.24920127795527156,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.3134,
+      "step": 156
+    },
+    {
+      "epoch": 0.2507987220447284,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0005,
+      "loss": 1.3115,
+      "step": 157
+    },
+    {
+      "epoch": 0.2523961661341853,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3001,
+      "step": 158
+    },
+    {
+      "epoch": 0.2539936102236422,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2944,
+      "step": 159
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0005,
+      "loss": 1.301,
+      "step": 160
+    },
+    {
+      "epoch": 0.2571884984025559,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.3009,
+      "step": 161
+    },
+    {
+      "epoch": 0.25878594249201275,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0005,
+      "loss": 1.318,
+      "step": 162
+    },
+    {
+      "epoch": 0.26038338658146964,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0005,
+      "loss": 1.3106,
+      "step": 163
+    },
+    {
+      "epoch": 0.26198083067092653,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.2947,
+      "step": 164
+    },
+    {
+      "epoch": 0.26357827476038337,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3142,
+      "step": 165
+    },
+    {
+      "epoch": 0.26517571884984026,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0005,
+      "loss": 1.2895,
+      "step": 166
+    },
+    {
+      "epoch": 0.26677316293929715,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2967,
+      "step": 167
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2979,
+      "step": 168
+    },
+    {
+      "epoch": 0.26996805111821087,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2905,
+      "step": 169
+    },
+    {
+      "epoch": 0.2715654952076677,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2862,
+      "step": 170
+    },
+    {
+      "epoch": 0.2731629392971246,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2825,
+      "step": 171
+    },
+    {
+      "epoch": 0.2747603833865815,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2758,
+      "step": 172
+    },
+    {
+      "epoch": 0.2763578274760383,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2866,
+      "step": 173
+    },
+    {
+      "epoch": 0.2779552715654952,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2919,
+      "step": 174
+    },
+    {
+      "epoch": 0.2795527156549521,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2861,
+      "step": 175
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2762,
+      "step": 176
+    },
+    {
+      "epoch": 0.2827476038338658,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2767,
+      "step": 177
+    },
+    {
+      "epoch": 0.28434504792332266,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2898,
+      "step": 178
+    },
+    {
+      "epoch": 0.28594249201277955,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2761,
+      "step": 179
+    },
+    {
+      "epoch": 0.28753993610223644,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.263,
+      "step": 180
+    },
+    {
+      "epoch": 0.28913738019169327,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2825,
+      "step": 181
+    },
+    {
+      "epoch": 0.29073482428115016,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2904,
+      "step": 182
+    },
+    {
+      "epoch": 0.29233226837060705,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2751,
+      "step": 183
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2914,
+      "step": 184
+    },
+    {
+      "epoch": 0.2955271565495208,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2696,
+      "step": 185
+    },
+    {
+      "epoch": 0.2971246006389776,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2712,
+      "step": 186
+    },
+    {
+      "epoch": 0.2987220447284345,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2787,
+      "step": 187
+    },
+    {
+      "epoch": 0.3003194888178914,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2769,
+      "step": 188
+    },
+    {
+      "epoch": 0.3019169329073482,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2653,
+      "step": 189
+    },
+    {
+      "epoch": 0.3035143769968051,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.2647,
+      "step": 190
+    },
+    {
+      "epoch": 0.305111821086262,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2877,
+      "step": 191
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2605,
+      "step": 192
+    },
+    {
+      "epoch": 0.3083067092651757,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2797,
+      "step": 193
+    },
+    {
+      "epoch": 0.30990415335463256,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2826,
+      "step": 194
+    },
+    {
+      "epoch": 0.31150159744408945,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2725,
+      "step": 195
+    },
+    {
+      "epoch": 0.31309904153354634,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2739,
+      "step": 196
+    },
+    {
+      "epoch": 0.3146964856230032,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2729,
+      "step": 197
+    },
+    {
+      "epoch": 0.31629392971246006,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2796,
+      "step": 198
+    },
+    {
+      "epoch": 0.31789137380191695,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2764,
+      "step": 199
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2799,
+      "step": 200
+    },
+    {
+      "epoch": 0.3210862619808307,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2899,
+      "step": 201
+    },
+    {
+      "epoch": 0.3226837060702875,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2804,
+      "step": 202
+    },
+    {
+      "epoch": 0.3242811501597444,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.258,
+      "step": 203
+    },
+    {
+      "epoch": 0.3258785942492013,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2633,
+      "step": 204
+    },
+    {
+      "epoch": 0.3274760383386581,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2743,
+      "step": 205
+    },
+    {
+      "epoch": 0.329073482428115,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2707,
+      "step": 206
+    },
+    {
+      "epoch": 0.3306709265175719,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2905,
+      "step": 207
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2899,
+      "step": 208
+    },
+    {
+      "epoch": 0.33386581469648563,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2614,
+      "step": 209
+    },
+    {
+      "epoch": 0.3354632587859425,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.0005,
+      "loss": 1.3266,
+      "step": 210
+    },
+    {
+      "epoch": 0.33706070287539935,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.3038,
+      "step": 211
+    },
+    {
+      "epoch": 0.33865814696485624,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0005,
+      "loss": 1.3769,
+      "step": 212
+    },
+    {
+      "epoch": 0.3402555910543131,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0005,
+      "loss": 1.3526,
+      "step": 213
+    },
+    {
+      "epoch": 0.34185303514376997,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0005,
+      "loss": 1.3073,
+      "step": 214
+    },
+    {
+      "epoch": 0.34345047923322686,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.3201,
+      "step": 215
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3147,
+      "step": 216
+    },
+    {
+      "epoch": 0.3466453674121406,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2978,
+      "step": 217
+    },
+    {
+      "epoch": 0.34824281150159747,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0005,
+      "loss": 1.301,
+      "step": 218
+    },
+    {
+      "epoch": 0.3498402555910543,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0005,
+      "loss": 1.3008,
+      "step": 219
+    },
+    {
+      "epoch": 0.3514376996805112,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0005,
+      "loss": 1.2982,
+      "step": 220
+    },
+    {
+      "epoch": 0.35303514376996803,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2868,
+      "step": 221
+    },
+    {
+      "epoch": 0.3546325878594249,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0005,
+      "loss": 1.296,
+      "step": 222
+    },
+    {
+      "epoch": 0.3562300319488818,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2881,
+      "step": 223
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0005,
+      "loss": 1.276,
+      "step": 224
+    },
+    {
+      "epoch": 0.35942492012779553,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2912,
+      "step": 225
+    },
+    {
+      "epoch": 0.3610223642172524,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2915,
+      "step": 226
+    },
+    {
+      "epoch": 0.36261980830670926,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.285,
+      "step": 227
+    },
+    {
+      "epoch": 0.36421725239616615,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2846,
+      "step": 228
+    },
+    {
+      "epoch": 0.365814696485623,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2714,
+      "step": 229
+    },
+    {
+      "epoch": 0.36741214057507987,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2824,
+      "step": 230
+    },
+    {
+      "epoch": 0.36900958466453676,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2615,
+      "step": 231
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.2791,
+      "step": 232
+    },
+    {
+      "epoch": 0.3722044728434505,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2774,
+      "step": 233
+    },
+    {
+      "epoch": 0.3738019169329074,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2565,
+      "step": 234
+    },
+    {
+      "epoch": 0.3753993610223642,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2719,
+      "step": 235
+    },
+    {
+      "epoch": 0.3769968051118211,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2579,
+      "step": 236
+    },
+    {
+      "epoch": 0.37859424920127793,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2587,
+      "step": 237
+    },
+    {
+      "epoch": 0.3801916932907348,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2789,
+      "step": 238
+    },
+    {
+      "epoch": 0.3817891373801917,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.2877,
+      "step": 239
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2614,
+      "step": 240
+    },
+    {
+      "epoch": 0.38498402555910544,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2491,
+      "step": 241
+    },
+    {
+      "epoch": 0.3865814696485623,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2494,
+      "step": 242
+    },
+    {
+      "epoch": 0.38817891373801916,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2728,
+      "step": 243
+    },
+    {
+      "epoch": 0.38977635782747605,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2759,
+      "step": 244
+    },
+    {
+      "epoch": 0.3913738019169329,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2795,
+      "step": 245
+    },
+    {
+      "epoch": 0.3929712460063898,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2761,
+      "step": 246
+    },
+    {
+      "epoch": 0.39456869009584666,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.2777,
+      "step": 247
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2729,
+      "step": 248
+    },
+    {
+      "epoch": 0.3977635782747604,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2576,
+      "step": 249
+    },
+    {
+      "epoch": 0.3993610223642173,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2693,
+      "step": 250
+    },
+    {
+      "epoch": 0.4009584664536741,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.275,
+      "step": 251
+    },
+    {
+      "epoch": 0.402555910543131,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.2864,
+      "step": 252
+    },
+    {
+      "epoch": 0.40415335463258784,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2694,
+      "step": 253
+    },
+    {
+      "epoch": 0.4057507987220447,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2527,
+      "step": 254
+    },
+    {
+      "epoch": 0.4073482428115016,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2652,
+      "step": 255
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2652,
+      "step": 256
+    },
+    {
+      "epoch": 0.41054313099041534,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2608,
+      "step": 257
+    },
+    {
+      "epoch": 0.41214057507987223,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.259,
+      "step": 258
+    },
+    {
+      "epoch": 0.41373801916932906,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2639,
+      "step": 259
+    },
+    {
+      "epoch": 0.41533546325878595,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2518,
+      "step": 260
+    },
+    {
+      "epoch": 0.4169329073482428,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.262,
+      "step": 261
+    },
+    {
+      "epoch": 0.4185303514376997,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2552,
+      "step": 262
+    },
+    {
+      "epoch": 0.42012779552715657,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2676,
+      "step": 263
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2714,
+      "step": 264
+    },
+    {
+      "epoch": 0.4233226837060703,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2573,
+      "step": 265
+    },
+    {
+      "epoch": 0.4249201277955272,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2589,
+      "step": 266
+    },
+    {
+      "epoch": 0.426517571884984,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.259,
+      "step": 267
+    },
+    {
+      "epoch": 0.4281150159744409,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2638,
+      "step": 268
+    },
+    {
+      "epoch": 0.42971246006389774,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.26,
+      "step": 269
+    },
+    {
+      "epoch": 0.43130990415335463,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2501,
+      "step": 270
+    },
+    {
+      "epoch": 0.4329073482428115,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2604,
+      "step": 271
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2709,
+      "step": 272
+    },
+    {
+      "epoch": 0.43610223642172524,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2619,
+      "step": 273
+    },
+    {
+      "epoch": 0.43769968051118213,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2524,
+      "step": 274
+    },
+    {
+      "epoch": 0.43929712460063897,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2523,
+      "step": 275
+    },
+    {
+      "epoch": 0.44089456869009586,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2679,
+      "step": 276
+    },
+    {
+      "epoch": 0.4424920127795527,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2602,
+      "step": 277
+    },
+    {
+      "epoch": 0.4440894568690096,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2411,
+      "step": 278
+    },
+    {
+      "epoch": 0.44568690095846647,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2453,
+      "step": 279
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.249,
+      "step": 280
+    },
+    {
+      "epoch": 0.4488817891373802,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2473,
+      "step": 281
+    },
+    {
+      "epoch": 0.4504792332268371,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2579,
+      "step": 282
+    },
+    {
+      "epoch": 0.4520766773162939,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2451,
+      "step": 283
+    },
+    {
+      "epoch": 0.4536741214057508,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2584,
+      "step": 284
+    },
+    {
+      "epoch": 0.45527156549520764,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2531,
+      "step": 285
+    },
+    {
+      "epoch": 0.45686900958466453,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2435,
+      "step": 286
+    },
+    {
+      "epoch": 0.4584664536741214,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2533,
+      "step": 287
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2606,
+      "step": 288
+    },
+    {
+      "epoch": 0.46166134185303515,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2414,
+      "step": 289
+    },
+    {
+      "epoch": 0.46325878594249204,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2306,
+      "step": 290
+    },
+    {
+      "epoch": 0.46485623003194887,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.24,
+      "step": 291
+    },
+    {
+      "epoch": 0.46645367412140576,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.249,
+      "step": 292
+    },
+    {
+      "epoch": 0.4680511182108626,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2594,
+      "step": 293
+    },
+    {
+      "epoch": 0.4696485623003195,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2494,
+      "step": 294
+    },
+    {
+      "epoch": 0.4712460063897764,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.255,
+      "step": 295
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2494,
+      "step": 296
+    },
+    {
+      "epoch": 0.4744408945686901,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2548,
+      "step": 297
+    },
+    {
+      "epoch": 0.476038338658147,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2556,
+      "step": 298
+    },
+    {
+      "epoch": 0.4776357827476038,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2378,
+      "step": 299
+    },
+    {
+      "epoch": 0.4792332268370607,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2482,
+      "step": 300
+    },
+    {
+      "epoch": 0.48083067092651754,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2494,
+      "step": 301
+    },
+    {
+      "epoch": 0.48242811501597443,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2491,
+      "step": 302
+    },
+    {
+      "epoch": 0.4840255591054313,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2386,
+      "step": 303
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2378,
+      "step": 304
+    },
+    {
+      "epoch": 0.48722044728434505,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2438,
+      "step": 305
+    },
+    {
+      "epoch": 0.48881789137380194,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2513,
+      "step": 306
+    },
+    {
+      "epoch": 0.4904153354632588,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.244,
+      "step": 307
+    },
+    {
+      "epoch": 0.49201277955271566,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2544,
+      "step": 308
+    },
+    {
+      "epoch": 0.4936102236421725,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2487,
+      "step": 309
+    },
+    {
+      "epoch": 0.4952076677316294,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2406,
+      "step": 310
+    },
+    {
+      "epoch": 0.4968051118210863,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2449,
+      "step": 311
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2574,
+      "step": 312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2479,
+      "step": 313
+    },
+    {
+      "epoch": 0.5015974440894568,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2544,
+      "step": 314
+    },
+    {
+      "epoch": 0.5031948881789138,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2539,
+      "step": 315
+    },
+    {
+      "epoch": 0.5047923322683706,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.2494,
+      "step": 316
+    },
+    {
+      "epoch": 0.5063897763578274,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2566,
+      "step": 317
+    },
+    {
+      "epoch": 0.5079872204472844,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2336,
+      "step": 318
+    },
+    {
+      "epoch": 0.5095846645367412,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2422,
+      "step": 319
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2523,
+      "step": 320
+    },
+    {
+      "epoch": 0.512779552715655,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2587,
+      "step": 321
+    },
+    {
+      "epoch": 0.5143769968051118,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.236,
+      "step": 322
+    },
+    {
+      "epoch": 0.5159744408945687,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2437,
+      "step": 323
+    },
+    {
+      "epoch": 0.5175718849840255,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2539,
+      "step": 324
+    },
+    {
+      "epoch": 0.5191693290734825,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.246,
+      "step": 325
+    },
+    {
+      "epoch": 0.5207667731629393,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2448,
+      "step": 326
+    },
+    {
+      "epoch": 0.5223642172523961,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2526,
+      "step": 327
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2489,
+      "step": 328
+    },
+    {
+      "epoch": 0.5255591054313099,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2433,
+      "step": 329
+    },
+    {
+      "epoch": 0.5271565495207667,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2434,
+      "step": 330
+    },
+    {
+      "epoch": 0.5287539936102237,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2508,
+      "step": 331
+    },
+    {
+      "epoch": 0.5303514376996805,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2312,
+      "step": 332
+    },
+    {
+      "epoch": 0.5319488817891374,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2311,
+      "step": 333
+    },
+    {
+      "epoch": 0.5335463258785943,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2363,
+      "step": 334
+    },
+    {
+      "epoch": 0.5351437699680511,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.2403,
+      "step": 335
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2487,
+      "step": 336
+    },
+    {
+      "epoch": 0.5383386581469649,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.232,
+      "step": 337
+    },
+    {
+      "epoch": 0.5399361022364217,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2562,
+      "step": 338
+    },
+    {
+      "epoch": 0.5415335463258786,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2409,
+      "step": 339
+    },
+    {
+      "epoch": 0.5431309904153354,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2453,
+      "step": 340
+    },
+    {
+      "epoch": 0.5447284345047924,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2379,
+      "step": 341
+    },
+    {
+      "epoch": 0.5463258785942492,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2415,
+      "step": 342
+    },
+    {
+      "epoch": 0.547923322683706,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2387,
+      "step": 343
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2519,
+      "step": 344
+    },
+    {
+      "epoch": 0.5511182108626198,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2419,
+      "step": 345
+    },
+    {
+      "epoch": 0.5527156549520766,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.237,
+      "step": 346
+    },
+    {
+      "epoch": 0.5543130990415336,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.231,
+      "step": 347
+    },
+    {
+      "epoch": 0.5559105431309904,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2388,
+      "step": 348
+    },
+    {
+      "epoch": 0.5575079872204473,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2432,
+      "step": 349
+    },
+    {
+      "epoch": 0.5591054313099042,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.246,
+      "step": 350
+    },
+    {
+      "epoch": 0.560702875399361,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.241,
+      "step": 351
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2322,
+      "step": 352
+    },
+    {
+      "epoch": 0.5638977635782748,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.242,
+      "step": 353
+    },
+    {
+      "epoch": 0.5654952076677316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2312,
+      "step": 354
+    },
+    {
+      "epoch": 0.5670926517571885,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2487,
+      "step": 355
+    },
+    {
+      "epoch": 0.5686900958466453,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2512,
+      "step": 356
+    },
+    {
+      "epoch": 0.5702875399361023,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2431,
+      "step": 357
+    },
+    {
+      "epoch": 0.5718849840255591,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2406,
+      "step": 358
+    },
+    {
+      "epoch": 0.5734824281150159,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2461,
+      "step": 359
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2397,
+      "step": 360
+    },
+    {
+      "epoch": 0.5766773162939297,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2357,
+      "step": 361
+    },
+    {
+      "epoch": 0.5782747603833865,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2474,
+      "step": 362
+    },
+    {
+      "epoch": 0.5798722044728435,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2354,
+      "step": 363
+    },
+    {
+      "epoch": 0.5814696485623003,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.234,
+      "step": 364
+    },
+    {
+      "epoch": 0.5830670926517572,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2246,
+      "step": 365
+    },
+    {
+      "epoch": 0.5846645367412141,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2457,
+      "step": 366
+    },
+    {
+      "epoch": 0.5862619808306709,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2301,
+      "step": 367
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2283,
+      "step": 368
+    },
+    {
+      "epoch": 0.5894568690095847,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2272,
+      "step": 369
+    },
+    {
+      "epoch": 0.5910543130990416,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2276,
+      "step": 370
+    },
+    {
+      "epoch": 0.5926517571884984,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2374,
+      "step": 371
+    },
+    {
+      "epoch": 0.5942492012779552,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2389,
+      "step": 372
+    },
+    {
+      "epoch": 0.5958466453674122,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2279,
+      "step": 373
+    },
+    {
+      "epoch": 0.597444089456869,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2157,
+      "step": 374
+    },
+    {
+      "epoch": 0.5990415335463258,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.247,
+      "step": 375
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2437,
+      "step": 376
+    },
+    {
+      "epoch": 0.6022364217252396,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2384,
+      "step": 377
+    },
+    {
+      "epoch": 0.6038338658146964,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2211,
+      "step": 378
+    },
+    {
+      "epoch": 0.6054313099041534,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2165,
+      "step": 379
+    },
+    {
+      "epoch": 0.6070287539936102,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2322,
+      "step": 380
+    },
+    {
+      "epoch": 0.6086261980830671,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2247,
+      "step": 381
+    },
+    {
+      "epoch": 0.610223642172524,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2296,
+      "step": 382
+    },
+    {
+      "epoch": 0.6118210862619808,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2247,
+      "step": 383
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2332,
+      "step": 384
+    },
+    {
+      "epoch": 0.6150159744408946,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2198,
+      "step": 385
+    },
+    {
+      "epoch": 0.6166134185303515,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2277,
+      "step": 386
+    },
+    {
+      "epoch": 0.6182108626198083,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2314,
+      "step": 387
+    },
+    {
+      "epoch": 0.6198083067092651,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2476,
+      "step": 388
+    },
+    {
+      "epoch": 0.6214057507987221,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2306,
+      "step": 389
+    },
+    {
+      "epoch": 0.6230031948881789,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2349,
+      "step": 390
+    },
+    {
+      "epoch": 0.6246006389776357,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2435,
+      "step": 391
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2416,
+      "step": 392
+    },
+    {
+      "epoch": 0.6277955271565495,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.251,
+      "step": 393
+    },
+    {
+      "epoch": 0.6293929712460063,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2508,
+      "step": 394
+    },
+    {
+      "epoch": 0.6309904153354633,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2359,
+      "step": 395
+    },
+    {
+      "epoch": 0.6325878594249201,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2269,
+      "step": 396
+    },
+    {
+      "epoch": 0.634185303514377,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.2446,
+      "step": 397
+    },
+    {
+      "epoch": 0.6357827476038339,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2332,
+      "step": 398
+    },
+    {
+      "epoch": 0.6373801916932907,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2298,
+      "step": 399
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2717,
+      "step": 400
+    },
+    {
+      "epoch": 0.6405750798722045,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2382,
+      "step": 401
+    },
+    {
+      "epoch": 0.6421725239616614,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2411,
+      "step": 402
+    },
+    {
+      "epoch": 0.6437699680511182,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0005,
+      "loss": 1.2314,
+      "step": 403
+    },
+    {
+      "epoch": 0.645367412140575,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.257,
+      "step": 404
+    },
+    {
+      "epoch": 0.646964856230032,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2554,
+      "step": 405
+    },
+    {
+      "epoch": 0.6485623003194888,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2387,
+      "step": 406
+    },
+    {
+      "epoch": 0.6501597444089456,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2366,
+      "step": 407
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2356,
+      "step": 408
+    },
+    {
+      "epoch": 0.6533546325878594,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2411,
+      "step": 409
+    },
+    {
+      "epoch": 0.6549520766773163,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2412,
+      "step": 410
+    },
+    {
+      "epoch": 0.6565495207667732,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2377,
+      "step": 411
+    },
+    {
+      "epoch": 0.65814696485623,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2367,
+      "step": 412
+    },
+    {
+      "epoch": 0.6597444089456869,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.2293,
+      "step": 413
+    },
+    {
+      "epoch": 0.6613418530351438,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.226,
+      "step": 414
+    },
+    {
+      "epoch": 0.6629392971246006,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.249,
+      "step": 415
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2318,
+      "step": 416
+    },
+    {
+      "epoch": 0.6661341853035144,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2394,
+      "step": 417
+    },
+    {
+      "epoch": 0.6677316293929713,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2215,
+      "step": 418
+    },
+    {
+      "epoch": 0.6693290734824281,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2394,
+      "step": 419
+    },
+    {
+      "epoch": 0.670926517571885,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.23,
+      "step": 420
+    },
+    {
+      "epoch": 0.6725239616613419,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2265,
+      "step": 421
+    },
+    {
+      "epoch": 0.6741214057507987,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2126,
+      "step": 422
+    },
+    {
+      "epoch": 0.6757188498402555,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2193,
+      "step": 423
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2248,
+      "step": 424
+    },
+    {
+      "epoch": 0.6789137380191693,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.2108,
+      "step": 425
+    },
+    {
+      "epoch": 0.6805111821086262,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2292,
+      "step": 426
+    },
+    {
+      "epoch": 0.6821086261980831,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.221,
+      "step": 427
+    },
+    {
+      "epoch": 0.6837060702875399,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2346,
+      "step": 428
+    },
+    {
+      "epoch": 0.6853035143769968,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.225,
+      "step": 429
+    },
+    {
+      "epoch": 0.6869009584664537,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.2268,
+      "step": 430
+    },
+    {
+      "epoch": 0.6884984025559105,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.229,
+      "step": 431
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2245,
+      "step": 432
+    },
+    {
+      "epoch": 0.6916932907348243,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.236,
+      "step": 433
+    },
+    {
+      "epoch": 0.6932907348242812,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2388,
+      "step": 434
+    },
+    {
+      "epoch": 0.694888178913738,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2175,
+      "step": 435
+    },
+    {
+      "epoch": 0.6964856230031949,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2129,
+      "step": 436
+    },
+    {
+      "epoch": 0.6980830670926518,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2151,
+      "step": 437
+    },
+    {
+      "epoch": 0.6996805111821086,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2211,
+      "step": 438
+    },
+    {
+      "epoch": 0.7012779552715654,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2271,
+      "step": 439
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2273,
+      "step": 440
+    },
+    {
+      "epoch": 0.7044728434504792,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2233,
+      "step": 441
+    },
+    {
+      "epoch": 0.7060702875399361,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2227,
+      "step": 442
+    },
+    {
+      "epoch": 0.707667731629393,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2495,
+      "step": 443
+    },
+    {
+      "epoch": 0.7092651757188498,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2402,
+      "step": 444
+    },
+    {
+      "epoch": 0.7108626198083067,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2351,
+      "step": 445
+    },
+    {
+      "epoch": 0.7124600638977636,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2172,
+      "step": 446
+    },
+    {
+      "epoch": 0.7140575079872205,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.227,
+      "step": 447
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2313,
+      "step": 448
+    },
+    {
+      "epoch": 0.7172523961661342,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.227,
+      "step": 449
+    },
+    {
+      "epoch": 0.7188498402555911,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2289,
+      "step": 450
+    },
+    {
+      "epoch": 0.7204472843450479,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2241,
+      "step": 451
+    },
+    {
+      "epoch": 0.7220447284345048,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2323,
+      "step": 452
+    },
+    {
+      "epoch": 0.7236421725239617,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2226,
+      "step": 453
+    },
+    {
+      "epoch": 0.7252396166134185,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2202,
+      "step": 454
+    },
+    {
+      "epoch": 0.7268370607028753,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2306,
+      "step": 455
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.2279,
+      "step": 456
+    },
+    {
+      "epoch": 0.7300319488817891,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2101,
+      "step": 457
+    },
+    {
+      "epoch": 0.731629392971246,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2256,
+      "step": 458
+    },
+    {
+      "epoch": 0.7332268370607029,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2226,
+      "step": 459
+    },
+    {
+      "epoch": 0.7348242811501597,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.226,
+      "step": 460
+    },
+    {
+      "epoch": 0.7364217252396166,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2322,
+      "step": 461
+    },
+    {
+      "epoch": 0.7380191693290735,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2144,
+      "step": 462
+    },
+    {
+      "epoch": 0.7396166134185304,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2206,
+      "step": 463
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2253,
+      "step": 464
+    },
+    {
+      "epoch": 0.7428115015974441,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2287,
+      "step": 465
+    },
+    {
+      "epoch": 0.744408945686901,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.227,
+      "step": 466
+    },
+    {
+      "epoch": 0.7460063897763578,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2251,
+      "step": 467
+    },
+    {
+      "epoch": 0.7476038338658147,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2191,
+      "step": 468
+    },
+    {
+      "epoch": 0.7492012779552716,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2219,
+      "step": 469
+    },
+    {
+      "epoch": 0.7507987220447284,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2282,
+      "step": 470
+    },
+    {
+      "epoch": 0.7523961661341853,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.227,
+      "step": 471
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2266,
+      "step": 472
+    },
+    {
+      "epoch": 0.755591054313099,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2343,
+      "step": 473
+    },
+    {
+      "epoch": 0.7571884984025559,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2186,
+      "step": 474
+    },
+    {
+      "epoch": 0.7587859424920128,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.219,
+      "step": 475
+    },
+    {
+      "epoch": 0.7603833865814696,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2134,
+      "step": 476
+    },
+    {
+      "epoch": 0.7619808306709265,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2156,
+      "step": 477
+    },
+    {
+      "epoch": 0.7635782747603834,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2204,
+      "step": 478
+    },
+    {
+      "epoch": 0.7651757188498403,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2169,
+      "step": 479
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2182,
+      "step": 480
+    },
+    {
+      "epoch": 0.768370607028754,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2304,
+      "step": 481
+    },
+    {
+      "epoch": 0.7699680511182109,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.229,
+      "step": 482
+    },
+    {
+      "epoch": 0.7715654952076677,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2234,
+      "step": 483
+    },
+    {
+      "epoch": 0.7731629392971247,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2345,
+      "step": 484
+    },
+    {
+      "epoch": 0.7747603833865815,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2168,
+      "step": 485
+    },
+    {
+      "epoch": 0.7763578274760383,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2089,
+      "step": 486
+    },
+    {
+      "epoch": 0.7779552715654952,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.217,
+      "step": 487
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2131,
+      "step": 488
+    },
+    {
+      "epoch": 0.7811501597444089,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2154,
+      "step": 489
+    },
+    {
+      "epoch": 0.7827476038338658,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2292,
+      "step": 490
+    },
+    {
+      "epoch": 0.7843450479233227,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2198,
+      "step": 491
+    },
+    {
+      "epoch": 0.7859424920127795,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2136,
+      "step": 492
+    },
+    {
+      "epoch": 0.7875399361022364,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2181,
+      "step": 493
+    },
+    {
+      "epoch": 0.7891373801916933,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2242,
+      "step": 494
+    },
+    {
+      "epoch": 0.7907348242811502,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2162,
+      "step": 495
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2111,
+      "step": 496
+    },
+    {
+      "epoch": 0.7939297124600639,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2317,
+      "step": 497
+    },
+    {
+      "epoch": 0.7955271565495208,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2091,
+      "step": 498
+    },
+    {
+      "epoch": 0.7971246006389776,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2159,
+      "step": 499
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2098,
+      "step": 500
+    },
+    {
+      "epoch": 0.8003194888178914,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2047,
+      "step": 501
+    },
+    {
+      "epoch": 0.8019169329073482,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2196,
+      "step": 502
+    },
+    {
+      "epoch": 0.8035143769968051,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2331,
+      "step": 503
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.219,
+      "step": 504
+    },
+    {
+      "epoch": 0.8067092651757188,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2203,
+      "step": 505
+    },
+    {
+      "epoch": 0.8083067092651757,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2112,
+      "step": 506
+    },
+    {
+      "epoch": 0.8099041533546326,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2189,
+      "step": 507
+    },
+    {
+      "epoch": 0.8115015974440895,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2205,
+      "step": 508
+    },
+    {
+      "epoch": 0.8130990415335463,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.204,
+      "step": 509
+    },
+    {
+      "epoch": 0.8146964856230032,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2242,
+      "step": 510
+    },
+    {
+      "epoch": 0.8162939297124601,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2271,
+      "step": 511
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2127,
+      "step": 512
+    },
+    {
+      "epoch": 0.8194888178913738,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2205,
+      "step": 513
+    },
+    {
+      "epoch": 0.8210862619808307,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2147,
+      "step": 514
+    },
+    {
+      "epoch": 0.8226837060702875,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.224,
+      "step": 515
+    },
+    {
+      "epoch": 0.8242811501597445,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.222,
+      "step": 516
+    },
+    {
+      "epoch": 0.8258785942492013,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2174,
+      "step": 517
+    },
+    {
+      "epoch": 0.8274760383386581,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2205,
+      "step": 518
+    },
+    {
+      "epoch": 0.829073482428115,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2308,
+      "step": 519
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2084,
+      "step": 520
+    },
+    {
+      "epoch": 0.8322683706070287,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2305,
+      "step": 521
+    },
+    {
+      "epoch": 0.8338658146964856,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2188,
+      "step": 522
+    },
+    {
+      "epoch": 0.8354632587859425,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2264,
+      "step": 523
+    },
+    {
+      "epoch": 0.8370607028753994,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.228,
+      "step": 524
+    },
+    {
+      "epoch": 0.8386581469648562,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2383,
+      "step": 525
+    },
+    {
+      "epoch": 0.8402555910543131,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2329,
+      "step": 526
+    },
+    {
+      "epoch": 0.84185303514377,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2319,
+      "step": 527
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2113,
+      "step": 528
+    },
+    {
+      "epoch": 0.8450479233226837,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2103,
+      "step": 529
+    },
+    {
+      "epoch": 0.8466453674121406,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.227,
+      "step": 530
+    },
+    {
+      "epoch": 0.8482428115015974,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2252,
+      "step": 531
+    },
+    {
+      "epoch": 0.8498402555910544,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2167,
+      "step": 532
+    },
+    {
+      "epoch": 0.8514376996805112,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2091,
+      "step": 533
+    },
+    {
+      "epoch": 0.853035143769968,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0005,
+      "loss": 1.205,
+      "step": 534
+    },
+    {
+      "epoch": 0.854632587859425,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2218,
+      "step": 535
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2282,
+      "step": 536
+    },
+    {
+      "epoch": 0.8578274760383386,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.2141,
+      "step": 537
+    },
+    {
+      "epoch": 0.8594249201277955,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2137,
+      "step": 538
+    },
+    {
+      "epoch": 0.8610223642172524,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.2134,
+      "step": 539
+    },
+    {
+      "epoch": 0.8626198083067093,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.211,
+      "step": 540
+    },
+    {
+      "epoch": 0.8642172523961661,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1993,
+      "step": 541
+    },
+    {
+      "epoch": 0.865814696485623,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2254,
+      "step": 542
+    },
+    {
+      "epoch": 0.8674121405750799,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2069,
+      "step": 543
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2111,
+      "step": 544
+    },
+    {
+      "epoch": 0.8706070287539937,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2177,
+      "step": 545
+    },
+    {
+      "epoch": 0.8722044728434505,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2157,
+      "step": 546
+    },
+    {
+      "epoch": 0.8738019169329073,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2055,
+      "step": 547
+    },
+    {
+      "epoch": 0.8753993610223643,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1984,
+      "step": 548
+    },
+    {
+      "epoch": 0.8769968051118211,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2235,
+      "step": 549
+    },
+    {
+      "epoch": 0.8785942492012779,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2257,
+      "step": 550
+    },
+    {
+      "epoch": 0.8801916932907349,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2255,
+      "step": 551
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2034,
+      "step": 552
+    },
+    {
+      "epoch": 0.8833865814696485,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.217,
+      "step": 553
+    },
+    {
+      "epoch": 0.8849840255591054,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2265,
+      "step": 554
+    },
+    {
+      "epoch": 0.8865814696485623,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2065,
+      "step": 555
+    },
+    {
+      "epoch": 0.8881789137380192,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2235,
+      "step": 556
+    },
+    {
+      "epoch": 0.889776357827476,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2217,
+      "step": 557
+    },
+    {
+      "epoch": 0.8913738019169329,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2213,
+      "step": 558
+    },
+    {
+      "epoch": 0.8929712460063898,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2185,
+      "step": 559
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2142,
+      "step": 560
+    },
+    {
+      "epoch": 0.8961661341853036,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2183,
+      "step": 561
+    },
+    {
+      "epoch": 0.8977635782747604,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2107,
+      "step": 562
+    },
+    {
+      "epoch": 0.8993610223642172,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.223,
+      "step": 563
+    },
+    {
+      "epoch": 0.9009584664536742,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2105,
+      "step": 564
+    },
+    {
+      "epoch": 0.902555910543131,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2182,
+      "step": 565
+    },
+    {
+      "epoch": 0.9041533546325878,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2147,
+      "step": 566
+    },
+    {
+      "epoch": 0.9057507987220448,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2023,
+      "step": 567
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2016,
+      "step": 568
+    },
+    {
+      "epoch": 0.9089456869009584,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2181,
+      "step": 569
+    },
+    {
+      "epoch": 0.9105431309904153,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2144,
+      "step": 570
+    },
+    {
+      "epoch": 0.9121405750798722,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2062,
+      "step": 571
+    },
+    {
+      "epoch": 0.9137380191693291,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2125,
+      "step": 572
+    },
+    {
+      "epoch": 0.9153354632587859,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2016,
+      "step": 573
+    },
+    {
+      "epoch": 0.9169329073482428,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2139,
+      "step": 574
+    },
+    {
+      "epoch": 0.9185303514376997,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2092,
+      "step": 575
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.2202,
+      "step": 576
+    },
+    {
+      "epoch": 0.9217252396166135,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2066,
+      "step": 577
+    },
+    {
+      "epoch": 0.9233226837060703,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1978,
+      "step": 578
+    },
+    {
+      "epoch": 0.9249201277955271,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2142,
+      "step": 579
+    },
+    {
+      "epoch": 0.9265175718849841,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2219,
+      "step": 580
+    },
+    {
+      "epoch": 0.9281150159744409,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2047,
+      "step": 581
+    },
+    {
+      "epoch": 0.9297124600638977,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2041,
+      "step": 582
+    },
+    {
+      "epoch": 0.9313099041533547,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2141,
+      "step": 583
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2029,
+      "step": 584
+    },
+    {
+      "epoch": 0.9345047923322684,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2095,
+      "step": 585
+    },
+    {
+      "epoch": 0.9361022364217252,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2021,
+      "step": 586
+    },
+    {
+      "epoch": 0.9376996805111821,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2047,
+      "step": 587
+    },
+    {
+      "epoch": 0.939297124600639,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2173,
+      "step": 588
+    },
+    {
+      "epoch": 0.9408945686900958,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2114,
+      "step": 589
+    },
+    {
+      "epoch": 0.9424920127795527,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2081,
+      "step": 590
+    },
+    {
+      "epoch": 0.9440894568690096,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2155,
+      "step": 591
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2206,
+      "step": 592
+    },
+    {
+      "epoch": 0.9472843450479234,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.2123,
+      "step": 593
+    },
+    {
+      "epoch": 0.9488817891373802,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2045,
+      "step": 594
+    },
+    {
+      "epoch": 0.950479233226837,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.203,
+      "step": 595
+    },
+    {
+      "epoch": 0.952076677316294,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2062,
+      "step": 596
+    },
+    {
+      "epoch": 0.9536741214057508,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2151,
+      "step": 597
+    },
+    {
+      "epoch": 0.9552715654952076,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2032,
+      "step": 598
+    },
+    {
+      "epoch": 0.9568690095846646,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2034,
+      "step": 599
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2022,
+      "step": 600
+    },
+    {
+      "epoch": 0.9600638977635783,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2153,
+      "step": 601
+    },
+    {
+      "epoch": 0.9616613418530351,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2076,
+      "step": 602
+    },
+    {
+      "epoch": 0.963258785942492,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2076,
+      "step": 603
+    },
+    {
+      "epoch": 0.9648562300319489,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2027,
+      "step": 604
+    },
+    {
+      "epoch": 0.9664536741214057,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2122,
+      "step": 605
+    },
+    {
+      "epoch": 0.9680511182108626,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2112,
+      "step": 606
+    },
+    {
+      "epoch": 0.9696485623003195,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2106,
+      "step": 607
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2122,
+      "step": 608
+    },
+    {
+      "epoch": 0.9728434504792333,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2075,
+      "step": 609
+    },
+    {
+      "epoch": 0.9744408945686901,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2229,
+      "step": 610
+    },
+    {
+      "epoch": 0.9760383386581469,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1965,
+      "step": 611
+    },
+    {
+      "epoch": 0.9776357827476039,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2247,
+      "step": 612
+    },
+    {
+      "epoch": 0.9792332268370607,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2146,
+      "step": 613
+    },
+    {
+      "epoch": 0.9808306709265175,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2017,
+      "step": 614
+    },
+    {
+      "epoch": 0.9824281150159745,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2085,
+      "step": 615
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2009,
+      "step": 616
+    },
+    {
+      "epoch": 0.9856230031948882,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2161,
+      "step": 617
+    },
+    {
+      "epoch": 0.987220447284345,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1982,
+      "step": 618
+    },
+    {
+      "epoch": 0.9888178913738019,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1993,
+      "step": 619
+    },
+    {
+      "epoch": 0.9904153354632588,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2067,
+      "step": 620
+    },
+    {
+      "epoch": 0.9920127795527156,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2065,
+      "step": 621
+    },
+    {
+      "epoch": 0.9936102236421726,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2101,
+      "step": 622
+    },
+    {
+      "epoch": 0.9952076677316294,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2232,
+      "step": 623
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2096,
+      "step": 624
+    },
+    {
+      "epoch": 0.9984025559105432,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2144,
+      "step": 625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2086,
+      "step": 626
+    },
+    {
+      "epoch": 1.0015974440894568,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2182,
+      "step": 627
+    },
+    {
+      "epoch": 1.0031948881789137,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2042,
+      "step": 628
+    },
+    {
+      "epoch": 1.0047923322683705,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2052,
+      "step": 629
+    },
+    {
+      "epoch": 1.0063897763578276,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2159,
+      "step": 630
+    },
+    {
+      "epoch": 1.0079872204472844,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.208,
+      "step": 631
+    },
+    {
+      "epoch": 1.0095846645367412,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2046,
+      "step": 632
+    },
+    {
+      "epoch": 1.011182108626198,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2092,
+      "step": 633
+    },
+    {
+      "epoch": 1.012779552715655,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1997,
+      "step": 634
+    },
+    {
+      "epoch": 1.0143769968051117,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1973,
+      "step": 635
+    },
+    {
+      "epoch": 1.0159744408945688,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2175,
+      "step": 636
+    },
+    {
+      "epoch": 1.0175718849840256,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2127,
+      "step": 637
+    },
+    {
+      "epoch": 1.0191693290734825,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2056,
+      "step": 638
+    },
+    {
+      "epoch": 1.0207667731629393,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1909,
+      "step": 639
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1979,
+      "step": 640
+    },
+    {
+      "epoch": 1.023961661341853,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2121,
+      "step": 641
+    },
+    {
+      "epoch": 1.0255591054313098,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2029,
+      "step": 642
+    },
+    {
+      "epoch": 1.0271565495207668,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1861,
+      "step": 643
+    },
+    {
+      "epoch": 1.0287539936102237,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2106,
+      "step": 644
+    },
+    {
+      "epoch": 1.0303514376996805,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1938,
+      "step": 645
+    },
+    {
+      "epoch": 1.0319488817891374,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1909,
+      "step": 646
+    },
+    {
+      "epoch": 1.0335463258785942,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2165,
+      "step": 647
+    },
+    {
+      "epoch": 1.035143769968051,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.2302,
+      "step": 648
+    },
+    {
+      "epoch": 1.036741214057508,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2243,
+      "step": 649
+    },
+    {
+      "epoch": 1.038338658146965,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.215,
+      "step": 650
+    },
+    {
+      "epoch": 1.0399361022364217,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0005,
+      "loss": 1.2044,
+      "step": 651
+    },
+    {
+      "epoch": 1.0415335463258786,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2108,
+      "step": 652
+    },
+    {
+      "epoch": 1.0431309904153354,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2042,
+      "step": 653
+    },
+    {
+      "epoch": 1.0447284345047922,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2202,
+      "step": 654
+    },
+    {
+      "epoch": 1.0463258785942493,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.214,
+      "step": 655
+    },
+    {
+      "epoch": 1.0479233226837061,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2131,
+      "step": 656
+    },
+    {
+      "epoch": 1.049520766773163,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2079,
+      "step": 657
+    },
+    {
+      "epoch": 1.0511182108626198,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.223,
+      "step": 658
+    },
+    {
+      "epoch": 1.0527156549520766,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0005,
+      "loss": 1.222,
+      "step": 659
+    },
+    {
+      "epoch": 1.0543130990415335,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2164,
+      "step": 660
+    },
+    {
+      "epoch": 1.0559105431309903,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2098,
+      "step": 661
+    },
+    {
+      "epoch": 1.0575079872204474,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1977,
+      "step": 662
+    },
+    {
+      "epoch": 1.0591054313099042,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2081,
+      "step": 663
+    },
+    {
+      "epoch": 1.060702875399361,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2083,
+      "step": 664
+    },
+    {
+      "epoch": 1.0623003194888179,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2028,
+      "step": 665
+    },
+    {
+      "epoch": 1.0638977635782747,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0005,
+      "loss": 1.218,
+      "step": 666
+    },
+    {
+      "epoch": 1.0654952076677315,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2138,
+      "step": 667
+    },
+    {
+      "epoch": 1.0670926517571886,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0005,
+      "loss": 1.203,
+      "step": 668
+    },
+    {
+      "epoch": 1.0686900958466454,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.2083,
+      "step": 669
+    },
+    {
+      "epoch": 1.0702875399361023,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2146,
+      "step": 670
+    },
+    {
+      "epoch": 1.071884984025559,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.21,
+      "step": 671
+    },
+    {
+      "epoch": 1.073482428115016,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.211,
+      "step": 672
+    },
+    {
+      "epoch": 1.0750798722044728,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 673
+    },
+    {
+      "epoch": 1.0766773162939298,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0005,
+      "loss": 1.2271,
+      "step": 674
+    },
+    {
+      "epoch": 1.0782747603833867,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2192,
+      "step": 675
+    },
+    {
+      "epoch": 1.0798722044728435,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2118,
+      "step": 676
+    },
+    {
+      "epoch": 1.0814696485623003,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2262,
+      "step": 677
+    },
+    {
+      "epoch": 1.0830670926517572,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2165,
+      "step": 678
+    },
+    {
+      "epoch": 1.084664536741214,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2024,
+      "step": 679
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2081,
+      "step": 680
+    },
+    {
+      "epoch": 1.0878594249201279,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2101,
+      "step": 681
+    },
+    {
+      "epoch": 1.0894568690095847,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.204,
+      "step": 682
+    },
+    {
+      "epoch": 1.0910543130990416,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2047,
+      "step": 683
+    },
+    {
+      "epoch": 1.0926517571884984,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2137,
+      "step": 684
+    },
+    {
+      "epoch": 1.0942492012779552,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2151,
+      "step": 685
+    },
+    {
+      "epoch": 1.095846645367412,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2164,
+      "step": 686
+    },
+    {
+      "epoch": 1.097444089456869,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2033,
+      "step": 687
+    },
+    {
+      "epoch": 1.099041533546326,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2044,
+      "step": 688
+    },
+    {
+      "epoch": 1.1006389776357828,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2042,
+      "step": 689
+    },
+    {
+      "epoch": 1.1022364217252396,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2121,
+      "step": 690
+    },
+    {
+      "epoch": 1.1038338658146964,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2062,
+      "step": 691
+    },
+    {
+      "epoch": 1.1054313099041533,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2,
+      "step": 692
+    },
+    {
+      "epoch": 1.1070287539936103,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1979,
+      "step": 693
+    },
+    {
+      "epoch": 1.1086261980830672,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2033,
+      "step": 694
+    },
+    {
+      "epoch": 1.110223642172524,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2053,
+      "step": 695
+    },
+    {
+      "epoch": 1.1118210862619808,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.191,
+      "step": 696
+    },
+    {
+      "epoch": 1.1134185303514377,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2108,
+      "step": 697
+    },
+    {
+      "epoch": 1.1150159744408945,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.209,
+      "step": 698
+    },
+    {
+      "epoch": 1.1166134185303513,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2103,
+      "step": 699
+    },
+    {
+      "epoch": 1.1182108626198084,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2089,
+      "step": 700
+    },
+    {
+      "epoch": 1.1198083067092652,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2077,
+      "step": 701
+    },
+    {
+      "epoch": 1.121405750798722,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2107,
+      "step": 702
+    },
+    {
+      "epoch": 1.123003194888179,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1997,
+      "step": 703
+    },
+    {
+      "epoch": 1.1246006389776357,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.203,
+      "step": 704
+    },
+    {
+      "epoch": 1.1261980830670926,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2064,
+      "step": 705
+    },
+    {
+      "epoch": 1.1277955271565494,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2079,
+      "step": 706
+    },
+    {
+      "epoch": 1.1293929712460065,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1951,
+      "step": 707
+    },
+    {
+      "epoch": 1.1309904153354633,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2044,
+      "step": 708
+    },
+    {
+      "epoch": 1.1325878594249201,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1963,
+      "step": 709
+    },
+    {
+      "epoch": 1.134185303514377,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2086,
+      "step": 710
+    },
+    {
+      "epoch": 1.1357827476038338,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2021,
+      "step": 711
+    },
+    {
+      "epoch": 1.1373801916932909,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2104,
+      "step": 712
+    },
+    {
+      "epoch": 1.1389776357827477,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2088,
+      "step": 713
+    },
+    {
+      "epoch": 1.1405750798722045,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1967,
+      "step": 714
+    },
+    {
+      "epoch": 1.1421725239616614,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1975,
+      "step": 715
+    },
+    {
+      "epoch": 1.1437699680511182,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2039,
+      "step": 716
+    },
+    {
+      "epoch": 1.145367412140575,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2157,
+      "step": 717
+    },
+    {
+      "epoch": 1.1469648562300319,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1995,
+      "step": 718
+    },
+    {
+      "epoch": 1.148562300319489,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1986,
+      "step": 719
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2016,
+      "step": 720
+    },
+    {
+      "epoch": 1.1517571884984026,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1961,
+      "step": 721
+    },
+    {
+      "epoch": 1.1533546325878594,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1998,
+      "step": 722
+    },
+    {
+      "epoch": 1.1549520766773163,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2137,
+      "step": 723
+    },
+    {
+      "epoch": 1.156549520766773,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2079,
+      "step": 724
+    },
+    {
+      "epoch": 1.15814696485623,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1933,
+      "step": 725
+    },
+    {
+      "epoch": 1.159744408945687,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2043,
+      "step": 726
+    },
+    {
+      "epoch": 1.1613418530351438,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1961,
+      "step": 727
+    },
+    {
+      "epoch": 1.1629392971246006,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.203,
+      "step": 728
+    },
+    {
+      "epoch": 1.1645367412140575,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2041,
+      "step": 729
+    },
+    {
+      "epoch": 1.1661341853035143,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2056,
+      "step": 730
+    },
+    {
+      "epoch": 1.1677316293929714,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2096,
+      "step": 731
+    },
+    {
+      "epoch": 1.1693290734824282,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1992,
+      "step": 732
+    },
+    {
+      "epoch": 1.170926517571885,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2091,
+      "step": 733
+    },
+    {
+      "epoch": 1.1725239616613419,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1962,
+      "step": 734
+    },
+    {
+      "epoch": 1.1741214057507987,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2127,
+      "step": 735
+    },
+    {
+      "epoch": 1.1757188498402555,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2033,
+      "step": 736
+    },
+    {
+      "epoch": 1.1773162939297124,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1925,
+      "step": 737
+    },
+    {
+      "epoch": 1.1789137380191694,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1968,
+      "step": 738
+    },
+    {
+      "epoch": 1.1805111821086263,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1955,
+      "step": 739
+    },
+    {
+      "epoch": 1.182108626198083,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2037,
+      "step": 740
+    },
+    {
+      "epoch": 1.18370607028754,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2036,
+      "step": 741
+    },
+    {
+      "epoch": 1.1853035143769968,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2045,
+      "step": 742
+    },
+    {
+      "epoch": 1.1869009584664536,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2038,
+      "step": 743
+    },
+    {
+      "epoch": 1.1884984025559104,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1948,
+      "step": 744
+    },
+    {
+      "epoch": 1.1900958466453675,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2123,
+      "step": 745
+    },
+    {
+      "epoch": 1.1916932907348243,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1914,
+      "step": 746
+    },
+    {
+      "epoch": 1.1932907348242812,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 747
+    },
+    {
+      "epoch": 1.194888178913738,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1951,
+      "step": 748
+    },
+    {
+      "epoch": 1.1964856230031948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2089,
+      "step": 749
+    },
+    {
+      "epoch": 1.1980830670926517,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2114,
+      "step": 750
+    },
+    {
+      "epoch": 1.1996805111821087,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2156,
+      "step": 751
+    },
+    {
+      "epoch": 1.2012779552715656,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0005,
+      "loss": 1.205,
+      "step": 752
+    },
+    {
+      "epoch": 1.2028753993610224,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2052,
+      "step": 753
+    },
+    {
+      "epoch": 1.2044728434504792,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2039,
+      "step": 754
+    },
+    {
+      "epoch": 1.206070287539936,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.1964,
+      "step": 755
+    },
+    {
+      "epoch": 1.207667731629393,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1937,
+      "step": 756
+    },
+    {
+      "epoch": 1.20926517571885,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.193,
+      "step": 757
+    },
+    {
+      "epoch": 1.2108626198083068,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2096,
+      "step": 758
+    },
+    {
+      "epoch": 1.2124600638977636,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2032,
+      "step": 759
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1986,
+      "step": 760
+    },
+    {
+      "epoch": 1.2156549520766773,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1981,
+      "step": 761
+    },
+    {
+      "epoch": 1.2172523961661341,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2094,
+      "step": 762
+    },
+    {
+      "epoch": 1.218849840255591,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2121,
+      "step": 763
+    },
+    {
+      "epoch": 1.220447284345048,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.21,
+      "step": 764
+    },
+    {
+      "epoch": 1.2220447284345048,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1916,
+      "step": 765
+    },
+    {
+      "epoch": 1.2236421725239617,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2016,
+      "step": 766
+    },
+    {
+      "epoch": 1.2252396166134185,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2113,
+      "step": 767
+    },
+    {
+      "epoch": 1.2268370607028753,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2104,
+      "step": 768
+    },
+    {
+      "epoch": 1.2284345047923322,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.209,
+      "step": 769
+    },
+    {
+      "epoch": 1.230031948881789,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2074,
+      "step": 770
+    },
+    {
+      "epoch": 1.231629392971246,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1816,
+      "step": 771
+    },
+    {
+      "epoch": 1.233226837060703,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1883,
+      "step": 772
+    },
+    {
+      "epoch": 1.2348242811501597,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1987,
+      "step": 773
+    },
+    {
+      "epoch": 1.2364217252396166,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2005,
+      "step": 774
+    },
+    {
+      "epoch": 1.2380191693290734,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.205,
+      "step": 775
+    },
+    {
+      "epoch": 1.2396166134185305,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2037,
+      "step": 776
+    },
+    {
+      "epoch": 1.2412140575079873,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.197,
+      "step": 777
+    },
+    {
+      "epoch": 1.2428115015974441,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1965,
+      "step": 778
+    },
+    {
+      "epoch": 1.244408945686901,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2184,
+      "step": 779
+    },
+    {
+      "epoch": 1.2460063897763578,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2137,
+      "step": 780
+    },
+    {
+      "epoch": 1.2476038338658146,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1955,
+      "step": 781
+    },
+    {
+      "epoch": 1.2492012779552715,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2116,
+      "step": 782
+    },
+    {
+      "epoch": 1.2507987220447285,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1972,
+      "step": 783
+    },
+    {
+      "epoch": 1.2523961661341854,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1898,
+      "step": 784
+    },
+    {
+      "epoch": 1.2539936102236422,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1908,
+      "step": 785
+    },
+    {
+      "epoch": 1.255591054313099,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1942,
+      "step": 786
+    },
+    {
+      "epoch": 1.2571884984025559,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2085,
+      "step": 787
+    },
+    {
+      "epoch": 1.2587859424920127,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1977,
+      "step": 788
+    },
+    {
+      "epoch": 1.2603833865814695,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1893,
+      "step": 789
+    },
+    {
+      "epoch": 1.2619808306709266,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1902,
+      "step": 790
+    },
+    {
+      "epoch": 1.2635782747603834,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1947,
+      "step": 791
+    },
+    {
+      "epoch": 1.2651757188498403,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2011,
+      "step": 792
+    },
+    {
+      "epoch": 1.266773162939297,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2064,
+      "step": 793
+    },
+    {
+      "epoch": 1.268370607028754,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1955,
+      "step": 794
+    },
+    {
+      "epoch": 1.269968051118211,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2073,
+      "step": 795
+    },
+    {
+      "epoch": 1.2715654952076676,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1947,
+      "step": 796
+    },
+    {
+      "epoch": 1.2731629392971247,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2112,
+      "step": 797
+    },
+    {
+      "epoch": 1.2747603833865815,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1889,
+      "step": 798
+    },
+    {
+      "epoch": 1.2763578274760383,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1926,
+      "step": 799
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1972,
+      "step": 800
+    },
+    {
+      "epoch": 1.279552715654952,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2028,
+      "step": 801
+    },
+    {
+      "epoch": 1.281150159744409,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2089,
+      "step": 802
+    },
+    {
+      "epoch": 1.2827476038338659,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1887,
+      "step": 803
+    },
+    {
+      "epoch": 1.2843450479233227,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1981,
+      "step": 804
+    },
+    {
+      "epoch": 1.2859424920127795,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1939,
+      "step": 805
+    },
+    {
+      "epoch": 1.2875399361022364,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1997,
+      "step": 806
+    },
+    {
+      "epoch": 1.2891373801916932,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1979,
+      "step": 807
+    },
+    {
+      "epoch": 1.29073482428115,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1855,
+      "step": 808
+    },
+    {
+      "epoch": 1.292332268370607,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2008,
+      "step": 809
+    },
+    {
+      "epoch": 1.293929712460064,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2103,
+      "step": 810
+    },
+    {
+      "epoch": 1.2955271565495208,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 811
+    },
+    {
+      "epoch": 1.2971246006389776,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.2008,
+      "step": 812
+    },
+    {
+      "epoch": 1.2987220447284344,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1932,
+      "step": 813
+    },
+    {
+      "epoch": 1.3003194888178915,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1956,
+      "step": 814
+    },
+    {
+      "epoch": 1.3019169329073481,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1848,
+      "step": 815
+    },
+    {
+      "epoch": 1.3035143769968052,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.197,
+      "step": 816
+    },
+    {
+      "epoch": 1.305111821086262,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2092,
+      "step": 817
+    },
+    {
+      "epoch": 1.3067092651757188,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1873,
+      "step": 818
+    },
+    {
+      "epoch": 1.3083067092651757,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1933,
+      "step": 819
+    },
+    {
+      "epoch": 1.3099041533546325,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2063,
+      "step": 820
+    },
+    {
+      "epoch": 1.3115015974440896,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1948,
+      "step": 821
+    },
+    {
+      "epoch": 1.3130990415335464,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2015,
+      "step": 822
+    },
+    {
+      "epoch": 1.3146964856230032,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1926,
+      "step": 823
+    },
+    {
+      "epoch": 1.31629392971246,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1849,
+      "step": 824
+    },
+    {
+      "epoch": 1.317891373801917,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2017,
+      "step": 825
+    },
+    {
+      "epoch": 1.3194888178913737,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2028,
+      "step": 826
+    },
+    {
+      "epoch": 1.3210862619808306,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1937,
+      "step": 827
+    },
+    {
+      "epoch": 1.3226837060702876,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1983,
+      "step": 828
+    },
+    {
+      "epoch": 1.3242811501597445,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1939,
+      "step": 829
+    },
+    {
+      "epoch": 1.3258785942492013,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1877,
+      "step": 830
+    },
+    {
+      "epoch": 1.3274760383386581,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1992,
+      "step": 831
+    },
+    {
+      "epoch": 1.329073482428115,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1921,
+      "step": 832
+    },
+    {
+      "epoch": 1.330670926517572,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2049,
+      "step": 833
+    },
+    {
+      "epoch": 1.3322683706070286,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1954,
+      "step": 834
+    },
+    {
+      "epoch": 1.3338658146964857,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1981,
+      "step": 835
+    },
+    {
+      "epoch": 1.3354632587859425,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2052,
+      "step": 836
+    },
+    {
+      "epoch": 1.3370607028753994,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 837
+    },
+    {
+      "epoch": 1.3386581469648562,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2004,
+      "step": 838
+    },
+    {
+      "epoch": 1.340255591054313,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.196,
+      "step": 839
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1906,
+      "step": 840
+    },
+    {
+      "epoch": 1.343450479233227,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1937,
+      "step": 841
+    },
+    {
+      "epoch": 1.3450479233226837,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.187,
+      "step": 842
+    },
+    {
+      "epoch": 1.3466453674121406,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2026,
+      "step": 843
+    },
+    {
+      "epoch": 1.3482428115015974,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1995,
+      "step": 844
+    },
+    {
+      "epoch": 1.3498402555910542,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1841,
+      "step": 845
+    },
+    {
+      "epoch": 1.351437699680511,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1862,
+      "step": 846
+    },
+    {
+      "epoch": 1.3530351437699681,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1948,
+      "step": 847
+    },
+    {
+      "epoch": 1.354632587859425,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1941,
+      "step": 848
+    },
+    {
+      "epoch": 1.3562300319488818,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1893,
+      "step": 849
+    },
+    {
+      "epoch": 1.3578274760383386,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.2119,
+      "step": 850
+    },
+    {
+      "epoch": 1.3594249201277955,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1918,
+      "step": 851
+    },
+    {
+      "epoch": 1.3610223642172525,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.193,
+      "step": 852
+    },
+    {
+      "epoch": 1.3626198083067091,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1887,
+      "step": 853
+    },
+    {
+      "epoch": 1.3642172523961662,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1916,
+      "step": 854
+    },
+    {
+      "epoch": 1.365814696485623,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2032,
+      "step": 855
+    },
+    {
+      "epoch": 1.3674121405750799,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2001,
+      "step": 856
+    },
+    {
+      "epoch": 1.3690095846645367,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1934,
+      "step": 857
+    },
+    {
+      "epoch": 1.3706070287539935,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1959,
+      "step": 858
+    },
+    {
+      "epoch": 1.3722044728434506,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1944,
+      "step": 859
+    },
+    {
+      "epoch": 1.3738019169329074,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1944,
+      "step": 860
+    },
+    {
+      "epoch": 1.3753993610223643,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2023,
+      "step": 861
+    },
+    {
+      "epoch": 1.376996805111821,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1851,
+      "step": 862
+    },
+    {
+      "epoch": 1.378594249201278,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1989,
+      "step": 863
+    },
+    {
+      "epoch": 1.3801916932907348,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2114,
+      "step": 864
+    },
+    {
+      "epoch": 1.3817891373801916,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1933,
+      "step": 865
+    },
+    {
+      "epoch": 1.3833865814696487,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1915,
+      "step": 866
+    },
+    {
+      "epoch": 1.3849840255591055,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1943,
+      "step": 867
+    },
+    {
+      "epoch": 1.3865814696485623,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 868
+    },
+    {
+      "epoch": 1.3881789137380192,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.187,
+      "step": 869
+    },
+    {
+      "epoch": 1.389776357827476,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2059,
+      "step": 870
+    },
+    {
+      "epoch": 1.3913738019169328,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1994,
+      "step": 871
+    },
+    {
+      "epoch": 1.3929712460063897,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1948,
+      "step": 872
+    },
+    {
+      "epoch": 1.3945686900958467,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2002,
+      "step": 873
+    },
+    {
+      "epoch": 1.3961661341853036,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2026,
+      "step": 874
+    },
+    {
+      "epoch": 1.3977635782747604,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1988,
+      "step": 875
+    },
+    {
+      "epoch": 1.3993610223642172,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2111,
+      "step": 876
+    },
+    {
+      "epoch": 1.400958466453674,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 877
+    },
+    {
+      "epoch": 1.4025559105431311,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1941,
+      "step": 878
+    },
+    {
+      "epoch": 1.4041533546325877,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1977,
+      "step": 879
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 880
+    },
+    {
+      "epoch": 1.4073482428115016,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1968,
+      "step": 881
+    },
+    {
+      "epoch": 1.4089456869009584,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1861,
+      "step": 882
+    },
+    {
+      "epoch": 1.4105431309904153,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.1983,
+      "step": 883
+    },
+    {
+      "epoch": 1.4121405750798721,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 884
+    },
+    {
+      "epoch": 1.4137380191693292,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2044,
+      "step": 885
+    },
+    {
+      "epoch": 1.415335463258786,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1986,
+      "step": 886
+    },
+    {
+      "epoch": 1.4169329073482428,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1891,
+      "step": 887
+    },
+    {
+      "epoch": 1.4185303514376997,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1881,
+      "step": 888
+    },
+    {
+      "epoch": 1.4201277955271565,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1975,
+      "step": 889
+    },
+    {
+      "epoch": 1.4217252396166133,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1962,
+      "step": 890
+    },
+    {
+      "epoch": 1.4233226837060702,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1927,
+      "step": 891
+    },
+    {
+      "epoch": 1.4249201277955272,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1797,
+      "step": 892
+    },
+    {
+      "epoch": 1.426517571884984,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 893
+    },
+    {
+      "epoch": 1.428115015974441,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1899,
+      "step": 894
+    },
+    {
+      "epoch": 1.4297124600638977,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1817,
+      "step": 895
+    },
+    {
+      "epoch": 1.4313099041533546,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1987,
+      "step": 896
+    },
+    {
+      "epoch": 1.4329073482428116,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1881,
+      "step": 897
+    },
+    {
+      "epoch": 1.4345047923322682,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2004,
+      "step": 898
+    },
+    {
+      "epoch": 1.4361022364217253,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.195,
+      "step": 899
+    },
+    {
+      "epoch": 1.4376996805111821,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2038,
+      "step": 900
+    },
+    {
+      "epoch": 1.439297124600639,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1758,
+      "step": 901
+    },
+    {
+      "epoch": 1.4408945686900958,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2012,
+      "step": 902
+    },
+    {
+      "epoch": 1.4424920127795526,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1952,
+      "step": 903
+    },
+    {
+      "epoch": 1.4440894568690097,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1964,
+      "step": 904
+    },
+    {
+      "epoch": 1.4456869009584665,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1926,
+      "step": 905
+    },
+    {
+      "epoch": 1.4472843450479234,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1979,
+      "step": 906
+    },
+    {
+      "epoch": 1.4488817891373802,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1855,
+      "step": 907
+    },
+    {
+      "epoch": 1.450479233226837,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2021,
+      "step": 908
+    },
+    {
+      "epoch": 1.4520766773162939,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1905,
+      "step": 909
+    },
+    {
+      "epoch": 1.4536741214057507,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.196,
+      "step": 910
+    },
+    {
+      "epoch": 1.4552715654952078,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1829,
+      "step": 911
+    },
+    {
+      "epoch": 1.4568690095846646,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1918,
+      "step": 912
+    },
+    {
+      "epoch": 1.4584664536741214,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1834,
+      "step": 913
+    },
+    {
+      "epoch": 1.4600638977635783,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.186,
+      "step": 914
+    },
+    {
+      "epoch": 1.461661341853035,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1894,
+      "step": 915
+    },
+    {
+      "epoch": 1.4632587859424921,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 916
+    },
+    {
+      "epoch": 1.4648562300319488,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1938,
+      "step": 917
+    },
+    {
+      "epoch": 1.4664536741214058,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1886,
+      "step": 918
+    },
+    {
+      "epoch": 1.4680511182108626,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.189,
+      "step": 919
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1969,
+      "step": 920
+    },
+    {
+      "epoch": 1.4712460063897763,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1848,
+      "step": 921
+    },
+    {
+      "epoch": 1.4728434504792332,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1883,
+      "step": 922
+    },
+    {
+      "epoch": 1.4744408945686902,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1888,
+      "step": 923
+    },
+    {
+      "epoch": 1.476038338658147,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0005,
+      "loss": 1.189,
+      "step": 924
+    },
+    {
+      "epoch": 1.4776357827476039,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1934,
+      "step": 925
+    },
+    {
+      "epoch": 1.4792332268370607,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1907,
+      "step": 926
+    },
+    {
+      "epoch": 1.4808306709265175,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2069,
+      "step": 927
+    },
+    {
+      "epoch": 1.4824281150159744,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1857,
+      "step": 928
+    },
+    {
+      "epoch": 1.4840255591054312,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1886,
+      "step": 929
+    },
+    {
+      "epoch": 1.4856230031948883,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1901,
+      "step": 930
+    },
+    {
+      "epoch": 1.487220447284345,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1956,
+      "step": 931
+    },
+    {
+      "epoch": 1.488817891373802,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1927,
+      "step": 932
+    },
+    {
+      "epoch": 1.4904153354632588,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2021,
+      "step": 933
+    },
+    {
+      "epoch": 1.4920127795527156,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.1878,
+      "step": 934
+    },
+    {
+      "epoch": 1.4936102236421724,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 935
+    },
+    {
+      "epoch": 1.4952076677316293,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1908,
+      "step": 936
+    },
+    {
+      "epoch": 1.4968051118210863,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.195,
+      "step": 937
+    },
+    {
+      "epoch": 1.4984025559105432,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1925,
+      "step": 938
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1983,
+      "step": 939
+    },
+    {
+      "epoch": 1.5015974440894568,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1902,
+      "step": 940
+    },
+    {
+      "epoch": 1.5031948881789137,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1921,
+      "step": 941
+    },
+    {
+      "epoch": 1.5047923322683707,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1884,
+      "step": 942
+    },
+    {
+      "epoch": 1.5063897763578273,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.2068,
+      "step": 943
+    },
+    {
+      "epoch": 1.5079872204472844,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1932,
+      "step": 944
+    },
+    {
+      "epoch": 1.5095846645367412,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1901,
+      "step": 945
+    },
+    {
+      "epoch": 1.511182108626198,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1998,
+      "step": 946
+    },
+    {
+      "epoch": 1.5127795527156551,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1814,
+      "step": 947
+    },
+    {
+      "epoch": 1.5143769968051117,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1977,
+      "step": 948
+    },
+    {
+      "epoch": 1.5159744408945688,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 949
+    },
+    {
+      "epoch": 1.5175718849840254,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 950
+    },
+    {
+      "epoch": 1.5191693290734825,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2066,
+      "step": 951
+    },
+    {
+      "epoch": 1.5207667731629393,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1869,
+      "step": 952
+    },
+    {
+      "epoch": 1.5223642172523961,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1941,
+      "step": 953
+    },
+    {
+      "epoch": 1.5239616613418532,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.193,
+      "step": 954
+    },
+    {
+      "epoch": 1.5255591054313098,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1856,
+      "step": 955
+    },
+    {
+      "epoch": 1.5271565495207668,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2031,
+      "step": 956
+    },
+    {
+      "epoch": 1.5287539936102237,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 957
+    },
+    {
+      "epoch": 1.5303514376996805,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1939,
+      "step": 958
+    },
+    {
+      "epoch": 1.5319488817891374,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.182,
+      "step": 959
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1963,
+      "step": 960
+    },
+    {
+      "epoch": 1.5351437699680512,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1971,
+      "step": 961
+    },
+    {
+      "epoch": 1.5367412140575079,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.1915,
+      "step": 962
+    },
+    {
+      "epoch": 1.538338658146965,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2075,
+      "step": 963
+    },
+    {
+      "epoch": 1.5399361022364217,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0005,
+      "loss": 1.1878,
+      "step": 964
+    },
+    {
+      "epoch": 1.5415335463258786,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.1963,
+      "step": 965
+    },
+    {
+      "epoch": 1.5431309904153354,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 966
+    },
+    {
+      "epoch": 1.5447284345047922,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.1907,
+      "step": 967
+    },
+    {
+      "epoch": 1.5463258785942493,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1865,
+      "step": 968
+    },
+    {
+      "epoch": 1.547923322683706,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1896,
+      "step": 969
+    },
+    {
+      "epoch": 1.549520766773163,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1987,
+      "step": 970
+    },
+    {
+      "epoch": 1.5511182108626198,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.199,
+      "step": 971
+    },
+    {
+      "epoch": 1.5527156549520766,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1912,
+      "step": 972
+    },
+    {
+      "epoch": 1.5543130990415337,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1834,
+      "step": 973
+    },
+    {
+      "epoch": 1.5559105431309903,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.183,
+      "step": 974
+    },
+    {
+      "epoch": 1.5575079872204474,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1843,
+      "step": 975
+    },
+    {
+      "epoch": 1.5591054313099042,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1966,
+      "step": 976
+    },
+    {
+      "epoch": 1.560702875399361,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1867,
+      "step": 977
+    },
+    {
+      "epoch": 1.5623003194888179,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1924,
+      "step": 978
+    },
+    {
+      "epoch": 1.5638977635782747,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2025,
+      "step": 979
+    },
+    {
+      "epoch": 1.5654952076677318,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1925,
+      "step": 980
+    },
+    {
+      "epoch": 1.5670926517571884,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1785,
+      "step": 981
+    },
+    {
+      "epoch": 1.5686900958466454,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 982
+    },
+    {
+      "epoch": 1.5702875399361023,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1911,
+      "step": 983
+    },
+    {
+      "epoch": 1.571884984025559,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1861,
+      "step": 984
+    },
+    {
+      "epoch": 1.573482428115016,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1864,
+      "step": 985
+    },
+    {
+      "epoch": 1.5750798722044728,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.203,
+      "step": 986
+    },
+    {
+      "epoch": 1.5766773162939298,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1927,
+      "step": 987
+    },
+    {
+      "epoch": 1.5782747603833864,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1851,
+      "step": 988
+    },
+    {
+      "epoch": 1.5798722044728435,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1905,
+      "step": 989
+    },
+    {
+      "epoch": 1.5814696485623003,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1771,
+      "step": 990
+    },
+    {
+      "epoch": 1.5830670926517572,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1816,
+      "step": 991
+    },
+    {
+      "epoch": 1.5846645367412142,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 992
+    },
+    {
+      "epoch": 1.5862619808306708,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1817,
+      "step": 993
+    },
+    {
+      "epoch": 1.5878594249201279,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.203,
+      "step": 994
+    },
+    {
+      "epoch": 1.5894568690095847,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.19,
+      "step": 995
+    },
+    {
+      "epoch": 1.5910543130990416,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1925,
+      "step": 996
+    },
+    {
+      "epoch": 1.5926517571884984,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1961,
+      "step": 997
+    },
+    {
+      "epoch": 1.5942492012779552,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1968,
+      "step": 998
+    },
+    {
+      "epoch": 1.5958466453674123,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2044,
+      "step": 999
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1874,
+      "step": 1000
+    },
+    {
+      "epoch": 1.599041533546326,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1936,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6006389776357828,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1904,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6022364217252396,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1832,
+      "step": 1003
+    },
+    {
+      "epoch": 1.6038338658146964,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2008,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6054313099041533,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6070287539936103,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1945,
+      "step": 1006
+    },
+    {
+      "epoch": 1.608626198083067,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1784,
+      "step": 1007
+    },
+    {
+      "epoch": 1.610223642172524,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1952,
+      "step": 1008
+    },
+    {
+      "epoch": 1.6118210862619808,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2027,
+      "step": 1009
+    },
+    {
+      "epoch": 1.6134185303514377,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1973,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6150159744408947,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1955,
+      "step": 1011
+    },
+    {
+      "epoch": 1.6166134185303513,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1948,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6182108626198084,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 1013
+    },
+    {
+      "epoch": 1.619808306709265,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1848,
+      "step": 1014
+    },
+    {
+      "epoch": 1.621405750798722,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1957,
+      "step": 1015
+    },
+    {
+      "epoch": 1.623003194888179,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.179,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6246006389776357,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1898,
+      "step": 1017
+    },
+    {
+      "epoch": 1.6261980830670928,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1832,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6277955271565494,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1942,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6293929712460065,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.192,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6309904153354633,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6325878594249201,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1874,
+      "step": 1022
+    },
+    {
+      "epoch": 1.634185303514377,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1887,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6357827476038338,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1924,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6373801916932909,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1971,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6389776357827475,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1908,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6405750798722045,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1861,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6421725239616614,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1873,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6437699680511182,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1964,
+      "step": 1029
+    },
+    {
+      "epoch": 1.645367412140575,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1904,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6469648562300319,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1891,
+      "step": 1031
+    },
+    {
+      "epoch": 1.648562300319489,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1896,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6501597444089455,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1916,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6517571884984026,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1859,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6533546325878594,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6549520766773163,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1851,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6565495207667733,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1963,
+      "step": 1037
+    },
+    {
+      "epoch": 1.65814696485623,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1813,
+      "step": 1038
+    },
+    {
+      "epoch": 1.659744408945687,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1941,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6613418530351438,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1957,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6629392971246006,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1912,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6645367412140575,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1906,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6661341853035143,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1857,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6677316293929714,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1902,
+      "step": 1044
+    },
+    {
+      "epoch": 1.669329073482428,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1863,
+      "step": 1045
+    },
+    {
+      "epoch": 1.670926517571885,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1897,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6725239616613419,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1894,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6741214057507987,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1926,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6757188498402555,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1854,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6773162939297124,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6789137380191694,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1842,
+      "step": 1051
+    },
+    {
+      "epoch": 1.680511182108626,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1989,
+      "step": 1052
+    },
+    {
+      "epoch": 1.682108626198083,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1866,
+      "step": 1053
+    },
+    {
+      "epoch": 1.68370607028754,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.191,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6853035143769968,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.2012,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6869009584664538,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6884984025559104,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1878,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6900958466453675,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6916932907348243,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1887,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6932907348242812,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 1060
+    },
+    {
+      "epoch": 1.694888178913738,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1873,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6964856230031948,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1901,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6980830670926519,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.183,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6996805111821085,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1801,
+      "step": 1064
+    },
+    {
+      "epoch": 1.7012779552715656,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1876,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7028753993610224,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1911,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7044728434504792,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1924,
+      "step": 1067
+    },
+    {
+      "epoch": 1.706070287539936,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1904,
+      "step": 1068
+    },
+    {
+      "epoch": 1.707667731629393,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1798,
+      "step": 1069
+    },
+    {
+      "epoch": 1.70926517571885,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 1070
+    },
+    {
+      "epoch": 1.7108626198083066,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.186,
+      "step": 1071
+    },
+    {
+      "epoch": 1.7124600638977636,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7140575079872205,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 1073
+    },
+    {
+      "epoch": 1.7156549520766773,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1074
+    },
+    {
+      "epoch": 1.7172523961661343,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1962,
+      "step": 1075
+    },
+    {
+      "epoch": 1.718849840255591,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1894,
+      "step": 1076
+    },
+    {
+      "epoch": 1.720447284345048,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 1077
+    },
+    {
+      "epoch": 1.7220447284345048,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1963,
+      "step": 1078
+    },
+    {
+      "epoch": 1.7236421725239617,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1804,
+      "step": 1079
+    },
+    {
+      "epoch": 1.7252396166134185,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2015,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7268370607028753,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1849,
+      "step": 1081
+    },
+    {
+      "epoch": 1.7284345047923324,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1082
+    },
+    {
+      "epoch": 1.730031948881789,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2014,
+      "step": 1083
+    },
+    {
+      "epoch": 1.731629392971246,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1843,
+      "step": 1084
+    },
+    {
+      "epoch": 1.733226837060703,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1801,
+      "step": 1085
+    },
+    {
+      "epoch": 1.7348242811501597,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1814,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7364217252396166,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1753,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7380191693290734,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1907,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7396166134185305,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1969,
+      "step": 1089
+    },
+    {
+      "epoch": 1.741214057507987,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1924,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7428115015974441,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1941,
+      "step": 1091
+    },
+    {
+      "epoch": 1.744408945686901,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1843,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7460063897763578,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1892,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7476038338658149,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2055,
+      "step": 1094
+    },
+    {
+      "epoch": 1.7492012779552715,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1923,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7507987220447285,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2033,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7523961661341851,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2107,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7539936102236422,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0005,
+      "loss": 1.186,
+      "step": 1098
+    },
+    {
+      "epoch": 1.755591054313099,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1777,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7571884984025559,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.1921,
+      "step": 1100
+    },
+    {
+      "epoch": 1.758785942492013,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2016,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7603833865814695,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.19,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7619808306709266,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.2012,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7635782747603834,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1842,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7651757188498403,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.182,
+      "step": 1105
+    },
+    {
+      "epoch": 1.766773162939297,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.1863,
+      "step": 1106
+    },
+    {
+      "epoch": 1.768370607028754,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1963,
+      "step": 1107
+    },
+    {
+      "epoch": 1.769968051118211,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1837,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7715654952076676,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1776,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7731629392971247,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0005,
+      "loss": 1.184,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7747603833865815,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1763,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7763578274760383,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.2062,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7779552715654952,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.175,
+      "step": 1113
+    },
+    {
+      "epoch": 1.779552715654952,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1897,
+      "step": 1114
+    },
+    {
+      "epoch": 1.781150159744409,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.2061,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7827476038338657,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1833,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7843450479233227,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1911,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7859424920127795,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1855,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7875399361022364,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1858,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7891373801916934,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1951,
+      "step": 1120
+    },
+    {
+      "epoch": 1.79073482428115,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1977,
+      "step": 1121
+    },
+    {
+      "epoch": 1.792332268370607,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1861,
+      "step": 1122
+    },
+    {
+      "epoch": 1.793929712460064,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1924,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7955271565495208,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1846,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7971246006389776,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.188,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7987220447284344,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8003194888178915,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1961,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8019169329073481,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1832,
+      "step": 1128
+    },
+    {
+      "epoch": 1.8035143769968052,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1997,
+      "step": 1129
+    },
+    {
+      "epoch": 1.805111821086262,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1886,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8067092651757188,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1871,
+      "step": 1131
+    },
+    {
+      "epoch": 1.8083067092651757,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1853,
+      "step": 1132
+    },
+    {
+      "epoch": 1.8099041533546325,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.178,
+      "step": 1133
+    },
+    {
+      "epoch": 1.8115015974440896,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1822,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8130990415335462,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1829,
+      "step": 1135
+    },
+    {
+      "epoch": 1.8146964856230032,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1802,
+      "step": 1136
+    },
+    {
+      "epoch": 1.81629392971246,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1804,
+      "step": 1137
+    },
+    {
+      "epoch": 1.817891373801917,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1857,
+      "step": 1138
+    },
+    {
+      "epoch": 1.819488817891374,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 1139
+    },
+    {
+      "epoch": 1.8210862619808306,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1886,
+      "step": 1140
+    },
+    {
+      "epoch": 1.8226837060702876,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.182,
+      "step": 1141
+    },
+    {
+      "epoch": 1.8242811501597445,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1884,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8258785942492013,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.187,
+      "step": 1143
+    },
+    {
+      "epoch": 1.8274760383386581,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1876,
+      "step": 1144
+    },
+    {
+      "epoch": 1.829073482428115,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1888,
+      "step": 1145
+    },
+    {
+      "epoch": 1.830670926517572,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1863,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8322683706070286,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1899,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8338658146964857,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1944,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8354632587859425,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1903,
+      "step": 1149
+    },
+    {
+      "epoch": 1.8370607028753994,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1808,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8386581469648562,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1797,
+      "step": 1151
+    },
+    {
+      "epoch": 1.840255591054313,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1763,
+      "step": 1152
+    },
+    {
+      "epoch": 1.84185303514377,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8434504792332267,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1911,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8450479233226837,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1932,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8466453674121406,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.187,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8482428115015974,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1826,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8498402555910545,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1818,
+      "step": 1158
+    },
+    {
+      "epoch": 1.851437699680511,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1989,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8530351437699681,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1160
+    },
+    {
+      "epoch": 1.854632587859425,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8562300319488818,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1802,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8578274760383386,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8594249201277955,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1854,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8610223642172525,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1828,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8626198083067091,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8642172523961662,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1979,
+      "step": 1167
+    },
+    {
+      "epoch": 1.865814696485623,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.185,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8674121405750799,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8690095846645367,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.183,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8706070287539935,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8722044728434506,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1894,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8738019169329072,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8753993610223643,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1862,
+      "step": 1174
+    },
+    {
+      "epoch": 1.876996805111821,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1772,
+      "step": 1175
+    },
+    {
+      "epoch": 1.878594249201278,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1764,
+      "step": 1176
+    },
+    {
+      "epoch": 1.880191693290735,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1785,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8817891373801916,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1802,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8833865814696487,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1918,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8849840255591053,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8865814696485623,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1908,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8881789137380192,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1863,
+      "step": 1182
+    },
+    {
+      "epoch": 1.889776357827476,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1862,
+      "step": 1183
+    },
+    {
+      "epoch": 1.891373801916933,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1829,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8929712460063897,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1942,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8945686900958467,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1777,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8961661341853036,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.187,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8977635782747604,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1821,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8993610223642172,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1823,
+      "step": 1189
+    },
+    {
+      "epoch": 1.900958466453674,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1818,
+      "step": 1190
+    },
+    {
+      "epoch": 1.9025559105431311,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1949,
+      "step": 1191
+    },
+    {
+      "epoch": 1.9041533546325877,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.174,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9057507987220448,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1867,
+      "step": 1193
+    },
+    {
+      "epoch": 1.9073482428115016,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1752,
+      "step": 1194
+    },
+    {
+      "epoch": 1.9089456869009584,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1845,
+      "step": 1195
+    },
+    {
+      "epoch": 1.9105431309904153,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2065,
+      "step": 1196
+    },
+    {
+      "epoch": 1.9121405750798721,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.2005,
+      "step": 1197
+    },
+    {
+      "epoch": 1.9137380191693292,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9153354632587858,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1868,
+      "step": 1199
+    },
+    {
+      "epoch": 1.9169329073482428,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1846,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9185303514376997,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1913,
+      "step": 1201
+    },
+    {
+      "epoch": 1.9201277955271565,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1842,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9217252396166136,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1877,
+      "step": 1203
+    },
+    {
+      "epoch": 1.9233226837060702,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1999,
+      "step": 1204
+    },
+    {
+      "epoch": 1.9249201277955272,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 1205
+    },
+    {
+      "epoch": 1.926517571884984,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1884,
+      "step": 1206
+    },
+    {
+      "epoch": 1.928115015974441,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1207
+    },
+    {
+      "epoch": 1.9297124600638977,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0005,
+      "loss": 1.194,
+      "step": 1208
+    },
+    {
+      "epoch": 1.9313099041533546,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1749,
+      "step": 1209
+    },
+    {
+      "epoch": 1.9329073482428116,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2043,
+      "step": 1210
+    },
+    {
+      "epoch": 1.9345047923322682,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1939,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9361022364217253,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1927,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9376996805111821,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1853,
+      "step": 1213
+    },
+    {
+      "epoch": 1.939297124600639,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1843,
+      "step": 1214
+    },
+    {
+      "epoch": 1.9408945686900958,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 1215
+    },
+    {
+      "epoch": 1.9424920127795526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1956,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9440894568690097,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1791,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9456869009584663,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1985,
+      "step": 1218
+    },
+    {
+      "epoch": 1.9472843450479234,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.1877,
+      "step": 1219
+    },
+    {
+      "epoch": 1.9488817891373802,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1913,
+      "step": 1220
+    },
+    {
+      "epoch": 1.950479233226837,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1813,
+      "step": 1221
+    },
+    {
+      "epoch": 1.952076677316294,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1822,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9536741214057507,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9552715654952078,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1819,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9568690095846646,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1828,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9584664536741214,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1785,
+      "step": 1226
+    },
+    {
+      "epoch": 1.9600638977635783,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.189,
+      "step": 1227
+    },
+    {
+      "epoch": 1.961661341853035,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1799,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9632587859424921,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1818,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9648562300319488,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1884,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9664536741214058,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1812,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9680511182108626,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9696485623003195,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9712460063897763,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9728434504792332,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1901,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9744408945686902,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9760383386581468,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1945,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9776357827476039,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1855,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9792332268370607,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1845,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9808306709265175,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.184,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9824281150159746,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1823,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9840255591054312,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1879,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9856230031948883,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1875,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9872204472843449,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 1244
+    },
+    {
+      "epoch": 1.988817891373802,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1922,
+      "step": 1245
+    },
+    {
+      "epoch": 1.9904153354632588,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9920127795527156,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1801,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9936102236421727,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1903,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9952076677316293,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9968051118210863,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1747,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9984025559105432,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2005,
+      "step": 1251
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1897,
+      "step": 1252
+    },
+    {
+      "epoch": 2.001597444089457,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1858,
+      "step": 1253
+    },
+    {
+      "epoch": 2.0031948881789137,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1254
+    },
+    {
+      "epoch": 2.0047923322683707,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1255
+    },
+    {
+      "epoch": 2.0063897763578273,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1798,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0079872204472844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.176,
+      "step": 1257
+    },
+    {
+      "epoch": 2.009584664536741,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1258
+    },
+    {
+      "epoch": 2.011182108626198,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1259
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1260
+    },
+    {
+      "epoch": 2.0143769968051117,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1757,
+      "step": 1261
+    },
+    {
+      "epoch": 2.015974440894569,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1821,
+      "step": 1262
+    },
+    {
+      "epoch": 2.0175718849840254,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1897,
+      "step": 1263
+    },
+    {
+      "epoch": 2.0191693290734825,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1264
+    },
+    {
+      "epoch": 2.0207667731629395,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 1265
+    },
+    {
+      "epoch": 2.022364217252396,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.18,
+      "step": 1266
+    },
+    {
+      "epoch": 2.023961661341853,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1267
+    },
+    {
+      "epoch": 2.02555910543131,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 1268
+    },
+    {
+      "epoch": 2.027156549520767,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1805,
+      "step": 1269
+    },
+    {
+      "epoch": 2.0287539936102235,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 1270
+    },
+    {
+      "epoch": 2.0303514376996805,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1271
+    },
+    {
+      "epoch": 2.0319488817891376,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1855,
+      "step": 1272
+    },
+    {
+      "epoch": 2.033546325878594,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1823,
+      "step": 1273
+    },
+    {
+      "epoch": 2.0351437699680512,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1856,
+      "step": 1274
+    },
+    {
+      "epoch": 2.036741214057508,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1275
+    },
+    {
+      "epoch": 2.038338658146965,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 1276
+    },
+    {
+      "epoch": 2.0399361022364215,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1277
+    },
+    {
+      "epoch": 2.0415335463258786,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1796,
+      "step": 1278
+    },
+    {
+      "epoch": 2.0431309904153356,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1848,
+      "step": 1279
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1883,
+      "step": 1280
+    },
+    {
+      "epoch": 2.0463258785942493,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1281
+    },
+    {
+      "epoch": 2.047923322683706,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1807,
+      "step": 1282
+    },
+    {
+      "epoch": 2.049520766773163,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1735,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0511182108626196,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1882,
+      "step": 1284
+    },
+    {
+      "epoch": 2.0527156549520766,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1897,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0543130990415337,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 1286
+    },
+    {
+      "epoch": 2.0559105431309903,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1791,
+      "step": 1287
+    },
+    {
+      "epoch": 2.0575079872204474,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 1288
+    },
+    {
+      "epoch": 2.059105431309904,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1705,
+      "step": 1289
+    },
+    {
+      "epoch": 2.060702875399361,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1953,
+      "step": 1290
+    },
+    {
+      "epoch": 2.062300319488818,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 1291
+    },
+    {
+      "epoch": 2.0638977635782747,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1845,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0654952076677318,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1709,
+      "step": 1293
+    },
+    {
+      "epoch": 2.0670926517571884,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1758,
+      "step": 1294
+    },
+    {
+      "epoch": 2.0686900958466454,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 1295
+    },
+    {
+      "epoch": 2.070287539936102,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1296
+    },
+    {
+      "epoch": 2.071884984025559,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1297
+    },
+    {
+      "epoch": 2.073482428115016,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1825,
+      "step": 1298
+    },
+    {
+      "epoch": 2.0750798722044728,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1299
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0782747603833864,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1847,
+      "step": 1301
+    },
+    {
+      "epoch": 2.0798722044728435,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1927,
+      "step": 1302
+    },
+    {
+      "epoch": 2.0814696485623,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1792,
+      "step": 1303
+    },
+    {
+      "epoch": 2.083067092651757,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1735,
+      "step": 1304
+    },
+    {
+      "epoch": 2.084664536741214,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1305
+    },
+    {
+      "epoch": 2.086261980830671,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1772,
+      "step": 1306
+    },
+    {
+      "epoch": 2.087859424920128,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1783,
+      "step": 1307
+    },
+    {
+      "epoch": 2.0894568690095845,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1739,
+      "step": 1308
+    },
+    {
+      "epoch": 2.0910543130990416,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 1309
+    },
+    {
+      "epoch": 2.0926517571884986,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1707,
+      "step": 1310
+    },
+    {
+      "epoch": 2.094249201277955,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1748,
+      "step": 1311
+    },
+    {
+      "epoch": 2.0958466453674123,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1825,
+      "step": 1312
+    },
+    {
+      "epoch": 2.097444089456869,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1762,
+      "step": 1313
+    },
+    {
+      "epoch": 2.099041533546326,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1828,
+      "step": 1314
+    },
+    {
+      "epoch": 2.1006389776357826,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1771,
+      "step": 1315
+    },
+    {
+      "epoch": 2.1022364217252396,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1796,
+      "step": 1316
+    },
+    {
+      "epoch": 2.1038338658146967,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 1317
+    },
+    {
+      "epoch": 2.1054313099041533,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1318
+    },
+    {
+      "epoch": 2.1070287539936103,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1858,
+      "step": 1319
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1805,
+      "step": 1320
+    },
+    {
+      "epoch": 2.110223642172524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1856,
+      "step": 1321
+    },
+    {
+      "epoch": 2.1118210862619806,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1926,
+      "step": 1322
+    },
+    {
+      "epoch": 2.1134185303514377,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1803,
+      "step": 1323
+    },
+    {
+      "epoch": 2.1150159744408947,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1885,
+      "step": 1324
+    },
+    {
+      "epoch": 2.1166134185303513,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1325
+    },
+    {
+      "epoch": 2.1182108626198084,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1864,
+      "step": 1326
+    },
+    {
+      "epoch": 2.119808306709265,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.179,
+      "step": 1327
+    },
+    {
+      "epoch": 2.121405750798722,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1862,
+      "step": 1328
+    },
+    {
+      "epoch": 2.123003194888179,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1881,
+      "step": 1329
+    },
+    {
+      "epoch": 2.1246006389776357,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1828,
+      "step": 1330
+    },
+    {
+      "epoch": 2.126198083067093,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1725,
+      "step": 1331
+    },
+    {
+      "epoch": 2.1277955271565494,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1825,
+      "step": 1332
+    },
+    {
+      "epoch": 2.1293929712460065,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1803,
+      "step": 1333
+    },
+    {
+      "epoch": 2.130990415335463,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1826,
+      "step": 1334
+    },
+    {
+      "epoch": 2.13258785942492,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1839,
+      "step": 1335
+    },
+    {
+      "epoch": 2.134185303514377,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1859,
+      "step": 1336
+    },
+    {
+      "epoch": 2.135782747603834,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1787,
+      "step": 1337
+    },
+    {
+      "epoch": 2.137380191693291,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 1338
+    },
+    {
+      "epoch": 2.1389776357827475,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1826,
+      "step": 1339
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 1340
+    },
+    {
+      "epoch": 2.142172523961661,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 1341
+    },
+    {
+      "epoch": 2.143769968051118,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 1342
+    },
+    {
+      "epoch": 2.1453674121405752,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1659,
+      "step": 1343
+    },
+    {
+      "epoch": 2.146964856230032,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1344
+    },
+    {
+      "epoch": 2.148562300319489,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1792,
+      "step": 1345
+    },
+    {
+      "epoch": 2.1501597444089455,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 1346
+    },
+    {
+      "epoch": 2.1517571884984026,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1347
+    },
+    {
+      "epoch": 2.1533546325878596,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.182,
+      "step": 1348
+    },
+    {
+      "epoch": 2.1549520766773163,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1349
+    },
+    {
+      "epoch": 2.1565495207667733,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1818,
+      "step": 1350
+    },
+    {
+      "epoch": 2.15814696485623,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 1351
+    },
+    {
+      "epoch": 2.159744408945687,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 1352
+    },
+    {
+      "epoch": 2.1613418530351436,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 1353
+    },
+    {
+      "epoch": 2.1629392971246006,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1843,
+      "step": 1354
+    },
+    {
+      "epoch": 2.1645367412140577,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.182,
+      "step": 1355
+    },
+    {
+      "epoch": 2.1661341853035143,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1748,
+      "step": 1356
+    },
+    {
+      "epoch": 2.1677316293929714,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1357
+    },
+    {
+      "epoch": 2.169329073482428,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1773,
+      "step": 1358
+    },
+    {
+      "epoch": 2.170926517571885,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1848,
+      "step": 1359
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1807,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1741214057507987,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1361
+    },
+    {
+      "epoch": 2.1757188498402558,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.178,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1773162939297124,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.186,
+      "step": 1363
+    },
+    {
+      "epoch": 2.1789137380191694,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1765,
+      "step": 1364
+    },
+    {
+      "epoch": 2.180511182108626,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1773,
+      "step": 1365
+    },
+    {
+      "epoch": 2.182108626198083,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1366
+    },
+    {
+      "epoch": 2.18370607028754,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 1367
+    },
+    {
+      "epoch": 2.1853035143769968,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 1368
+    },
+    {
+      "epoch": 2.186900958466454,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1842,
+      "step": 1369
+    },
+    {
+      "epoch": 2.1884984025559104,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1779,
+      "step": 1370
+    },
+    {
+      "epoch": 2.1900958466453675,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.202,
+      "step": 1371
+    },
+    {
+      "epoch": 2.191693290734824,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.174,
+      "step": 1372
+    },
+    {
+      "epoch": 2.193290734824281,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.18,
+      "step": 1373
+    },
+    {
+      "epoch": 2.194888178913738,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1749,
+      "step": 1374
+    },
+    {
+      "epoch": 2.196485623003195,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 1375
+    },
+    {
+      "epoch": 2.198083067092652,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1797,
+      "step": 1376
+    },
+    {
+      "epoch": 2.1996805111821085,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1767,
+      "step": 1377
+    },
+    {
+      "epoch": 2.2012779552715656,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1808,
+      "step": 1378
+    },
+    {
+      "epoch": 2.202875399361022,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1833,
+      "step": 1379
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1623,
+      "step": 1380
+    },
+    {
+      "epoch": 2.2060702875399363,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1381
+    },
+    {
+      "epoch": 2.207667731629393,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1862,
+      "step": 1382
+    },
+    {
+      "epoch": 2.20926517571885,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1775,
+      "step": 1383
+    },
+    {
+      "epoch": 2.2108626198083066,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1384
+    },
+    {
+      "epoch": 2.2124600638977636,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1878,
+      "step": 1385
+    },
+    {
+      "epoch": 2.2140575079872207,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 1386
+    },
+    {
+      "epoch": 2.2156549520766773,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1804,
+      "step": 1387
+    },
+    {
+      "epoch": 2.2172523961661343,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 1388
+    },
+    {
+      "epoch": 2.218849840255591,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1773,
+      "step": 1389
+    },
+    {
+      "epoch": 2.220447284345048,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 1390
+    },
+    {
+      "epoch": 2.2220447284345046,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1826,
+      "step": 1391
+    },
+    {
+      "epoch": 2.2236421725239617,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1842,
+      "step": 1392
+    },
+    {
+      "epoch": 2.2252396166134187,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1772,
+      "step": 1393
+    },
+    {
+      "epoch": 2.2268370607028753,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1786,
+      "step": 1394
+    },
+    {
+      "epoch": 2.2284345047923324,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1888,
+      "step": 1395
+    },
+    {
+      "epoch": 2.230031948881789,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1714,
+      "step": 1396
+    },
+    {
+      "epoch": 2.231629392971246,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1791,
+      "step": 1397
+    },
+    {
+      "epoch": 2.2332268370607027,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1755,
+      "step": 1398
+    },
+    {
+      "epoch": 2.2348242811501597,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1399
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 1400
+    },
+    {
+      "epoch": 2.2380191693290734,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1401
+    },
+    {
+      "epoch": 2.2396166134185305,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1777,
+      "step": 1402
+    },
+    {
+      "epoch": 2.241214057507987,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1856,
+      "step": 1403
+    },
+    {
+      "epoch": 2.242811501597444,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 1404
+    },
+    {
+      "epoch": 2.244408945686901,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1742,
+      "step": 1405
+    },
+    {
+      "epoch": 2.246006389776358,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 1406
+    },
+    {
+      "epoch": 2.247603833865815,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1407
+    },
+    {
+      "epoch": 2.2492012779552715,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1408
+    },
+    {
+      "epoch": 2.2507987220447285,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1804,
+      "step": 1409
+    },
+    {
+      "epoch": 2.252396166134185,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 1410
+    },
+    {
+      "epoch": 2.253993610223642,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1856,
+      "step": 1411
+    },
+    {
+      "epoch": 2.255591054313099,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1832,
+      "step": 1412
+    },
+    {
+      "epoch": 2.257188498402556,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1883,
+      "step": 1413
+    },
+    {
+      "epoch": 2.258785942492013,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1968,
+      "step": 1414
+    },
+    {
+      "epoch": 2.2603833865814695,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 1415
+    },
+    {
+      "epoch": 2.2619808306709266,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1706,
+      "step": 1416
+    },
+    {
+      "epoch": 2.263578274760383,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1672,
+      "step": 1417
+    },
+    {
+      "epoch": 2.2651757188498403,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.187,
+      "step": 1418
+    },
+    {
+      "epoch": 2.2667731629392973,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 1419
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1795,
+      "step": 1420
+    },
+    {
+      "epoch": 2.269968051118211,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1841,
+      "step": 1421
+    },
+    {
+      "epoch": 2.2715654952076676,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1758,
+      "step": 1422
+    },
+    {
+      "epoch": 2.2731629392971247,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2747603833865817,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1785,
+      "step": 1424
+    },
+    {
+      "epoch": 2.2763578274760383,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 1425
+    },
+    {
+      "epoch": 2.2779552715654954,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1829,
+      "step": 1426
+    },
+    {
+      "epoch": 2.279552715654952,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1427
+    },
+    {
+      "epoch": 2.281150159744409,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 1428
+    },
+    {
+      "epoch": 2.2827476038338657,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1429
+    },
+    {
+      "epoch": 2.2843450479233227,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1854,
+      "step": 1430
+    },
+    {
+      "epoch": 2.2859424920127793,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1832,
+      "step": 1431
+    },
+    {
+      "epoch": 2.2875399361022364,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1789,
+      "step": 1432
+    },
+    {
+      "epoch": 2.2891373801916934,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 1433
+    },
+    {
+      "epoch": 2.29073482428115,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1899,
+      "step": 1434
+    },
+    {
+      "epoch": 2.292332268370607,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 1435
+    },
+    {
+      "epoch": 2.2939297124600637,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1841,
+      "step": 1436
+    },
+    {
+      "epoch": 2.2955271565495208,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 1437
+    },
+    {
+      "epoch": 2.297124600638978,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 1438
+    },
+    {
+      "epoch": 2.2987220447284344,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1847,
+      "step": 1439
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.183,
+      "step": 1440
+    },
+    {
+      "epoch": 2.301916932907348,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1798,
+      "step": 1441
+    },
+    {
+      "epoch": 2.303514376996805,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1442
+    },
+    {
+      "epoch": 2.3051118210862622,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 1443
+    },
+    {
+      "epoch": 2.306709265175719,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1843,
+      "step": 1444
+    },
+    {
+      "epoch": 2.308306709265176,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1445
+    },
+    {
+      "epoch": 2.3099041533546325,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1714,
+      "step": 1446
+    },
+    {
+      "epoch": 2.3115015974440896,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2003,
+      "step": 1447
+    },
+    {
+      "epoch": 2.313099041533546,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0005,
+      "loss": 1.1867,
+      "step": 1448
+    },
+    {
+      "epoch": 2.3146964856230032,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1768,
+      "step": 1449
+    },
+    {
+      "epoch": 2.31629392971246,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1450
+    },
+    {
+      "epoch": 2.317891373801917,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1925,
+      "step": 1451
+    },
+    {
+      "epoch": 2.319488817891374,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1778,
+      "step": 1452
+    },
+    {
+      "epoch": 2.3210862619808306,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 1453
+    },
+    {
+      "epoch": 2.3226837060702876,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1867,
+      "step": 1454
+    },
+    {
+      "epoch": 2.3242811501597442,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1876,
+      "step": 1455
+    },
+    {
+      "epoch": 2.3258785942492013,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1804,
+      "step": 1456
+    },
+    {
+      "epoch": 2.3274760383386583,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1865,
+      "step": 1457
+    },
+    {
+      "epoch": 2.329073482428115,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1458
+    },
+    {
+      "epoch": 2.330670926517572,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1716,
+      "step": 1459
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1814,
+      "step": 1460
+    },
+    {
+      "epoch": 2.3338658146964857,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1461
+    },
+    {
+      "epoch": 2.3354632587859427,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1462
+    },
+    {
+      "epoch": 2.3370607028753994,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 1463
+    },
+    {
+      "epoch": 2.3386581469648564,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1903,
+      "step": 1464
+    },
+    {
+      "epoch": 2.340255591054313,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 1465
+    },
+    {
+      "epoch": 2.34185303514377,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3434504792332267,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1867,
+      "step": 1467
+    },
+    {
+      "epoch": 2.3450479233226837,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1771,
+      "step": 1468
+    },
+    {
+      "epoch": 2.3466453674121404,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1901,
+      "step": 1469
+    },
+    {
+      "epoch": 2.3482428115015974,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1757,
+      "step": 1470
+    },
+    {
+      "epoch": 2.3498402555910545,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1471
+    },
+    {
+      "epoch": 2.351437699680511,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1472
+    },
+    {
+      "epoch": 2.353035143769968,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1705,
+      "step": 1473
+    },
+    {
+      "epoch": 2.3546325878594248,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 1474
+    },
+    {
+      "epoch": 2.356230031948882,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 1475
+    },
+    {
+      "epoch": 2.357827476038339,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1476
+    },
+    {
+      "epoch": 2.3594249201277955,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 1477
+    },
+    {
+      "epoch": 2.3610223642172525,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1725,
+      "step": 1478
+    },
+    {
+      "epoch": 2.362619808306709,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1479
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1738,
+      "step": 1480
+    },
+    {
+      "epoch": 2.365814696485623,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1481
+    },
+    {
+      "epoch": 2.36741214057508,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1775,
+      "step": 1482
+    },
+    {
+      "epoch": 2.369009584664537,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 1483
+    },
+    {
+      "epoch": 2.3706070287539935,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 1484
+    },
+    {
+      "epoch": 2.3722044728434506,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 1485
+    },
+    {
+      "epoch": 2.373801916932907,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1781,
+      "step": 1486
+    },
+    {
+      "epoch": 2.3753993610223643,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1487
+    },
+    {
+      "epoch": 2.376996805111821,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1488
+    },
+    {
+      "epoch": 2.378594249201278,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 1489
+    },
+    {
+      "epoch": 2.380191693290735,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1490
+    },
+    {
+      "epoch": 2.3817891373801916,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1776,
+      "step": 1491
+    },
+    {
+      "epoch": 2.3833865814696487,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1775,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3849840255591053,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1794,
+      "step": 1493
+    },
+    {
+      "epoch": 2.3865814696485623,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1741,
+      "step": 1494
+    },
+    {
+      "epoch": 2.3881789137380194,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1734,
+      "step": 1495
+    },
+    {
+      "epoch": 2.389776357827476,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1496
+    },
+    {
+      "epoch": 2.391373801916933,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.17,
+      "step": 1497
+    },
+    {
+      "epoch": 2.3929712460063897,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 1498
+    },
+    {
+      "epoch": 2.3945686900958467,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 1499
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1837,
+      "step": 1500
+    },
+    {
+      "epoch": 2.3977635782747604,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1755,
+      "step": 1501
+    },
+    {
+      "epoch": 2.3993610223642174,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1744,
+      "step": 1502
+    },
+    {
+      "epoch": 2.400958466453674,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1702,
+      "step": 1503
+    },
+    {
+      "epoch": 2.402555910543131,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 1504
+    },
+    {
+      "epoch": 2.4041533546325877,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1713,
+      "step": 1505
+    },
+    {
+      "epoch": 2.405750798722045,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1506
+    },
+    {
+      "epoch": 2.4073482428115014,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1507
+    },
+    {
+      "epoch": 2.4089456869009584,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1791,
+      "step": 1508
+    },
+    {
+      "epoch": 2.4105431309904155,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 1509
+    },
+    {
+      "epoch": 2.412140575079872,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 1510
+    },
+    {
+      "epoch": 2.413738019169329,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1511
+    },
+    {
+      "epoch": 2.415335463258786,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 1512
+    },
+    {
+      "epoch": 2.416932907348243,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1834,
+      "step": 1513
+    },
+    {
+      "epoch": 2.4185303514377,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 1514
+    },
+    {
+      "epoch": 2.4201277955271565,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1644,
+      "step": 1515
+    },
+    {
+      "epoch": 2.4217252396166136,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1516
+    },
+    {
+      "epoch": 2.42332268370607,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1848,
+      "step": 1517
+    },
+    {
+      "epoch": 2.4249201277955272,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1652,
+      "step": 1518
+    },
+    {
+      "epoch": 2.426517571884984,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1824,
+      "step": 1519
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1764,
+      "step": 1520
+    },
+    {
+      "epoch": 2.4297124600638975,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1747,
+      "step": 1521
+    },
+    {
+      "epoch": 2.4313099041533546,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1716,
+      "step": 1522
+    },
+    {
+      "epoch": 2.4329073482428116,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1792,
+      "step": 1523
+    },
+    {
+      "epoch": 2.4345047923322682,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 1524
+    },
+    {
+      "epoch": 2.4361022364217253,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1738,
+      "step": 1525
+    },
+    {
+      "epoch": 2.437699680511182,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1526
+    },
+    {
+      "epoch": 2.439297124600639,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 1527
+    },
+    {
+      "epoch": 2.440894568690096,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1721,
+      "step": 1528
+    },
+    {
+      "epoch": 2.4424920127795526,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 1529
+    },
+    {
+      "epoch": 2.4440894568690097,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1530
+    },
+    {
+      "epoch": 2.4456869009584663,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1736,
+      "step": 1531
+    },
+    {
+      "epoch": 2.4472843450479234,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1532
+    },
+    {
+      "epoch": 2.4488817891373804,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1788,
+      "step": 1533
+    },
+    {
+      "epoch": 2.450479233226837,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 1534
+    },
+    {
+      "epoch": 2.452076677316294,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 1535
+    },
+    {
+      "epoch": 2.4536741214057507,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1736,
+      "step": 1536
+    },
+    {
+      "epoch": 2.4552715654952078,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1767,
+      "step": 1537
+    },
+    {
+      "epoch": 2.4568690095846644,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1792,
+      "step": 1538
+    },
+    {
+      "epoch": 2.4584664536741214,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 1539
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1540
+    },
+    {
+      "epoch": 2.461661341853035,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1772,
+      "step": 1541
+    },
+    {
+      "epoch": 2.463258785942492,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1732,
+      "step": 1542
+    },
+    {
+      "epoch": 2.4648562300319488,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1767,
+      "step": 1543
+    },
+    {
+      "epoch": 2.466453674121406,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1544
+    },
+    {
+      "epoch": 2.4680511182108624,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1545
+    },
+    {
+      "epoch": 2.4696485623003195,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 1546
+    },
+    {
+      "epoch": 2.4712460063897765,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1709,
+      "step": 1547
+    },
+    {
+      "epoch": 2.472843450479233,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1738,
+      "step": 1548
+    },
+    {
+      "epoch": 2.47444089456869,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1755,
+      "step": 1549
+    },
+    {
+      "epoch": 2.476038338658147,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1753,
+      "step": 1550
+    },
+    {
+      "epoch": 2.477635782747604,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 1551
+    },
+    {
+      "epoch": 2.479233226837061,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1897,
+      "step": 1552
+    },
+    {
+      "epoch": 2.4808306709265175,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1817,
+      "step": 1553
+    },
+    {
+      "epoch": 2.4824281150159746,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1765,
+      "step": 1554
+    },
+    {
+      "epoch": 2.484025559105431,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1555
+    },
+    {
+      "epoch": 2.4856230031948883,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.168,
+      "step": 1556
+    },
+    {
+      "epoch": 2.487220447284345,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 1557
+    },
+    {
+      "epoch": 2.488817891373802,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1558
+    },
+    {
+      "epoch": 2.4904153354632586,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1559
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1742,
+      "step": 1560
+    },
+    {
+      "epoch": 2.4936102236421727,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1561
+    },
+    {
+      "epoch": 2.4952076677316293,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 1562
+    },
+    {
+      "epoch": 2.4968051118210863,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1563
+    },
+    {
+      "epoch": 2.498402555910543,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1707,
+      "step": 1564
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1805,
+      "step": 1565
+    },
+    {
+      "epoch": 2.501597444089457,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 1566
+    },
+    {
+      "epoch": 2.5031948881789137,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1735,
+      "step": 1567
+    },
+    {
+      "epoch": 2.5047923322683707,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1878,
+      "step": 1568
+    },
+    {
+      "epoch": 2.5063897763578273,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1569
+    },
+    {
+      "epoch": 2.5079872204472844,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1771,
+      "step": 1570
+    },
+    {
+      "epoch": 2.5095846645367414,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1856,
+      "step": 1571
+    },
+    {
+      "epoch": 2.511182108626198,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1767,
+      "step": 1572
+    },
+    {
+      "epoch": 2.512779552715655,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1676,
+      "step": 1573
+    },
+    {
+      "epoch": 2.5143769968051117,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 1574
+    },
+    {
+      "epoch": 2.515974440894569,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 1575
+    },
+    {
+      "epoch": 2.5175718849840254,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1698,
+      "step": 1576
+    },
+    {
+      "epoch": 2.5191693290734825,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1688,
+      "step": 1577
+    },
+    {
+      "epoch": 2.520766773162939,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1788,
+      "step": 1578
+    },
+    {
+      "epoch": 2.522364217252396,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 1579
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1816,
+      "step": 1580
+    },
+    {
+      "epoch": 2.52555910543131,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 1581
+    },
+    {
+      "epoch": 2.527156549520767,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1755,
+      "step": 1582
+    },
+    {
+      "epoch": 2.5287539936102235,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1812,
+      "step": 1583
+    },
+    {
+      "epoch": 2.5303514376996805,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1736,
+      "step": 1584
+    },
+    {
+      "epoch": 2.5319488817891376,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1801,
+      "step": 1585
+    },
+    {
+      "epoch": 2.533546325878594,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 1586
+    },
+    {
+      "epoch": 2.5351437699680512,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1587
+    },
+    {
+      "epoch": 2.536741214057508,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 1588
+    },
+    {
+      "epoch": 2.538338658146965,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1721,
+      "step": 1589
+    },
+    {
+      "epoch": 2.539936102236422,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 1590
+    },
+    {
+      "epoch": 2.5415335463258786,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1591
+    },
+    {
+      "epoch": 2.543130990415335,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1592
+    },
+    {
+      "epoch": 2.5447284345047922,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1742,
+      "step": 1593
+    },
+    {
+      "epoch": 2.5463258785942493,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1594
+    },
+    {
+      "epoch": 2.547923322683706,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1714,
+      "step": 1595
+    },
+    {
+      "epoch": 2.549520766773163,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1771,
+      "step": 1596
+    },
+    {
+      "epoch": 2.5511182108626196,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1597
+    },
+    {
+      "epoch": 2.5527156549520766,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 1598
+    },
+    {
+      "epoch": 2.5543130990415337,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 1599
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1600
+    },
+    {
+      "epoch": 2.5575079872204474,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1732,
+      "step": 1601
+    },
+    {
+      "epoch": 2.559105431309904,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1602
+    },
+    {
+      "epoch": 2.560702875399361,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 1603
+    },
+    {
+      "epoch": 2.562300319488818,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1741,
+      "step": 1604
+    },
+    {
+      "epoch": 2.5638977635782747,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5654952076677318,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1671,
+      "step": 1606
+    },
+    {
+      "epoch": 2.5670926517571884,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1839,
+      "step": 1607
+    },
+    {
+      "epoch": 2.5686900958466454,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5702875399361025,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1609
+    },
+    {
+      "epoch": 2.571884984025559,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.179,
+      "step": 1610
+    },
+    {
+      "epoch": 2.5734824281150157,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5750798722044728,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1731,
+      "step": 1612
+    },
+    {
+      "epoch": 2.57667731629393,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 1613
+    },
+    {
+      "epoch": 2.5782747603833864,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1685,
+      "step": 1614
+    },
+    {
+      "epoch": 2.5798722044728435,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 1615
+    },
+    {
+      "epoch": 2.5814696485623,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1616
+    },
+    {
+      "epoch": 2.583067092651757,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 1617
+    },
+    {
+      "epoch": 2.584664536741214,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1826,
+      "step": 1618
+    },
+    {
+      "epoch": 2.586261980830671,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1619
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1807,
+      "step": 1620
+    },
+    {
+      "epoch": 2.5894568690095845,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1621
+    },
+    {
+      "epoch": 2.5910543130990416,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 1622
+    },
+    {
+      "epoch": 2.5926517571884986,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 1623
+    },
+    {
+      "epoch": 2.594249201277955,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 1624
+    },
+    {
+      "epoch": 2.5958466453674123,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 1625
+    },
+    {
+      "epoch": 2.597444089456869,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1773,
+      "step": 1626
+    },
+    {
+      "epoch": 2.599041533546326,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 1627
+    },
+    {
+      "epoch": 2.600638977635783,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1765,
+      "step": 1628
+    },
+    {
+      "epoch": 2.6022364217252396,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 1629
+    },
+    {
+      "epoch": 2.6038338658146962,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1763,
+      "step": 1630
+    },
+    {
+      "epoch": 2.6054313099041533,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1631
+    },
+    {
+      "epoch": 2.6070287539936103,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1773,
+      "step": 1632
+    },
+    {
+      "epoch": 2.608626198083067,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1633
+    },
+    {
+      "epoch": 2.610223642172524,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 1634
+    },
+    {
+      "epoch": 2.6118210862619806,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1721,
+      "step": 1635
+    },
+    {
+      "epoch": 2.6134185303514377,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 1636
+    },
+    {
+      "epoch": 2.6150159744408947,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1776,
+      "step": 1637
+    },
+    {
+      "epoch": 2.6166134185303513,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1783,
+      "step": 1638
+    },
+    {
+      "epoch": 2.6182108626198084,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.168,
+      "step": 1639
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1640
+    },
+    {
+      "epoch": 2.621405750798722,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1641
+    },
+    {
+      "epoch": 2.623003194888179,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1642
+    },
+    {
+      "epoch": 2.6246006389776357,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 1643
+    },
+    {
+      "epoch": 2.626198083067093,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1707,
+      "step": 1644
+    },
+    {
+      "epoch": 2.6277955271565494,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.175,
+      "step": 1645
+    },
+    {
+      "epoch": 2.6293929712460065,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.168,
+      "step": 1646
+    },
+    {
+      "epoch": 2.6309904153354635,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1647
+    },
+    {
+      "epoch": 2.63258785942492,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 1648
+    },
+    {
+      "epoch": 2.6341853035143767,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 1649
+    },
+    {
+      "epoch": 2.635782747603834,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 1650
+    },
+    {
+      "epoch": 2.637380191693291,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1708,
+      "step": 1651
+    },
+    {
+      "epoch": 2.6389776357827475,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1705,
+      "step": 1652
+    },
+    {
+      "epoch": 2.6405750798722045,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 1653
+    },
+    {
+      "epoch": 2.642172523961661,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1777,
+      "step": 1654
+    },
+    {
+      "epoch": 2.643769968051118,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1838,
+      "step": 1655
+    },
+    {
+      "epoch": 2.6453674121405752,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 1656
+    },
+    {
+      "epoch": 2.646964856230032,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1767,
+      "step": 1657
+    },
+    {
+      "epoch": 2.648562300319489,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1658
+    },
+    {
+      "epoch": 2.6501597444089455,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 1659
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1660
+    },
+    {
+      "epoch": 2.6533546325878596,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1661
+    },
+    {
+      "epoch": 2.6549520766773163,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.169,
+      "step": 1662
+    },
+    {
+      "epoch": 2.6565495207667733,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1663
+    },
+    {
+      "epoch": 2.65814696485623,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1797,
+      "step": 1664
+    },
+    {
+      "epoch": 2.659744408945687,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1788,
+      "step": 1665
+    },
+    {
+      "epoch": 2.661341853035144,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.174,
+      "step": 1666
+    },
+    {
+      "epoch": 2.6629392971246006,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 1667
+    },
+    {
+      "epoch": 2.6645367412140573,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6661341853035143,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 1669
+    },
+    {
+      "epoch": 2.6677316293929714,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1826,
+      "step": 1670
+    },
+    {
+      "epoch": 2.669329073482428,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1783,
+      "step": 1671
+    },
+    {
+      "epoch": 2.670926517571885,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6725239616613417,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1755,
+      "step": 1673
+    },
+    {
+      "epoch": 2.6741214057507987,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 1674
+    },
+    {
+      "epoch": 2.6757188498402558,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1889,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6773162939297124,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 1676
+    },
+    {
+      "epoch": 2.6789137380191694,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1779,
+      "step": 1677
+    },
+    {
+      "epoch": 2.680511182108626,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 1678
+    },
+    {
+      "epoch": 2.682108626198083,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1679
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1841,
+      "step": 1680
+    },
+    {
+      "epoch": 2.6853035143769968,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1681
+    },
+    {
+      "epoch": 2.686900958466454,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1682
+    },
+    {
+      "epoch": 2.6884984025559104,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1794,
+      "step": 1683
+    },
+    {
+      "epoch": 2.6900958466453675,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.181,
+      "step": 1684
+    },
+    {
+      "epoch": 2.6916932907348246,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1804,
+      "step": 1685
+    },
+    {
+      "epoch": 2.693290734824281,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1686
+    },
+    {
+      "epoch": 2.6948881789137378,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 1687
+    },
+    {
+      "epoch": 2.696485623003195,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 1688
+    },
+    {
+      "epoch": 2.698083067092652,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1663,
+      "step": 1689
+    },
+    {
+      "epoch": 2.6996805111821085,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 1690
+    },
+    {
+      "epoch": 2.7012779552715656,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 1691
+    },
+    {
+      "epoch": 2.702875399361022,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.165,
+      "step": 1692
+    },
+    {
+      "epoch": 2.7044728434504792,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 1693
+    },
+    {
+      "epoch": 2.7060702875399363,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1779,
+      "step": 1694
+    },
+    {
+      "epoch": 2.707667731629393,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1835,
+      "step": 1695
+    },
+    {
+      "epoch": 2.70926517571885,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 1696
+    },
+    {
+      "epoch": 2.7108626198083066,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1658,
+      "step": 1697
+    },
+    {
+      "epoch": 2.7124600638977636,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 1698
+    },
+    {
+      "epoch": 2.7140575079872207,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1713,
+      "step": 1699
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1700
+    },
+    {
+      "epoch": 2.7172523961661343,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1788,
+      "step": 1701
+    },
+    {
+      "epoch": 2.718849840255591,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1823,
+      "step": 1702
+    },
+    {
+      "epoch": 2.720447284345048,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.17,
+      "step": 1703
+    },
+    {
+      "epoch": 2.722044728434505,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 1704
+    },
+    {
+      "epoch": 2.7236421725239617,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1781,
+      "step": 1705
+    },
+    {
+      "epoch": 2.7252396166134183,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1706
+    },
+    {
+      "epoch": 2.7268370607028753,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1828,
+      "step": 1707
+    },
+    {
+      "epoch": 2.7284345047923324,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 1708
+    },
+    {
+      "epoch": 2.730031948881789,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1737,
+      "step": 1709
+    },
+    {
+      "epoch": 2.731629392971246,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 1710
+    },
+    {
+      "epoch": 2.7332268370607027,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 1711
+    },
+    {
+      "epoch": 2.7348242811501597,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1738,
+      "step": 1712
+    },
+    {
+      "epoch": 2.736421725239617,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 1713
+    },
+    {
+      "epoch": 2.7380191693290734,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1714
+    },
+    {
+      "epoch": 2.7396166134185305,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 1715
+    },
+    {
+      "epoch": 2.741214057507987,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1729,
+      "step": 1716
+    },
+    {
+      "epoch": 2.742811501597444,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 1717
+    },
+    {
+      "epoch": 2.744408945686901,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 1718
+    },
+    {
+      "epoch": 2.746006389776358,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 1719
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1752,
+      "step": 1720
+    },
+    {
+      "epoch": 2.7492012779552715,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1829,
+      "step": 1721
+    },
+    {
+      "epoch": 2.7507987220447285,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 1722
+    },
+    {
+      "epoch": 2.752396166134185,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1833,
+      "step": 1723
+    },
+    {
+      "epoch": 2.753993610223642,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1747,
+      "step": 1724
+    },
+    {
+      "epoch": 2.755591054313099,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 1725
+    },
+    {
+      "epoch": 2.757188498402556,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 1726
+    },
+    {
+      "epoch": 2.758785942492013,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 1727
+    },
+    {
+      "epoch": 2.7603833865814695,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 1728
+    },
+    {
+      "epoch": 2.7619808306709266,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 1729
+    },
+    {
+      "epoch": 2.763578274760383,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.175,
+      "step": 1730
+    },
+    {
+      "epoch": 2.7651757188498403,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1741,
+      "step": 1731
+    },
+    {
+      "epoch": 2.7667731629392973,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1702,
+      "step": 1732
+    },
+    {
+      "epoch": 2.768370607028754,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 1733
+    },
+    {
+      "epoch": 2.769968051118211,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 1734
+    },
+    {
+      "epoch": 2.7715654952076676,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 1735
+    },
+    {
+      "epoch": 2.7731629392971247,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 1736
+    },
+    {
+      "epoch": 2.7747603833865817,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1764,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7763578274760383,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1603,
+      "step": 1738
+    },
+    {
+      "epoch": 2.777955271565495,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 1739
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.165,
+      "step": 1740
+    },
+    {
+      "epoch": 2.781150159744409,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 1741
+    },
+    {
+      "epoch": 2.7827476038338657,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1742
+    },
+    {
+      "epoch": 2.7843450479233227,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.175,
+      "step": 1743
+    },
+    {
+      "epoch": 2.7859424920127793,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1731,
+      "step": 1744
+    },
+    {
+      "epoch": 2.7875399361022364,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1652,
+      "step": 1745
+    },
+    {
+      "epoch": 2.7891373801916934,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1671,
+      "step": 1746
+    },
+    {
+      "epoch": 2.79073482428115,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 1747
+    },
+    {
+      "epoch": 2.792332268370607,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1719,
+      "step": 1748
+    },
+    {
+      "epoch": 2.7939297124600637,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1707,
+      "step": 1749
+    },
+    {
+      "epoch": 2.7955271565495208,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1705,
+      "step": 1750
+    },
+    {
+      "epoch": 2.797124600638978,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 1751
+    },
+    {
+      "epoch": 2.7987220447284344,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 1752
+    },
+    {
+      "epoch": 2.8003194888178915,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1753
+    },
+    {
+      "epoch": 2.801916932907348,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 1754
+    },
+    {
+      "epoch": 2.803514376996805,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 1755
+    },
+    {
+      "epoch": 2.8051118210862622,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1801,
+      "step": 1756
+    },
+    {
+      "epoch": 2.806709265175719,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 1757
+    },
+    {
+      "epoch": 2.8083067092651754,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 1758
+    },
+    {
+      "epoch": 2.8099041533546325,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1688,
+      "step": 1759
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1763,
+      "step": 1760
+    },
+    {
+      "epoch": 2.813099041533546,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1725,
+      "step": 1761
+    },
+    {
+      "epoch": 2.8146964856230032,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 1762
+    },
+    {
+      "epoch": 2.81629392971246,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 1763
+    },
+    {
+      "epoch": 2.817891373801917,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1771,
+      "step": 1764
+    },
+    {
+      "epoch": 2.819488817891374,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 1765
+    },
+    {
+      "epoch": 2.8210862619808306,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 1766
+    },
+    {
+      "epoch": 2.8226837060702876,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1751,
+      "step": 1767
+    },
+    {
+      "epoch": 2.8242811501597442,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.174,
+      "step": 1768
+    },
+    {
+      "epoch": 2.8258785942492013,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1739,
+      "step": 1769
+    },
+    {
+      "epoch": 2.8274760383386583,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 1770
+    },
+    {
+      "epoch": 2.829073482428115,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1771
+    },
+    {
+      "epoch": 2.830670926517572,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1646,
+      "step": 1772
+    },
+    {
+      "epoch": 2.8322683706070286,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.176,
+      "step": 1773
+    },
+    {
+      "epoch": 2.8338658146964857,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1629,
+      "step": 1774
+    },
+    {
+      "epoch": 2.8354632587859427,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1775
+    },
+    {
+      "epoch": 2.8370607028753994,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 1776
+    },
+    {
+      "epoch": 2.838658146964856,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 1777
+    },
+    {
+      "epoch": 2.840255591054313,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1749,
+      "step": 1778
+    },
+    {
+      "epoch": 2.84185303514377,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1706,
+      "step": 1779
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1780
+    },
+    {
+      "epoch": 2.8450479233226837,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 1781
+    },
+    {
+      "epoch": 2.8466453674121404,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 1782
+    },
+    {
+      "epoch": 2.8482428115015974,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1658,
+      "step": 1783
+    },
+    {
+      "epoch": 2.8498402555910545,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1762,
+      "step": 1784
+    },
+    {
+      "epoch": 2.851437699680511,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 1785
+    },
+    {
+      "epoch": 2.853035143769968,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1762,
+      "step": 1786
+    },
+    {
+      "epoch": 2.8546325878594248,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 1787
+    },
+    {
+      "epoch": 2.856230031948882,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1752,
+      "step": 1788
+    },
+    {
+      "epoch": 2.857827476038339,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1789
+    },
+    {
+      "epoch": 2.8594249201277955,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1790
+    },
+    {
+      "epoch": 2.8610223642172525,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1655,
+      "step": 1791
+    },
+    {
+      "epoch": 2.862619808306709,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 1792
+    },
+    {
+      "epoch": 2.864217252396166,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1793
+    },
+    {
+      "epoch": 2.8658146964856233,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.166,
+      "step": 1794
+    },
+    {
+      "epoch": 2.86741214057508,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1773,
+      "step": 1795
+    },
+    {
+      "epoch": 2.8690095846645365,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1821,
+      "step": 1796
+    },
+    {
+      "epoch": 2.8706070287539935,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1781,
+      "step": 1797
+    },
+    {
+      "epoch": 2.8722044728434506,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1706,
+      "step": 1798
+    },
+    {
+      "epoch": 2.873801916932907,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1764,
+      "step": 1799
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1800
+    },
+    {
+      "epoch": 2.876996805111821,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 1801
+    },
+    {
+      "epoch": 2.878594249201278,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1857,
+      "step": 1802
+    },
+    {
+      "epoch": 2.880191693290735,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 1803
+    },
+    {
+      "epoch": 2.8817891373801916,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 1804
+    },
+    {
+      "epoch": 2.8833865814696487,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 1805
+    },
+    {
+      "epoch": 2.8849840255591053,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1751,
+      "step": 1806
+    },
+    {
+      "epoch": 2.8865814696485623,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.166,
+      "step": 1807
+    },
+    {
+      "epoch": 2.8881789137380194,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 1808
+    },
+    {
+      "epoch": 2.889776357827476,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1809
+    },
+    {
+      "epoch": 2.891373801916933,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1748,
+      "step": 1810
+    },
+    {
+      "epoch": 2.8929712460063897,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1819,
+      "step": 1811
+    },
+    {
+      "epoch": 2.8945686900958467,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 1812
+    },
+    {
+      "epoch": 2.8961661341853038,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 1813
+    },
+    {
+      "epoch": 2.8977635782747604,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 1814
+    },
+    {
+      "epoch": 2.899361022364217,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 1815
+    },
+    {
+      "epoch": 2.900958466453674,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1684,
+      "step": 1816
+    },
+    {
+      "epoch": 2.902555910543131,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1685,
+      "step": 1817
+    },
+    {
+      "epoch": 2.9041533546325877,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1734,
+      "step": 1818
+    },
+    {
+      "epoch": 2.905750798722045,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 1819
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1708,
+      "step": 1820
+    },
+    {
+      "epoch": 2.9089456869009584,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1783,
+      "step": 1821
+    },
+    {
+      "epoch": 2.9105431309904155,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 1822
+    },
+    {
+      "epoch": 2.912140575079872,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1758,
+      "step": 1823
+    },
+    {
+      "epoch": 2.913738019169329,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 1824
+    },
+    {
+      "epoch": 2.915335463258786,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 1825
+    },
+    {
+      "epoch": 2.916932907348243,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1732,
+      "step": 1826
+    },
+    {
+      "epoch": 2.9185303514377,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 1827
+    },
+    {
+      "epoch": 2.9201277955271565,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1783,
+      "step": 1828
+    },
+    {
+      "epoch": 2.9217252396166136,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 1829
+    },
+    {
+      "epoch": 2.92332268370607,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1646,
+      "step": 1830
+    },
+    {
+      "epoch": 2.9249201277955272,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.168,
+      "step": 1831
+    },
+    {
+      "epoch": 2.9265175718849843,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1721,
+      "step": 1832
+    },
+    {
+      "epoch": 2.928115015974441,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1827,
+      "step": 1833
+    },
+    {
+      "epoch": 2.9297124600638975,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1719,
+      "step": 1834
+    },
+    {
+      "epoch": 2.9313099041533546,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1835
+    },
+    {
+      "epoch": 2.9329073482428116,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 1836
+    },
+    {
+      "epoch": 2.9345047923322682,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1729,
+      "step": 1837
+    },
+    {
+      "epoch": 2.9361022364217253,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 1838
+    },
+    {
+      "epoch": 2.937699680511182,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 1839
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1714,
+      "step": 1840
+    },
+    {
+      "epoch": 2.940894568690096,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1691,
+      "step": 1841
+    },
+    {
+      "epoch": 2.9424920127795526,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 1842
+    },
+    {
+      "epoch": 2.9440894568690097,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 1843
+    },
+    {
+      "epoch": 2.9456869009584663,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1737,
+      "step": 1844
+    },
+    {
+      "epoch": 2.9472843450479234,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1684,
+      "step": 1845
+    },
+    {
+      "epoch": 2.9488817891373804,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1846
+    },
+    {
+      "epoch": 2.950479233226837,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 1847
+    },
+    {
+      "epoch": 2.952076677316294,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.175,
+      "step": 1848
+    },
+    {
+      "epoch": 2.9536741214057507,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 1849
+    },
+    {
+      "epoch": 2.9552715654952078,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 1850
+    },
+    {
+      "epoch": 2.956869009584665,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 1851
+    },
+    {
+      "epoch": 2.9584664536741214,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 1852
+    },
+    {
+      "epoch": 2.960063897763578,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1853
+    },
+    {
+      "epoch": 2.961661341853035,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1761,
+      "step": 1854
+    },
+    {
+      "epoch": 2.963258785942492,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.166,
+      "step": 1855
+    },
+    {
+      "epoch": 2.9648562300319488,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 1856
+    },
+    {
+      "epoch": 2.966453674121406,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 1857
+    },
+    {
+      "epoch": 2.9680511182108624,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1858
+    },
+    {
+      "epoch": 2.9696485623003195,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1789,
+      "step": 1859
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 1860
+    },
+    {
+      "epoch": 2.972843450479233,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1817,
+      "step": 1861
+    },
+    {
+      "epoch": 2.97444089456869,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 1862
+    },
+    {
+      "epoch": 2.976038338658147,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 1863
+    },
+    {
+      "epoch": 2.977635782747604,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 1864
+    },
+    {
+      "epoch": 2.979233226837061,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1765,
+      "step": 1865
+    },
+    {
+      "epoch": 2.9808306709265175,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 1866
+    },
+    {
+      "epoch": 2.9824281150159746,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1867
+    },
+    {
+      "epoch": 2.984025559105431,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1708,
+      "step": 1868
+    },
+    {
+      "epoch": 2.9856230031948883,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1869
+    },
+    {
+      "epoch": 2.987220447284345,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 1870
+    },
+    {
+      "epoch": 2.988817891373802,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 1871
+    },
+    {
+      "epoch": 2.9904153354632586,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1801,
+      "step": 1872
+    },
+    {
+      "epoch": 2.9920127795527156,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1873
+    },
+    {
+      "epoch": 2.9936102236421727,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.178,
+      "step": 1874
+    },
+    {
+      "epoch": 2.9952076677316293,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 1875
+    },
+    {
+      "epoch": 2.9968051118210863,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1688,
+      "step": 1876
+    },
+    {
+      "epoch": 2.998402555910543,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1734,
+      "step": 1877
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1853,
+      "step": 1878
+    },
+    {
+      "epoch": 3.001597444089457,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 1879
+    },
+    {
+      "epoch": 3.0031948881789137,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1701,
+      "step": 1880
+    },
+    {
+      "epoch": 3.0047923322683707,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1677,
+      "step": 1881
+    },
+    {
+      "epoch": 3.0063897763578273,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1667,
+      "step": 1882
+    },
+    {
+      "epoch": 3.0079872204472844,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1675,
+      "step": 1883
+    },
+    {
+      "epoch": 3.009584664536741,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1741,
+      "step": 1884
+    },
+    {
+      "epoch": 3.011182108626198,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 1885
+    },
+    {
+      "epoch": 3.012779552715655,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1758,
+      "step": 1886
+    },
+    {
+      "epoch": 3.0143769968051117,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 1887
+    },
+    {
+      "epoch": 3.015974440894569,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 1888
+    },
+    {
+      "epoch": 3.0175718849840254,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 1889
+    },
+    {
+      "epoch": 3.0191693290734825,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 1890
+    },
+    {
+      "epoch": 3.0207667731629395,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 1891
+    },
+    {
+      "epoch": 3.022364217252396,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1753,
+      "step": 1892
+    },
+    {
+      "epoch": 3.023961661341853,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1742,
+      "step": 1893
+    },
+    {
+      "epoch": 3.02555910543131,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 1894
+    },
+    {
+      "epoch": 3.027156549520767,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 1895
+    },
+    {
+      "epoch": 3.0287539936102235,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 1896
+    },
+    {
+      "epoch": 3.0303514376996805,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1616,
+      "step": 1897
+    },
+    {
+      "epoch": 3.0319488817891376,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1691,
+      "step": 1898
+    },
+    {
+      "epoch": 3.033546325878594,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 1899
+    },
+    {
+      "epoch": 3.0351437699680512,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1900
+    },
+    {
+      "epoch": 3.036741214057508,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 1901
+    },
+    {
+      "epoch": 3.038338658146965,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1708,
+      "step": 1902
+    },
+    {
+      "epoch": 3.0399361022364215,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 1903
+    },
+    {
+      "epoch": 3.0415335463258786,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 1904
+    },
+    {
+      "epoch": 3.0431309904153356,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 1905
+    },
+    {
+      "epoch": 3.0447284345047922,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 1906
+    },
+    {
+      "epoch": 3.0463258785942493,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1806,
+      "step": 1907
+    },
+    {
+      "epoch": 3.047923322683706,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 1908
+    },
+    {
+      "epoch": 3.049520766773163,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1735,
+      "step": 1909
+    },
+    {
+      "epoch": 3.0511182108626196,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 1910
+    },
+    {
+      "epoch": 3.0527156549520766,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1684,
+      "step": 1911
+    },
+    {
+      "epoch": 3.0543130990415337,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1673,
+      "step": 1912
+    },
+    {
+      "epoch": 3.0559105431309903,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.168,
+      "step": 1913
+    },
+    {
+      "epoch": 3.0575079872204474,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1781,
+      "step": 1914
+    },
+    {
+      "epoch": 3.059105431309904,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1675,
+      "step": 1915
+    },
+    {
+      "epoch": 3.060702875399361,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1671,
+      "step": 1916
+    },
+    {
+      "epoch": 3.062300319488818,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1671,
+      "step": 1917
+    },
+    {
+      "epoch": 3.0638977635782747,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 1918
+    },
+    {
+      "epoch": 3.0654952076677318,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1698,
+      "step": 1919
+    },
+    {
+      "epoch": 3.0670926517571884,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 1920
+    },
+    {
+      "epoch": 3.0686900958466454,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1762,
+      "step": 1921
+    },
+    {
+      "epoch": 3.070287539936102,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 1922
+    },
+    {
+      "epoch": 3.071884984025559,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 1923
+    },
+    {
+      "epoch": 3.073482428115016,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 1924
+    },
+    {
+      "epoch": 3.0750798722044728,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1702,
+      "step": 1925
+    },
+    {
+      "epoch": 3.07667731629393,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 1926
+    },
+    {
+      "epoch": 3.0782747603833864,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 1927
+    },
+    {
+      "epoch": 3.0798722044728435,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1672,
+      "step": 1928
+    },
+    {
+      "epoch": 3.0814696485623,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 1929
+    },
+    {
+      "epoch": 3.083067092651757,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.181,
+      "step": 1930
+    },
+    {
+      "epoch": 3.084664536741214,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 1931
+    },
+    {
+      "epoch": 3.086261980830671,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1816,
+      "step": 1932
+    },
+    {
+      "epoch": 3.087859424920128,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 1933
+    },
+    {
+      "epoch": 3.0894568690095845,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1742,
+      "step": 1934
+    },
+    {
+      "epoch": 3.0910543130990416,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 1935
+    },
+    {
+      "epoch": 3.0926517571884986,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1701,
+      "step": 1936
+    },
+    {
+      "epoch": 3.094249201277955,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 1937
+    },
+    {
+      "epoch": 3.0958466453674123,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1754,
+      "step": 1938
+    },
+    {
+      "epoch": 3.097444089456869,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1733,
+      "step": 1939
+    },
+    {
+      "epoch": 3.099041533546326,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 1940
+    },
+    {
+      "epoch": 3.1006389776357826,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1729,
+      "step": 1941
+    },
+    {
+      "epoch": 3.1022364217252396,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.163,
+      "step": 1942
+    },
+    {
+      "epoch": 3.1038338658146967,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 1943
+    },
+    {
+      "epoch": 3.1054313099041533,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 1944
+    },
+    {
+      "epoch": 3.1070287539936103,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.163,
+      "step": 1945
+    },
+    {
+      "epoch": 3.108626198083067,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 1946
+    },
+    {
+      "epoch": 3.110223642172524,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1883,
+      "step": 1947
+    },
+    {
+      "epoch": 3.1118210862619806,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1737,
+      "step": 1948
+    },
+    {
+      "epoch": 3.1134185303514377,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 1949
+    },
+    {
+      "epoch": 3.1150159744408947,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1665,
+      "step": 1950
+    },
+    {
+      "epoch": 3.1166134185303513,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1652,
+      "step": 1951
+    },
+    {
+      "epoch": 3.1182108626198084,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 1952
+    },
+    {
+      "epoch": 3.119808306709265,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 1953
+    },
+    {
+      "epoch": 3.121405750798722,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1954
+    },
+    {
+      "epoch": 3.123003194888179,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 1955
+    },
+    {
+      "epoch": 3.1246006389776357,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 1956
+    },
+    {
+      "epoch": 3.126198083067093,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1774,
+      "step": 1957
+    },
+    {
+      "epoch": 3.1277955271565494,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 1958
+    },
+    {
+      "epoch": 3.1293929712460065,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 1959
+    },
+    {
+      "epoch": 3.130990415335463,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 1960
+    },
+    {
+      "epoch": 3.13258785942492,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 1961
+    },
+    {
+      "epoch": 3.134185303514377,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0005,
+      "loss": 1.188,
+      "step": 1962
+    },
+    {
+      "epoch": 3.135782747603834,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1709,
+      "step": 1963
+    },
+    {
+      "epoch": 3.137380191693291,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 1964
+    },
+    {
+      "epoch": 3.1389776357827475,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1811,
+      "step": 1965
+    },
+    {
+      "epoch": 3.1405750798722045,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1658,
+      "step": 1966
+    },
+    {
+      "epoch": 3.142172523961661,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0005,
+      "loss": 1.176,
+      "step": 1967
+    },
+    {
+      "epoch": 3.143769968051118,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1925,
+      "step": 1968
+    },
+    {
+      "epoch": 3.1453674121405752,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 1969
+    },
+    {
+      "epoch": 3.146964856230032,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 1970
+    },
+    {
+      "epoch": 3.148562300319489,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0005,
+      "loss": 1.1757,
+      "step": 1971
+    },
+    {
+      "epoch": 3.1501597444089455,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 1972
+    },
+    {
+      "epoch": 3.1517571884984026,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1852,
+      "step": 1973
+    },
+    {
+      "epoch": 3.1533546325878596,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 1974
+    },
+    {
+      "epoch": 3.1549520766773163,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 1975
+    },
+    {
+      "epoch": 3.1565495207667733,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 1976
+    },
+    {
+      "epoch": 3.15814696485623,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1737,
+      "step": 1977
+    },
+    {
+      "epoch": 3.159744408945687,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1725,
+      "step": 1978
+    },
+    {
+      "epoch": 3.1613418530351436,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 1979
+    },
+    {
+      "epoch": 3.1629392971246006,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1742,
+      "step": 1980
+    },
+    {
+      "epoch": 3.1645367412140577,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.168,
+      "step": 1981
+    },
+    {
+      "epoch": 3.1661341853035143,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 1982
+    },
+    {
+      "epoch": 3.1677316293929714,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1664,
+      "step": 1983
+    },
+    {
+      "epoch": 3.169329073482428,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 1984
+    },
+    {
+      "epoch": 3.170926517571885,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 1985
+    },
+    {
+      "epoch": 3.1725239616613417,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 1986
+    },
+    {
+      "epoch": 3.1741214057507987,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1718,
+      "step": 1987
+    },
+    {
+      "epoch": 3.1757188498402558,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1836,
+      "step": 1988
+    },
+    {
+      "epoch": 3.1773162939297124,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1659,
+      "step": 1989
+    },
+    {
+      "epoch": 3.1789137380191694,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1684,
+      "step": 1990
+    },
+    {
+      "epoch": 3.180511182108626,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1659,
+      "step": 1991
+    },
+    {
+      "epoch": 3.182108626198083,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 1992
+    },
+    {
+      "epoch": 3.18370607028754,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1993
+    },
+    {
+      "epoch": 3.1853035143769968,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1721,
+      "step": 1994
+    },
+    {
+      "epoch": 3.186900958466454,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1707,
+      "step": 1995
+    },
+    {
+      "epoch": 3.1884984025559104,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1713,
+      "step": 1996
+    },
+    {
+      "epoch": 3.1900958466453675,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 1997
+    },
+    {
+      "epoch": 3.191693290734824,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 1998
+    },
+    {
+      "epoch": 3.193290734824281,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 1999
+    },
+    {
+      "epoch": 3.194888178913738,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 2000
+    },
+    {
+      "epoch": 3.196485623003195,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 2001
+    },
+    {
+      "epoch": 3.198083067092652,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1713,
+      "step": 2002
+    },
+    {
+      "epoch": 3.1996805111821085,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 2003
+    },
+    {
+      "epoch": 3.2012779552715656,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2004
+    },
+    {
+      "epoch": 3.202875399361022,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1638,
+      "step": 2005
+    },
+    {
+      "epoch": 3.2044728434504792,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 2006
+    },
+    {
+      "epoch": 3.2060702875399363,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2007
+    },
+    {
+      "epoch": 3.207667731629393,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2008
+    },
+    {
+      "epoch": 3.20926517571885,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1762,
+      "step": 2009
+    },
+    {
+      "epoch": 3.2108626198083066,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1818,
+      "step": 2010
+    },
+    {
+      "epoch": 3.2124600638977636,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 2011
+    },
+    {
+      "epoch": 3.2140575079872207,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1736,
+      "step": 2012
+    },
+    {
+      "epoch": 3.2156549520766773,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1677,
+      "step": 2013
+    },
+    {
+      "epoch": 3.2172523961661343,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 2014
+    },
+    {
+      "epoch": 3.218849840255591,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1691,
+      "step": 2015
+    },
+    {
+      "epoch": 3.220447284345048,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1769,
+      "step": 2016
+    },
+    {
+      "epoch": 3.2220447284345046,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1725,
+      "step": 2017
+    },
+    {
+      "epoch": 3.2236421725239617,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2018
+    },
+    {
+      "epoch": 3.2252396166134187,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 2019
+    },
+    {
+      "epoch": 3.2268370607028753,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1634,
+      "step": 2020
+    },
+    {
+      "epoch": 3.2284345047923324,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 2021
+    },
+    {
+      "epoch": 3.230031948881789,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 2022
+    },
+    {
+      "epoch": 3.231629392971246,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2023
+    },
+    {
+      "epoch": 3.2332268370607027,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 2024
+    },
+    {
+      "epoch": 3.2348242811501597,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2025
+    },
+    {
+      "epoch": 3.236421725239617,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 2026
+    },
+    {
+      "epoch": 3.2380191693290734,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.165,
+      "step": 2027
+    },
+    {
+      "epoch": 3.2396166134185305,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1684,
+      "step": 2028
+    },
+    {
+      "epoch": 3.241214057507987,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1659,
+      "step": 2029
+    },
+    {
+      "epoch": 3.242811501597444,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 2030
+    },
+    {
+      "epoch": 3.244408945686901,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2031
+    },
+    {
+      "epoch": 3.246006389776358,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1616,
+      "step": 2032
+    },
+    {
+      "epoch": 3.247603833865815,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1731,
+      "step": 2033
+    },
+    {
+      "epoch": 3.2492012779552715,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2034
+    },
+    {
+      "epoch": 3.2507987220447285,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.182,
+      "step": 2035
+    },
+    {
+      "epoch": 3.252396166134185,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2036
+    },
+    {
+      "epoch": 3.253993610223642,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 2037
+    },
+    {
+      "epoch": 3.255591054313099,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 2038
+    },
+    {
+      "epoch": 3.257188498402556,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1672,
+      "step": 2039
+    },
+    {
+      "epoch": 3.258785942492013,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1782,
+      "step": 2040
+    },
+    {
+      "epoch": 3.2603833865814695,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 2041
+    },
+    {
+      "epoch": 3.2619808306709266,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 2042
+    },
+    {
+      "epoch": 3.263578274760383,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 2043
+    },
+    {
+      "epoch": 3.2651757188498403,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 2044
+    },
+    {
+      "epoch": 3.2667731629392973,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 2045
+    },
+    {
+      "epoch": 3.268370607028754,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 2046
+    },
+    {
+      "epoch": 3.269968051118211,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 2047
+    },
+    {
+      "epoch": 3.2715654952076676,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 2048
+    },
+    {
+      "epoch": 3.2731629392971247,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2049
+    },
+    {
+      "epoch": 3.2747603833865817,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 2050
+    },
+    {
+      "epoch": 3.2763578274760383,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1751,
+      "step": 2051
+    },
+    {
+      "epoch": 3.2779552715654954,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 2052
+    },
+    {
+      "epoch": 3.279552715654952,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 2053
+    },
+    {
+      "epoch": 3.281150159744409,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 2054
+    },
+    {
+      "epoch": 3.2827476038338657,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 2055
+    },
+    {
+      "epoch": 3.2843450479233227,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 2056
+    },
+    {
+      "epoch": 3.2859424920127793,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2057
+    },
+    {
+      "epoch": 3.2875399361022364,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 2058
+    },
+    {
+      "epoch": 3.2891373801916934,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 2059
+    },
+    {
+      "epoch": 3.29073482428115,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 2060
+    },
+    {
+      "epoch": 3.292332268370607,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2061
+    },
+    {
+      "epoch": 3.2939297124600637,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1714,
+      "step": 2062
+    },
+    {
+      "epoch": 3.2955271565495208,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2063
+    },
+    {
+      "epoch": 3.297124600638978,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2064
+    },
+    {
+      "epoch": 3.2987220447284344,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1749,
+      "step": 2065
+    },
+    {
+      "epoch": 3.3003194888178915,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.17,
+      "step": 2066
+    },
+    {
+      "epoch": 3.301916932907348,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1691,
+      "step": 2067
+    },
+    {
+      "epoch": 3.303514376996805,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1729,
+      "step": 2068
+    },
+    {
+      "epoch": 3.3051118210862622,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1792,
+      "step": 2069
+    },
+    {
+      "epoch": 3.306709265175719,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 2070
+    },
+    {
+      "epoch": 3.308306709265176,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1821,
+      "step": 2071
+    },
+    {
+      "epoch": 3.3099041533546325,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1759,
+      "step": 2072
+    },
+    {
+      "epoch": 3.3115015974440896,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 2073
+    },
+    {
+      "epoch": 3.313099041533546,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2074
+    },
+    {
+      "epoch": 3.3146964856230032,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 2075
+    },
+    {
+      "epoch": 3.31629392971246,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 2076
+    },
+    {
+      "epoch": 3.317891373801917,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 2077
+    },
+    {
+      "epoch": 3.319488817891374,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 2078
+    },
+    {
+      "epoch": 3.3210862619808306,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 2079
+    },
+    {
+      "epoch": 3.3226837060702876,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1677,
+      "step": 2080
+    },
+    {
+      "epoch": 3.3242811501597442,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 2081
+    },
+    {
+      "epoch": 3.3258785942492013,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1758,
+      "step": 2082
+    },
+    {
+      "epoch": 3.3274760383386583,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 2083
+    },
+    {
+      "epoch": 3.329073482428115,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1691,
+      "step": 2084
+    },
+    {
+      "epoch": 3.330670926517572,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 2085
+    },
+    {
+      "epoch": 3.3322683706070286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2086
+    },
+    {
+      "epoch": 3.3338658146964857,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 2087
+    },
+    {
+      "epoch": 3.3354632587859427,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1685,
+      "step": 2088
+    },
+    {
+      "epoch": 3.3370607028753994,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 2089
+    },
+    {
+      "epoch": 3.3386581469648564,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2090
+    },
+    {
+      "epoch": 3.340255591054313,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1623,
+      "step": 2091
+    },
+    {
+      "epoch": 3.34185303514377,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 2092
+    },
+    {
+      "epoch": 3.3434504792332267,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1729,
+      "step": 2093
+    },
+    {
+      "epoch": 3.3450479233226837,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2094
+    },
+    {
+      "epoch": 3.3466453674121404,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1739,
+      "step": 2095
+    },
+    {
+      "epoch": 3.3482428115015974,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 2096
+    },
+    {
+      "epoch": 3.3498402555910545,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2097
+    },
+    {
+      "epoch": 3.351437699680511,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 2098
+    },
+    {
+      "epoch": 3.353035143769968,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 2099
+    },
+    {
+      "epoch": 3.3546325878594248,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2100
+    },
+    {
+      "epoch": 3.356230031948882,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1652,
+      "step": 2101
+    },
+    {
+      "epoch": 3.357827476038339,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 2102
+    },
+    {
+      "epoch": 3.3594249201277955,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 2103
+    },
+    {
+      "epoch": 3.3610223642172525,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 2104
+    },
+    {
+      "epoch": 3.362619808306709,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2105
+    },
+    {
+      "epoch": 3.364217252396166,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 2106
+    },
+    {
+      "epoch": 3.365814696485623,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2107
+    },
+    {
+      "epoch": 3.36741214057508,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2108
+    },
+    {
+      "epoch": 3.369009584664537,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 2109
+    },
+    {
+      "epoch": 3.3706070287539935,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2110
+    },
+    {
+      "epoch": 3.3722044728434506,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 2111
+    },
+    {
+      "epoch": 3.373801916932907,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 2112
+    },
+    {
+      "epoch": 3.3753993610223643,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 2113
+    },
+    {
+      "epoch": 3.376996805111821,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 2114
+    },
+    {
+      "epoch": 3.378594249201278,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 2115
+    },
+    {
+      "epoch": 3.380191693290735,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1706,
+      "step": 2116
+    },
+    {
+      "epoch": 3.3817891373801916,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1719,
+      "step": 2117
+    },
+    {
+      "epoch": 3.3833865814696487,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2118
+    },
+    {
+      "epoch": 3.3849840255591053,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 2119
+    },
+    {
+      "epoch": 3.3865814696485623,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 2120
+    },
+    {
+      "epoch": 3.3881789137380194,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1644,
+      "step": 2121
+    },
+    {
+      "epoch": 3.389776357827476,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 2122
+    },
+    {
+      "epoch": 3.391373801916933,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2123
+    },
+    {
+      "epoch": 3.3929712460063897,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2124
+    },
+    {
+      "epoch": 3.3945686900958467,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 2125
+    },
+    {
+      "epoch": 3.3961661341853033,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2126
+    },
+    {
+      "epoch": 3.3977635782747604,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 2127
+    },
+    {
+      "epoch": 3.3993610223642174,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 2128
+    },
+    {
+      "epoch": 3.400958466453674,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2129
+    },
+    {
+      "epoch": 3.402555910543131,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2130
+    },
+    {
+      "epoch": 3.4041533546325877,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 2131
+    },
+    {
+      "epoch": 3.405750798722045,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2132
+    },
+    {
+      "epoch": 3.4073482428115014,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2133
+    },
+    {
+      "epoch": 3.4089456869009584,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1633,
+      "step": 2134
+    },
+    {
+      "epoch": 3.4105431309904155,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2135
+    },
+    {
+      "epoch": 3.412140575079872,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1677,
+      "step": 2136
+    },
+    {
+      "epoch": 3.413738019169329,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2137
+    },
+    {
+      "epoch": 3.415335463258786,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1663,
+      "step": 2138
+    },
+    {
+      "epoch": 3.416932907348243,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 2139
+    },
+    {
+      "epoch": 3.4185303514377,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 2140
+    },
+    {
+      "epoch": 3.4201277955271565,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1667,
+      "step": 2141
+    },
+    {
+      "epoch": 3.4217252396166136,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 2142
+    },
+    {
+      "epoch": 3.42332268370607,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1652,
+      "step": 2143
+    },
+    {
+      "epoch": 3.4249201277955272,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 2144
+    },
+    {
+      "epoch": 3.426517571884984,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2145
+    },
+    {
+      "epoch": 3.428115015974441,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 2146
+    },
+    {
+      "epoch": 3.4297124600638975,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2147
+    },
+    {
+      "epoch": 3.4313099041533546,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 2148
+    },
+    {
+      "epoch": 3.4329073482428116,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 2149
+    },
+    {
+      "epoch": 3.4345047923322682,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1768,
+      "step": 2150
+    },
+    {
+      "epoch": 3.4361022364217253,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 2151
+    },
+    {
+      "epoch": 3.437699680511182,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 2152
+    },
+    {
+      "epoch": 3.439297124600639,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 2153
+    },
+    {
+      "epoch": 3.440894568690096,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1672,
+      "step": 2154
+    },
+    {
+      "epoch": 3.4424920127795526,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.169,
+      "step": 2155
+    },
+    {
+      "epoch": 3.4440894568690097,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1671,
+      "step": 2156
+    },
+    {
+      "epoch": 3.4456869009584663,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 2157
+    },
+    {
+      "epoch": 3.4472843450479234,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2158
+    },
+    {
+      "epoch": 3.4488817891373804,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1673,
+      "step": 2159
+    },
+    {
+      "epoch": 3.450479233226837,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2160
+    },
+    {
+      "epoch": 3.452076677316294,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 2161
+    },
+    {
+      "epoch": 3.4536741214057507,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2162
+    },
+    {
+      "epoch": 3.4552715654952078,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1675,
+      "step": 2163
+    },
+    {
+      "epoch": 3.4568690095846644,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 2164
+    },
+    {
+      "epoch": 3.4584664536741214,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2165
+    },
+    {
+      "epoch": 3.460063897763578,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 2166
+    },
+    {
+      "epoch": 3.461661341853035,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 2167
+    },
+    {
+      "epoch": 3.463258785942492,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1663,
+      "step": 2168
+    },
+    {
+      "epoch": 3.4648562300319488,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1701,
+      "step": 2169
+    },
+    {
+      "epoch": 3.466453674121406,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2170
+    },
+    {
+      "epoch": 3.4680511182108624,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2171
+    },
+    {
+      "epoch": 3.4696485623003195,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2172
+    },
+    {
+      "epoch": 3.4712460063897765,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1731,
+      "step": 2173
+    },
+    {
+      "epoch": 3.472843450479233,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 2174
+    },
+    {
+      "epoch": 3.47444089456869,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2175
+    },
+    {
+      "epoch": 3.476038338658147,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2176
+    },
+    {
+      "epoch": 3.477635782747604,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2177
+    },
+    {
+      "epoch": 3.479233226837061,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 2178
+    },
+    {
+      "epoch": 3.4808306709265175,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 2179
+    },
+    {
+      "epoch": 3.4824281150159746,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 2180
+    },
+    {
+      "epoch": 3.484025559105431,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 2181
+    },
+    {
+      "epoch": 3.4856230031948883,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 2182
+    },
+    {
+      "epoch": 3.487220447284345,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 2183
+    },
+    {
+      "epoch": 3.488817891373802,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 2184
+    },
+    {
+      "epoch": 3.4904153354632586,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 2185
+    },
+    {
+      "epoch": 3.4920127795527156,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 2186
+    },
+    {
+      "epoch": 3.4936102236421727,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1675,
+      "step": 2187
+    },
+    {
+      "epoch": 3.4952076677316293,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 2188
+    },
+    {
+      "epoch": 3.4968051118210863,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 2189
+    },
+    {
+      "epoch": 3.498402555910543,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 2190
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1747,
+      "step": 2191
+    },
+    {
+      "epoch": 3.501597444089457,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 2192
+    },
+    {
+      "epoch": 3.5031948881789137,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1664,
+      "step": 2193
+    },
+    {
+      "epoch": 3.5047923322683707,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1721,
+      "step": 2194
+    },
+    {
+      "epoch": 3.5063897763578273,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1688,
+      "step": 2195
+    },
+    {
+      "epoch": 3.5079872204472844,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2196
+    },
+    {
+      "epoch": 3.5095846645367414,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 2197
+    },
+    {
+      "epoch": 3.511182108626198,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 2198
+    },
+    {
+      "epoch": 3.512779552715655,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 2199
+    },
+    {
+      "epoch": 3.5143769968051117,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 2200
+    },
+    {
+      "epoch": 3.515974440894569,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1784,
+      "step": 2201
+    },
+    {
+      "epoch": 3.5175718849840254,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 2202
+    },
+    {
+      "epoch": 3.5191693290734825,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 2203
+    },
+    {
+      "epoch": 3.520766773162939,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1634,
+      "step": 2204
+    },
+    {
+      "epoch": 3.522364217252396,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1608,
+      "step": 2205
+    },
+    {
+      "epoch": 3.523961661341853,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 2206
+    },
+    {
+      "epoch": 3.52555910543131,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1684,
+      "step": 2207
+    },
+    {
+      "epoch": 3.527156549520767,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 2208
+    },
+    {
+      "epoch": 3.5287539936102235,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2209
+    },
+    {
+      "epoch": 3.5303514376996805,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1617,
+      "step": 2210
+    },
+    {
+      "epoch": 3.5319488817891376,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 2211
+    },
+    {
+      "epoch": 3.533546325878594,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2212
+    },
+    {
+      "epoch": 3.5351437699680512,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 2213
+    },
+    {
+      "epoch": 3.536741214057508,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 2214
+    },
+    {
+      "epoch": 3.538338658146965,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1732,
+      "step": 2215
+    },
+    {
+      "epoch": 3.539936102236422,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 2216
+    },
+    {
+      "epoch": 3.5415335463258786,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2217
+    },
+    {
+      "epoch": 3.543130990415335,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1755,
+      "step": 2218
+    },
+    {
+      "epoch": 3.5447284345047922,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1709,
+      "step": 2219
+    },
+    {
+      "epoch": 3.5463258785942493,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1726,
+      "step": 2220
+    },
+    {
+      "epoch": 3.547923322683706,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 2221
+    },
+    {
+      "epoch": 3.549520766773163,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 2222
+    },
+    {
+      "epoch": 3.5511182108626196,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1633,
+      "step": 2223
+    },
+    {
+      "epoch": 3.5527156549520766,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 2224
+    },
+    {
+      "epoch": 3.5543130990415337,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2225
+    },
+    {
+      "epoch": 3.5559105431309903,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1629,
+      "step": 2226
+    },
+    {
+      "epoch": 3.5575079872204474,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 2227
+    },
+    {
+      "epoch": 3.559105431309904,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2228
+    },
+    {
+      "epoch": 3.560702875399361,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 2229
+    },
+    {
+      "epoch": 3.562300319488818,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1739,
+      "step": 2230
+    },
+    {
+      "epoch": 3.5638977635782747,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2231
+    },
+    {
+      "epoch": 3.5654952076677318,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2232
+    },
+    {
+      "epoch": 3.5670926517571884,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 2233
+    },
+    {
+      "epoch": 3.5686900958466454,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1664,
+      "step": 2234
+    },
+    {
+      "epoch": 3.5702875399361025,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2235
+    },
+    {
+      "epoch": 3.571884984025559,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1706,
+      "step": 2236
+    },
+    {
+      "epoch": 3.5734824281150157,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1707,
+      "step": 2237
+    },
+    {
+      "epoch": 3.5750798722044728,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.163,
+      "step": 2238
+    },
+    {
+      "epoch": 3.57667731629393,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 2239
+    },
+    {
+      "epoch": 3.5782747603833864,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 2240
+    },
+    {
+      "epoch": 3.5798722044728435,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 2241
+    },
+    {
+      "epoch": 3.5814696485623,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1676,
+      "step": 2242
+    },
+    {
+      "epoch": 3.583067092651757,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1678,
+      "step": 2243
+    },
+    {
+      "epoch": 3.584664536741214,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2244
+    },
+    {
+      "epoch": 3.586261980830671,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 2245
+    },
+    {
+      "epoch": 3.587859424920128,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 2246
+    },
+    {
+      "epoch": 3.5894568690095845,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2247
+    },
+    {
+      "epoch": 3.5910543130990416,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1734,
+      "step": 2248
+    },
+    {
+      "epoch": 3.5926517571884986,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 2249
+    },
+    {
+      "epoch": 3.594249201277955,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1713,
+      "step": 2250
+    },
+    {
+      "epoch": 3.5958466453674123,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 2251
+    },
+    {
+      "epoch": 3.597444089456869,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 2252
+    },
+    {
+      "epoch": 3.599041533546326,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1621,
+      "step": 2253
+    },
+    {
+      "epoch": 3.600638977635783,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 2254
+    },
+    {
+      "epoch": 3.6022364217252396,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1744,
+      "step": 2255
+    },
+    {
+      "epoch": 3.6038338658146962,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1659,
+      "step": 2256
+    },
+    {
+      "epoch": 3.6054313099041533,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1623,
+      "step": 2257
+    },
+    {
+      "epoch": 3.6070287539936103,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 2258
+    },
+    {
+      "epoch": 3.608626198083067,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1694,
+      "step": 2259
+    },
+    {
+      "epoch": 3.610223642172524,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.166,
+      "step": 2260
+    },
+    {
+      "epoch": 3.6118210862619806,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 2261
+    },
+    {
+      "epoch": 3.6134185303514377,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 2262
+    },
+    {
+      "epoch": 3.6150159744408947,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1723,
+      "step": 2263
+    },
+    {
+      "epoch": 3.6166134185303513,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1724,
+      "step": 2264
+    },
+    {
+      "epoch": 3.6182108626198084,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 2265
+    },
+    {
+      "epoch": 3.619808306709265,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1708,
+      "step": 2266
+    },
+    {
+      "epoch": 3.621405750798722,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1748,
+      "step": 2267
+    },
+    {
+      "epoch": 3.623003194888179,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 2268
+    },
+    {
+      "epoch": 3.6246006389776357,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 2269
+    },
+    {
+      "epoch": 3.626198083067093,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 2270
+    },
+    {
+      "epoch": 3.6277955271565494,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2271
+    },
+    {
+      "epoch": 3.6293929712460065,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2272
+    },
+    {
+      "epoch": 3.6309904153354635,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1734,
+      "step": 2273
+    },
+    {
+      "epoch": 3.63258785942492,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 2274
+    },
+    {
+      "epoch": 3.6341853035143767,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2275
+    },
+    {
+      "epoch": 3.635782747603834,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 2276
+    },
+    {
+      "epoch": 3.637380191693291,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2277
+    },
+    {
+      "epoch": 3.6389776357827475,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 2278
+    },
+    {
+      "epoch": 3.6405750798722045,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1731,
+      "step": 2279
+    },
+    {
+      "epoch": 3.642172523961661,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2280
+    },
+    {
+      "epoch": 3.643769968051118,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2281
+    },
+    {
+      "epoch": 3.6453674121405752,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 2282
+    },
+    {
+      "epoch": 3.646964856230032,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 2283
+    },
+    {
+      "epoch": 3.648562300319489,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 2284
+    },
+    {
+      "epoch": 3.6501597444089455,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1746,
+      "step": 2285
+    },
+    {
+      "epoch": 3.6517571884984026,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 2286
+    },
+    {
+      "epoch": 3.6533546325878596,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2287
+    },
+    {
+      "epoch": 3.6549520766773163,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 2288
+    },
+    {
+      "epoch": 3.6565495207667733,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1592,
+      "step": 2289
+    },
+    {
+      "epoch": 3.65814696485623,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2290
+    },
+    {
+      "epoch": 3.659744408945687,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 2291
+    },
+    {
+      "epoch": 3.661341853035144,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 2292
+    },
+    {
+      "epoch": 3.6629392971246006,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 2293
+    },
+    {
+      "epoch": 3.6645367412140573,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1621,
+      "step": 2294
+    },
+    {
+      "epoch": 3.6661341853035143,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 2295
+    },
+    {
+      "epoch": 3.6677316293929714,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2296
+    },
+    {
+      "epoch": 3.669329073482428,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1711,
+      "step": 2297
+    },
+    {
+      "epoch": 3.670926517571885,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2298
+    },
+    {
+      "epoch": 3.6725239616613417,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 2299
+    },
+    {
+      "epoch": 3.6741214057507987,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 2300
+    },
+    {
+      "epoch": 3.6757188498402558,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1621,
+      "step": 2301
+    },
+    {
+      "epoch": 3.6773162939297124,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2302
+    },
+    {
+      "epoch": 3.6789137380191694,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 2303
+    },
+    {
+      "epoch": 3.680511182108626,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 2304
+    },
+    {
+      "epoch": 3.682108626198083,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2305
+    },
+    {
+      "epoch": 3.68370607028754,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 2306
+    },
+    {
+      "epoch": 3.6853035143769968,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 2307
+    },
+    {
+      "epoch": 3.686900958466454,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 2308
+    },
+    {
+      "epoch": 3.6884984025559104,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 2309
+    },
+    {
+      "epoch": 3.6900958466453675,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 2310
+    },
+    {
+      "epoch": 3.6916932907348246,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1783,
+      "step": 2311
+    },
+    {
+      "epoch": 3.693290734824281,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1539,
+      "step": 2312
+    },
+    {
+      "epoch": 3.6948881789137378,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1708,
+      "step": 2313
+    },
+    {
+      "epoch": 3.696485623003195,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1688,
+      "step": 2314
+    },
+    {
+      "epoch": 3.698083067092652,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 2315
+    },
+    {
+      "epoch": 3.6996805111821085,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2316
+    },
+    {
+      "epoch": 3.7012779552715656,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 2317
+    },
+    {
+      "epoch": 3.702875399361022,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 2318
+    },
+    {
+      "epoch": 3.7044728434504792,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1638,
+      "step": 2319
+    },
+    {
+      "epoch": 3.7060702875399363,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1633,
+      "step": 2320
+    },
+    {
+      "epoch": 3.707667731629393,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2321
+    },
+    {
+      "epoch": 3.70926517571885,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 2322
+    },
+    {
+      "epoch": 3.7108626198083066,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2323
+    },
+    {
+      "epoch": 3.7124600638977636,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 2324
+    },
+    {
+      "epoch": 3.7140575079872207,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2325
+    },
+    {
+      "epoch": 3.7156549520766773,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1729,
+      "step": 2326
+    },
+    {
+      "epoch": 3.7172523961661343,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 2327
+    },
+    {
+      "epoch": 3.718849840255591,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 2328
+    },
+    {
+      "epoch": 3.720447284345048,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1738,
+      "step": 2329
+    },
+    {
+      "epoch": 3.722044728434505,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 2330
+    },
+    {
+      "epoch": 3.7236421725239617,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.17,
+      "step": 2331
+    },
+    {
+      "epoch": 3.7252396166134183,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 2332
+    },
+    {
+      "epoch": 3.7268370607028753,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 2333
+    },
+    {
+      "epoch": 3.7284345047923324,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 2334
+    },
+    {
+      "epoch": 3.730031948881789,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 2335
+    },
+    {
+      "epoch": 3.731629392971246,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2336
+    },
+    {
+      "epoch": 3.7332268370607027,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2337
+    },
+    {
+      "epoch": 3.7348242811501597,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 2338
+    },
+    {
+      "epoch": 3.736421725239617,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2339
+    },
+    {
+      "epoch": 3.7380191693290734,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 2340
+    },
+    {
+      "epoch": 3.7396166134185305,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 2341
+    },
+    {
+      "epoch": 3.741214057507987,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2342
+    },
+    {
+      "epoch": 3.742811501597444,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2343
+    },
+    {
+      "epoch": 3.744408945686901,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 2344
+    },
+    {
+      "epoch": 3.746006389776358,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 2345
+    },
+    {
+      "epoch": 3.747603833865815,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 2346
+    },
+    {
+      "epoch": 3.7492012779552715,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 2347
+    },
+    {
+      "epoch": 3.7507987220447285,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 2348
+    },
+    {
+      "epoch": 3.752396166134185,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 2349
+    },
+    {
+      "epoch": 3.753993610223642,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 2350
+    },
+    {
+      "epoch": 3.755591054313099,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2351
+    },
+    {
+      "epoch": 3.757188498402556,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1656,
+      "step": 2352
+    },
+    {
+      "epoch": 3.758785942492013,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 2353
+    },
+    {
+      "epoch": 3.7603833865814695,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 2354
+    },
+    {
+      "epoch": 3.7619808306709266,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 2355
+    },
+    {
+      "epoch": 3.763578274760383,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 2356
+    },
+    {
+      "epoch": 3.7651757188498403,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1655,
+      "step": 2357
+    },
+    {
+      "epoch": 3.7667731629392973,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 2358
+    },
+    {
+      "epoch": 3.768370607028754,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 2359
+    },
+    {
+      "epoch": 3.769968051118211,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 2360
+    },
+    {
+      "epoch": 3.7715654952076676,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2361
+    },
+    {
+      "epoch": 3.7731629392971247,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2362
+    },
+    {
+      "epoch": 3.7747603833865817,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 2363
+    },
+    {
+      "epoch": 3.7763578274760383,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.167,
+      "step": 2364
+    },
+    {
+      "epoch": 3.777955271565495,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1725,
+      "step": 2365
+    },
+    {
+      "epoch": 3.779552715654952,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 2366
+    },
+    {
+      "epoch": 3.781150159744409,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 2367
+    },
+    {
+      "epoch": 3.7827476038338657,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2368
+    },
+    {
+      "epoch": 3.7843450479233227,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1616,
+      "step": 2369
+    },
+    {
+      "epoch": 3.7859424920127793,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1728,
+      "step": 2370
+    },
+    {
+      "epoch": 3.7875399361022364,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2371
+    },
+    {
+      "epoch": 3.7891373801916934,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 2372
+    },
+    {
+      "epoch": 3.79073482428115,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2373
+    },
+    {
+      "epoch": 3.792332268370607,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2374
+    },
+    {
+      "epoch": 3.7939297124600637,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2375
+    },
+    {
+      "epoch": 3.7955271565495208,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2376
+    },
+    {
+      "epoch": 3.797124600638978,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2377
+    },
+    {
+      "epoch": 3.7987220447284344,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2378
+    },
+    {
+      "epoch": 3.8003194888178915,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2379
+    },
+    {
+      "epoch": 3.801916932907348,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2380
+    },
+    {
+      "epoch": 3.803514376996805,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 2381
+    },
+    {
+      "epoch": 3.8051118210862622,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 2382
+    },
+    {
+      "epoch": 3.806709265175719,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 2383
+    },
+    {
+      "epoch": 3.8083067092651754,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.163,
+      "step": 2384
+    },
+    {
+      "epoch": 3.8099041533546325,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 2385
+    },
+    {
+      "epoch": 3.8115015974440896,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 2386
+    },
+    {
+      "epoch": 3.813099041533546,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2387
+    },
+    {
+      "epoch": 3.8146964856230032,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2388
+    },
+    {
+      "epoch": 3.81629392971246,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 2389
+    },
+    {
+      "epoch": 3.817891373801917,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1765,
+      "step": 2390
+    },
+    {
+      "epoch": 3.819488817891374,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 2391
+    },
+    {
+      "epoch": 3.8210862619808306,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1672,
+      "step": 2392
+    },
+    {
+      "epoch": 3.8226837060702876,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 2393
+    },
+    {
+      "epoch": 3.8242811501597442,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 2394
+    },
+    {
+      "epoch": 3.8258785942492013,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2395
+    },
+    {
+      "epoch": 3.8274760383386583,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1706,
+      "step": 2396
+    },
+    {
+      "epoch": 3.829073482428115,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 2397
+    },
+    {
+      "epoch": 3.830670926517572,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2398
+    },
+    {
+      "epoch": 3.8322683706070286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 2399
+    },
+    {
+      "epoch": 3.8338658146964857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2400
+    },
+    {
+      "epoch": 3.8354632587859427,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2401
+    },
+    {
+      "epoch": 3.8370607028753994,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 2402
+    },
+    {
+      "epoch": 3.838658146964856,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1683,
+      "step": 2403
+    },
+    {
+      "epoch": 3.840255591054313,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2404
+    },
+    {
+      "epoch": 3.84185303514377,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2405
+    },
+    {
+      "epoch": 3.8434504792332267,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2406
+    },
+    {
+      "epoch": 3.8450479233226837,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 2407
+    },
+    {
+      "epoch": 3.8466453674121404,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 2408
+    },
+    {
+      "epoch": 3.8482428115015974,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 2409
+    },
+    {
+      "epoch": 3.8498402555910545,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 2410
+    },
+    {
+      "epoch": 3.851437699680511,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1698,
+      "step": 2411
+    },
+    {
+      "epoch": 3.853035143769968,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2412
+    },
+    {
+      "epoch": 3.8546325878594248,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 2413
+    },
+    {
+      "epoch": 3.856230031948882,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 2414
+    },
+    {
+      "epoch": 3.857827476038339,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 2415
+    },
+    {
+      "epoch": 3.8594249201277955,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 2416
+    },
+    {
+      "epoch": 3.8610223642172525,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1595,
+      "step": 2417
+    },
+    {
+      "epoch": 3.862619808306709,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 2418
+    },
+    {
+      "epoch": 3.864217252396166,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2419
+    },
+    {
+      "epoch": 3.8658146964856233,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 2420
+    },
+    {
+      "epoch": 3.86741214057508,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1655,
+      "step": 2421
+    },
+    {
+      "epoch": 3.8690095846645365,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1735,
+      "step": 2422
+    },
+    {
+      "epoch": 3.8706070287539935,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 2423
+    },
+    {
+      "epoch": 3.8722044728434506,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2424
+    },
+    {
+      "epoch": 3.873801916932907,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 2425
+    },
+    {
+      "epoch": 3.8753993610223643,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2426
+    },
+    {
+      "epoch": 3.876996805111821,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2427
+    },
+    {
+      "epoch": 3.878594249201278,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 2428
+    },
+    {
+      "epoch": 3.880191693290735,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1672,
+      "step": 2429
+    },
+    {
+      "epoch": 3.8817891373801916,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 2430
+    },
+    {
+      "epoch": 3.8833865814696487,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 2431
+    },
+    {
+      "epoch": 3.8849840255591053,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 2432
+    },
+    {
+      "epoch": 3.8865814696485623,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1623,
+      "step": 2433
+    },
+    {
+      "epoch": 3.8881789137380194,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 2434
+    },
+    {
+      "epoch": 3.889776357827476,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2435
+    },
+    {
+      "epoch": 3.891373801916933,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 2436
+    },
+    {
+      "epoch": 3.8929712460063897,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 2437
+    },
+    {
+      "epoch": 3.8945686900958467,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1644,
+      "step": 2438
+    },
+    {
+      "epoch": 3.8961661341853038,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 2439
+    },
+    {
+      "epoch": 3.8977635782747604,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 2440
+    },
+    {
+      "epoch": 3.899361022364217,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 2441
+    },
+    {
+      "epoch": 3.900958466453674,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 2442
+    },
+    {
+      "epoch": 3.902555910543131,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 2443
+    },
+    {
+      "epoch": 3.9041533546325877,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 2444
+    },
+    {
+      "epoch": 3.905750798722045,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 2445
+    },
+    {
+      "epoch": 3.9073482428115014,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 2446
+    },
+    {
+      "epoch": 3.9089456869009584,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 2447
+    },
+    {
+      "epoch": 3.9105431309904155,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 2448
+    },
+    {
+      "epoch": 3.912140575079872,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 2449
+    },
+    {
+      "epoch": 3.913738019169329,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 2450
+    },
+    {
+      "epoch": 3.915335463258786,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 2451
+    },
+    {
+      "epoch": 3.916932907348243,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1715,
+      "step": 2452
+    },
+    {
+      "epoch": 3.9185303514377,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 2453
+    },
+    {
+      "epoch": 3.9201277955271565,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 2454
+    },
+    {
+      "epoch": 3.9217252396166136,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 2455
+    },
+    {
+      "epoch": 3.92332268370607,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 2456
+    },
+    {
+      "epoch": 3.9249201277955272,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 2457
+    },
+    {
+      "epoch": 3.9265175718849843,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2458
+    },
+    {
+      "epoch": 3.928115015974441,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 2459
+    },
+    {
+      "epoch": 3.9297124600638975,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 2460
+    },
+    {
+      "epoch": 3.9313099041533546,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 2461
+    },
+    {
+      "epoch": 3.9329073482428116,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2462
+    },
+    {
+      "epoch": 3.9345047923322682,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 2463
+    },
+    {
+      "epoch": 3.9361022364217253,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 2464
+    },
+    {
+      "epoch": 3.937699680511182,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 2465
+    },
+    {
+      "epoch": 3.939297124600639,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 2466
+    },
+    {
+      "epoch": 3.940894568690096,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 2467
+    },
+    {
+      "epoch": 3.9424920127795526,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1608,
+      "step": 2468
+    },
+    {
+      "epoch": 3.9440894568690097,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2469
+    },
+    {
+      "epoch": 3.9456869009584663,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1595,
+      "step": 2470
+    },
+    {
+      "epoch": 3.9472843450479234,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2471
+    },
+    {
+      "epoch": 3.9488817891373804,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 2472
+    },
+    {
+      "epoch": 3.950479233226837,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 2473
+    },
+    {
+      "epoch": 3.952076677316294,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 2474
+    },
+    {
+      "epoch": 3.9536741214057507,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 2475
+    },
+    {
+      "epoch": 3.9552715654952078,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.155,
+      "step": 2476
+    },
+    {
+      "epoch": 3.956869009584665,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 2477
+    },
+    {
+      "epoch": 3.9584664536741214,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2478
+    },
+    {
+      "epoch": 3.960063897763578,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1667,
+      "step": 2479
+    },
+    {
+      "epoch": 3.961661341853035,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 2480
+    },
+    {
+      "epoch": 3.963258785942492,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2481
+    },
+    {
+      "epoch": 3.9648562300319488,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 2482
+    },
+    {
+      "epoch": 3.966453674121406,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1644,
+      "step": 2483
+    },
+    {
+      "epoch": 3.9680511182108624,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 2484
+    },
+    {
+      "epoch": 3.9696485623003195,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2485
+    },
+    {
+      "epoch": 3.9712460063897765,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1698,
+      "step": 2486
+    },
+    {
+      "epoch": 3.972843450479233,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2487
+    },
+    {
+      "epoch": 3.97444089456869,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2488
+    },
+    {
+      "epoch": 3.976038338658147,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 2489
+    },
+    {
+      "epoch": 3.977635782747604,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 2490
+    },
+    {
+      "epoch": 3.979233226837061,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2491
+    },
+    {
+      "epoch": 3.9808306709265175,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1608,
+      "step": 2492
+    },
+    {
+      "epoch": 3.9824281150159746,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 2493
+    },
+    {
+      "epoch": 3.984025559105431,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2494
+    },
+    {
+      "epoch": 3.9856230031948883,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 2495
+    },
+    {
+      "epoch": 3.987220447284345,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.173,
+      "step": 2496
+    },
+    {
+      "epoch": 3.988817891373802,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 2497
+    },
+    {
+      "epoch": 3.9904153354632586,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2498
+    },
+    {
+      "epoch": 3.9920127795527156,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2499
+    },
+    {
+      "epoch": 3.9936102236421727,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2500
+    },
+    {
+      "epoch": 3.9952076677316293,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2501
+    },
+    {
+      "epoch": 3.9968051118210863,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2502
+    },
+    {
+      "epoch": 3.998402555910543,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2503
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2504
+    },
+    {
+      "epoch": 4.001597444089457,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 2505
+    },
+    {
+      "epoch": 4.003194888178914,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2506
+    },
+    {
+      "epoch": 4.00479233226837,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1664,
+      "step": 2507
+    },
+    {
+      "epoch": 4.006389776357827,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2508
+    },
+    {
+      "epoch": 4.007987220447284,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2509
+    },
+    {
+      "epoch": 4.0095846645367414,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2510
+    },
+    {
+      "epoch": 4.0111821086261985,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 2511
+    },
+    {
+      "epoch": 4.012779552715655,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2512
+    },
+    {
+      "epoch": 4.014376996805112,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2513
+    },
+    {
+      "epoch": 4.015974440894569,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2514
+    },
+    {
+      "epoch": 4.017571884984026,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1616,
+      "step": 2515
+    },
+    {
+      "epoch": 4.019169329073482,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2516
+    },
+    {
+      "epoch": 4.020766773162939,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 2517
+    },
+    {
+      "epoch": 4.022364217252396,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 2518
+    },
+    {
+      "epoch": 4.023961661341853,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1629,
+      "step": 2519
+    },
+    {
+      "epoch": 4.02555910543131,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 2520
+    },
+    {
+      "epoch": 4.027156549520766,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 2521
+    },
+    {
+      "epoch": 4.0287539936102235,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2522
+    },
+    {
+      "epoch": 4.0303514376996805,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 2523
+    },
+    {
+      "epoch": 4.031948881789138,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2524
+    },
+    {
+      "epoch": 4.033546325878595,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 2525
+    },
+    {
+      "epoch": 4.035143769968051,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2526
+    },
+    {
+      "epoch": 4.036741214057508,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2527
+    },
+    {
+      "epoch": 4.038338658146965,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 2528
+    },
+    {
+      "epoch": 4.039936102236422,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 2529
+    },
+    {
+      "epoch": 4.041533546325879,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 2530
+    },
+    {
+      "epoch": 4.043130990415335,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 2531
+    },
+    {
+      "epoch": 4.044728434504792,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 2532
+    },
+    {
+      "epoch": 4.046325878594249,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 2533
+    },
+    {
+      "epoch": 4.047923322683706,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2534
+    },
+    {
+      "epoch": 4.0495207667731625,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 2535
+    },
+    {
+      "epoch": 4.05111821086262,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2536
+    },
+    {
+      "epoch": 4.052715654952077,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 2537
+    },
+    {
+      "epoch": 4.054313099041534,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.165,
+      "step": 2538
+    },
+    {
+      "epoch": 4.055910543130991,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 2539
+    },
+    {
+      "epoch": 4.057507987220447,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 2540
+    },
+    {
+      "epoch": 4.059105431309904,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 2541
+    },
+    {
+      "epoch": 4.060702875399361,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1608,
+      "step": 2542
+    },
+    {
+      "epoch": 4.062300319488818,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 2543
+    },
+    {
+      "epoch": 4.063897763578275,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1646,
+      "step": 2544
+    },
+    {
+      "epoch": 4.065495207667731,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 2545
+    },
+    {
+      "epoch": 4.067092651757188,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2546
+    },
+    {
+      "epoch": 4.068690095846645,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 2547
+    },
+    {
+      "epoch": 4.0702875399361025,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 2548
+    },
+    {
+      "epoch": 4.0718849840255595,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2549
+    },
+    {
+      "epoch": 4.073482428115016,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 2550
+    },
+    {
+      "epoch": 4.075079872204473,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 2551
+    },
+    {
+      "epoch": 4.07667731629393,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2552
+    },
+    {
+      "epoch": 4.078274760383387,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1731,
+      "step": 2553
+    },
+    {
+      "epoch": 4.079872204472843,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 2554
+    },
+    {
+      "epoch": 4.0814696485623,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 2555
+    },
+    {
+      "epoch": 4.083067092651757,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1617,
+      "step": 2556
+    },
+    {
+      "epoch": 4.084664536741214,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2557
+    },
+    {
+      "epoch": 4.086261980830671,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 2558
+    },
+    {
+      "epoch": 4.087859424920127,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 2559
+    },
+    {
+      "epoch": 4.0894568690095845,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 2560
+    },
+    {
+      "epoch": 4.0910543130990416,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 2561
+    },
+    {
+      "epoch": 4.092651757188499,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2562
+    },
+    {
+      "epoch": 4.094249201277956,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 2563
+    },
+    {
+      "epoch": 4.095846645367412,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 2564
+    },
+    {
+      "epoch": 4.097444089456869,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 2565
+    },
+    {
+      "epoch": 4.099041533546326,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 2566
+    },
+    {
+      "epoch": 4.100638977635783,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.156,
+      "step": 2567
+    },
+    {
+      "epoch": 4.102236421725239,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 2568
+    },
+    {
+      "epoch": 4.103833865814696,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1592,
+      "step": 2569
+    },
+    {
+      "epoch": 4.105431309904153,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2570
+    },
+    {
+      "epoch": 4.10702875399361,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 2571
+    },
+    {
+      "epoch": 4.108626198083067,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1634,
+      "step": 2572
+    },
+    {
+      "epoch": 4.110223642172524,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 2573
+    },
+    {
+      "epoch": 4.111821086261981,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 2574
+    },
+    {
+      "epoch": 4.113418530351438,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 2575
+    },
+    {
+      "epoch": 4.115015974440895,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 2576
+    },
+    {
+      "epoch": 4.116613418530352,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 2577
+    },
+    {
+      "epoch": 4.118210862619808,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 2578
+    },
+    {
+      "epoch": 4.119808306709265,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2579
+    },
+    {
+      "epoch": 4.121405750798722,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 2580
+    },
+    {
+      "epoch": 4.123003194888179,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 2581
+    },
+    {
+      "epoch": 4.124600638977636,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 2582
+    },
+    {
+      "epoch": 4.126198083067092,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2583
+    },
+    {
+      "epoch": 4.127795527156549,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 2584
+    },
+    {
+      "epoch": 4.1293929712460065,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1621,
+      "step": 2585
+    },
+    {
+      "epoch": 4.1309904153354635,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 2586
+    },
+    {
+      "epoch": 4.13258785942492,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2587
+    },
+    {
+      "epoch": 4.134185303514377,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2588
+    },
+    {
+      "epoch": 4.135782747603834,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.171,
+      "step": 2589
+    },
+    {
+      "epoch": 4.137380191693291,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 2590
+    },
+    {
+      "epoch": 4.138977635782748,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2591
+    },
+    {
+      "epoch": 4.140575079872204,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 2592
+    },
+    {
+      "epoch": 4.142172523961661,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.155,
+      "step": 2593
+    },
+    {
+      "epoch": 4.143769968051118,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2594
+    },
+    {
+      "epoch": 4.145367412140575,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1665,
+      "step": 2595
+    },
+    {
+      "epoch": 4.146964856230032,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2596
+    },
+    {
+      "epoch": 4.1485623003194885,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 2597
+    },
+    {
+      "epoch": 4.1501597444089455,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2598
+    },
+    {
+      "epoch": 4.151757188498403,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 2599
+    },
+    {
+      "epoch": 4.15335463258786,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2600
+    },
+    {
+      "epoch": 4.154952076677317,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2601
+    },
+    {
+      "epoch": 4.156549520766773,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 2602
+    },
+    {
+      "epoch": 4.15814696485623,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 2603
+    },
+    {
+      "epoch": 4.159744408945687,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1666,
+      "step": 2604
+    },
+    {
+      "epoch": 4.161341853035144,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 2605
+    },
+    {
+      "epoch": 4.1629392971246,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1629,
+      "step": 2606
+    },
+    {
+      "epoch": 4.164536741214057,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2607
+    },
+    {
+      "epoch": 4.166134185303514,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 2608
+    },
+    {
+      "epoch": 4.167731629392971,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 2609
+    },
+    {
+      "epoch": 4.169329073482428,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 2610
+    },
+    {
+      "epoch": 4.170926517571885,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2611
+    },
+    {
+      "epoch": 4.172523961661342,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1667,
+      "step": 2612
+    },
+    {
+      "epoch": 4.174121405750799,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 2613
+    },
+    {
+      "epoch": 4.175718849840256,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2614
+    },
+    {
+      "epoch": 4.177316293929713,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 2615
+    },
+    {
+      "epoch": 4.178913738019169,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 2616
+    },
+    {
+      "epoch": 4.180511182108626,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 2617
+    },
+    {
+      "epoch": 4.182108626198083,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 2618
+    },
+    {
+      "epoch": 4.18370607028754,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 2619
+    },
+    {
+      "epoch": 4.185303514376997,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 2620
+    },
+    {
+      "epoch": 4.186900958466453,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 2621
+    },
+    {
+      "epoch": 4.18849840255591,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2622
+    },
+    {
+      "epoch": 4.1900958466453675,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 2623
+    },
+    {
+      "epoch": 4.1916932907348246,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 2624
+    },
+    {
+      "epoch": 4.193290734824281,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 2625
+    },
+    {
+      "epoch": 4.194888178913738,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 2626
+    },
+    {
+      "epoch": 4.196485623003195,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 2627
+    },
+    {
+      "epoch": 4.198083067092652,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1655,
+      "step": 2628
+    },
+    {
+      "epoch": 4.199680511182109,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 2629
+    },
+    {
+      "epoch": 4.201277955271565,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1646,
+      "step": 2630
+    },
+    {
+      "epoch": 4.202875399361022,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1592,
+      "step": 2631
+    },
+    {
+      "epoch": 4.204472843450479,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2632
+    },
+    {
+      "epoch": 4.206070287539936,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2633
+    },
+    {
+      "epoch": 4.207667731629393,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2634
+    },
+    {
+      "epoch": 4.2092651757188495,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 2635
+    },
+    {
+      "epoch": 4.210862619808307,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1603,
+      "step": 2636
+    },
+    {
+      "epoch": 4.212460063897764,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 2637
+    },
+    {
+      "epoch": 4.214057507987221,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 2638
+    },
+    {
+      "epoch": 4.215654952076678,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 2639
+    },
+    {
+      "epoch": 4.217252396166134,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 2640
+    },
+    {
+      "epoch": 4.218849840255591,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 2641
+    },
+    {
+      "epoch": 4.220447284345048,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2642
+    },
+    {
+      "epoch": 4.222044728434505,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 2643
+    },
+    {
+      "epoch": 4.223642172523961,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2644
+    },
+    {
+      "epoch": 4.225239616613418,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2645
+    },
+    {
+      "epoch": 4.226837060702875,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2646
+    },
+    {
+      "epoch": 4.228434504792332,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 2647
+    },
+    {
+      "epoch": 4.2300319488817895,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 2648
+    },
+    {
+      "epoch": 4.231629392971246,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1623,
+      "step": 2649
+    },
+    {
+      "epoch": 4.233226837060703,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2650
+    },
+    {
+      "epoch": 4.23482428115016,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1703,
+      "step": 2651
+    },
+    {
+      "epoch": 4.236421725239617,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2652
+    },
+    {
+      "epoch": 4.238019169329074,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2653
+    },
+    {
+      "epoch": 4.23961661341853,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2654
+    },
+    {
+      "epoch": 4.241214057507987,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1675,
+      "step": 2655
+    },
+    {
+      "epoch": 4.242811501597444,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2656
+    },
+    {
+      "epoch": 4.244408945686901,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1701,
+      "step": 2657
+    },
+    {
+      "epoch": 4.246006389776358,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2658
+    },
+    {
+      "epoch": 4.247603833865814,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1645,
+      "step": 2659
+    },
+    {
+      "epoch": 4.2492012779552715,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 2660
+    },
+    {
+      "epoch": 4.2507987220447285,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2661
+    },
+    {
+      "epoch": 4.252396166134186,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 2662
+    },
+    {
+      "epoch": 4.253993610223642,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 2663
+    },
+    {
+      "epoch": 4.255591054313099,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 2664
+    },
+    {
+      "epoch": 4.257188498402556,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 2665
+    },
+    {
+      "epoch": 4.258785942492013,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 2666
+    },
+    {
+      "epoch": 4.26038338658147,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 2667
+    },
+    {
+      "epoch": 4.261980830670926,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 2668
+    },
+    {
+      "epoch": 4.263578274760383,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 2669
+    },
+    {
+      "epoch": 4.26517571884984,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1629,
+      "step": 2670
+    },
+    {
+      "epoch": 4.266773162939297,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 2671
+    },
+    {
+      "epoch": 4.268370607028754,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2672
+    },
+    {
+      "epoch": 4.2699680511182105,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2673
+    },
+    {
+      "epoch": 4.271565495207668,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 2674
+    },
+    {
+      "epoch": 4.273162939297125,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2675
+    },
+    {
+      "epoch": 4.274760383386582,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 2676
+    },
+    {
+      "epoch": 4.276357827476039,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 2677
+    },
+    {
+      "epoch": 4.277955271565495,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 2678
+    },
+    {
+      "epoch": 4.279552715654952,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 2679
+    },
+    {
+      "epoch": 4.281150159744409,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1617,
+      "step": 2680
+    },
+    {
+      "epoch": 4.282747603833866,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2681
+    },
+    {
+      "epoch": 4.284345047923322,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 2682
+    },
+    {
+      "epoch": 4.285942492012779,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 2683
+    },
+    {
+      "epoch": 4.287539936102236,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2684
+    },
+    {
+      "epoch": 4.289137380191693,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1633,
+      "step": 2685
+    },
+    {
+      "epoch": 4.2907348242811505,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 2686
+    },
+    {
+      "epoch": 4.292332268370607,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 2687
+    },
+    {
+      "epoch": 4.293929712460064,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 2688
+    },
+    {
+      "epoch": 4.295527156549521,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2689
+    },
+    {
+      "epoch": 4.297124600638978,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2690
+    },
+    {
+      "epoch": 4.298722044728435,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 2691
+    },
+    {
+      "epoch": 4.300319488817891,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2692
+    },
+    {
+      "epoch": 4.301916932907348,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 2693
+    },
+    {
+      "epoch": 4.303514376996805,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1698,
+      "step": 2694
+    },
+    {
+      "epoch": 4.305111821086262,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 2695
+    },
+    {
+      "epoch": 4.306709265175719,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 2696
+    },
+    {
+      "epoch": 4.3083067092651754,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 2697
+    },
+    {
+      "epoch": 4.3099041533546325,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 2698
+    },
+    {
+      "epoch": 4.31150159744409,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2699
+    },
+    {
+      "epoch": 4.313099041533547,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 2700
+    },
+    {
+      "epoch": 4.314696485623003,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 2701
+    },
+    {
+      "epoch": 4.31629392971246,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 2702
+    },
+    {
+      "epoch": 4.317891373801917,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 2703
+    },
+    {
+      "epoch": 4.319488817891374,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 2704
+    },
+    {
+      "epoch": 4.321086261980831,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 2705
+    },
+    {
+      "epoch": 4.322683706070287,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2706
+    },
+    {
+      "epoch": 4.324281150159744,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 2707
+    },
+    {
+      "epoch": 4.325878594249201,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 2708
+    },
+    {
+      "epoch": 4.327476038338658,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 2709
+    },
+    {
+      "epoch": 4.329073482428115,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 2710
+    },
+    {
+      "epoch": 4.330670926517572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 2711
+    },
+    {
+      "epoch": 4.332268370607029,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 2712
+    },
+    {
+      "epoch": 4.333865814696486,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 2713
+    },
+    {
+      "epoch": 4.335463258785943,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 2714
+    },
+    {
+      "epoch": 4.3370607028754,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 2715
+    },
+    {
+      "epoch": 4.338658146964856,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 2716
+    },
+    {
+      "epoch": 4.340255591054313,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 2717
+    },
+    {
+      "epoch": 4.34185303514377,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 2718
+    },
+    {
+      "epoch": 4.343450479233227,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1629,
+      "step": 2719
+    },
+    {
+      "epoch": 4.345047923322683,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 2720
+    },
+    {
+      "epoch": 4.34664536741214,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 2721
+    },
+    {
+      "epoch": 4.348242811501597,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 2722
+    },
+    {
+      "epoch": 4.3498402555910545,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2723
+    },
+    {
+      "epoch": 4.3514376996805115,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1603,
+      "step": 2724
+    },
+    {
+      "epoch": 4.353035143769968,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 2725
+    },
+    {
+      "epoch": 4.354632587859425,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 2726
+    },
+    {
+      "epoch": 4.356230031948882,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 2727
+    },
+    {
+      "epoch": 4.357827476038339,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 2728
+    },
+    {
+      "epoch": 4.359424920127796,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 2729
+    },
+    {
+      "epoch": 4.361022364217252,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1576,
+      "step": 2730
+    },
+    {
+      "epoch": 4.362619808306709,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 2731
+    },
+    {
+      "epoch": 4.364217252396166,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2732
+    },
+    {
+      "epoch": 4.365814696485623,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 2733
+    },
+    {
+      "epoch": 4.36741214057508,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2734
+    },
+    {
+      "epoch": 4.3690095846645365,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 2735
+    },
+    {
+      "epoch": 4.3706070287539935,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2736
+    },
+    {
+      "epoch": 4.372204472843451,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 2737
+    },
+    {
+      "epoch": 4.373801916932908,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 2738
+    },
+    {
+      "epoch": 4.375399361022364,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 2739
+    },
+    {
+      "epoch": 4.376996805111821,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 2740
+    },
+    {
+      "epoch": 4.378594249201278,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 2741
+    },
+    {
+      "epoch": 4.380191693290735,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2742
+    },
+    {
+      "epoch": 4.381789137380192,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 2743
+    },
+    {
+      "epoch": 4.383386581469648,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 2744
+    },
+    {
+      "epoch": 4.384984025559105,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2745
+    },
+    {
+      "epoch": 4.386581469648562,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1661,
+      "step": 2746
+    },
+    {
+      "epoch": 4.388178913738019,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 2747
+    },
+    {
+      "epoch": 4.389776357827476,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 2748
+    },
+    {
+      "epoch": 4.391373801916933,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 2749
+    },
+    {
+      "epoch": 4.39297124600639,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 2750
+    },
+    {
+      "epoch": 4.394568690095847,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 2751
+    },
+    {
+      "epoch": 4.396166134185304,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2752
+    },
+    {
+      "epoch": 4.397763578274761,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 2753
+    },
+    {
+      "epoch": 4.399361022364217,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 2754
+    },
+    {
+      "epoch": 4.400958466453674,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 2755
+    },
+    {
+      "epoch": 4.402555910543131,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 2756
+    },
+    {
+      "epoch": 4.404153354632588,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 2757
+    },
+    {
+      "epoch": 4.405750798722044,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 2758
+    },
+    {
+      "epoch": 4.407348242811501,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 2759
+    },
+    {
+      "epoch": 4.4089456869009584,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 2760
+    },
+    {
+      "epoch": 4.4105431309904155,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 2761
+    },
+    {
+      "epoch": 4.412140575079873,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 2762
+    },
+    {
+      "epoch": 4.413738019169329,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2763
+    },
+    {
+      "epoch": 4.415335463258786,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2764
+    },
+    {
+      "epoch": 4.416932907348243,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 2765
+    },
+    {
+      "epoch": 4.4185303514377,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2766
+    },
+    {
+      "epoch": 4.420127795527157,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 2767
+    },
+    {
+      "epoch": 4.421725239616613,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 2768
+    },
+    {
+      "epoch": 4.42332268370607,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 2769
+    },
+    {
+      "epoch": 4.424920127795527,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 2770
+    },
+    {
+      "epoch": 4.426517571884984,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 2771
+    },
+    {
+      "epoch": 4.428115015974441,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 2772
+    },
+    {
+      "epoch": 4.4297124600638975,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 2773
+    },
+    {
+      "epoch": 4.431309904153355,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 2774
+    },
+    {
+      "epoch": 4.432907348242812,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.169,
+      "step": 2775
+    },
+    {
+      "epoch": 4.434504792332269,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1695,
+      "step": 2776
+    },
+    {
+      "epoch": 4.436102236421725,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 2777
+    },
+    {
+      "epoch": 4.437699680511182,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2778
+    },
+    {
+      "epoch": 4.439297124600639,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2779
+    },
+    {
+      "epoch": 4.440894568690096,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 2780
+    },
+    {
+      "epoch": 4.442492012779553,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 2781
+    },
+    {
+      "epoch": 4.444089456869009,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2782
+    },
+    {
+      "epoch": 4.445686900958466,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 2783
+    },
+    {
+      "epoch": 4.447284345047923,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 2784
+    },
+    {
+      "epoch": 4.44888178913738,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 2785
+    },
+    {
+      "epoch": 4.4504792332268375,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2786
+    },
+    {
+      "epoch": 4.452076677316294,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2787
+    },
+    {
+      "epoch": 4.453674121405751,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 2788
+    },
+    {
+      "epoch": 4.455271565495208,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1633,
+      "step": 2789
+    },
+    {
+      "epoch": 4.456869009584665,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2790
+    },
+    {
+      "epoch": 4.458466453674122,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 2791
+    },
+    {
+      "epoch": 4.460063897763578,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2792
+    },
+    {
+      "epoch": 4.461661341853035,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 2793
+    },
+    {
+      "epoch": 4.463258785942492,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 2794
+    },
+    {
+      "epoch": 4.464856230031949,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 2795
+    },
+    {
+      "epoch": 4.466453674121405,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2796
+    },
+    {
+      "epoch": 4.468051118210862,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 2797
+    },
+    {
+      "epoch": 4.4696485623003195,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 2798
+    },
+    {
+      "epoch": 4.4712460063897765,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 2799
+    },
+    {
+      "epoch": 4.472843450479234,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 2800
+    },
+    {
+      "epoch": 4.47444089456869,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 2801
+    },
+    {
+      "epoch": 4.476038338658147,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 2802
+    },
+    {
+      "epoch": 4.477635782747604,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 2803
+    },
+    {
+      "epoch": 4.479233226837061,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 2804
+    },
+    {
+      "epoch": 4.480830670926517,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 2805
+    },
+    {
+      "epoch": 4.482428115015974,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 2806
+    },
+    {
+      "epoch": 4.484025559105431,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 2807
+    },
+    {
+      "epoch": 4.485623003194888,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2808
+    },
+    {
+      "epoch": 4.487220447284345,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2809
+    },
+    {
+      "epoch": 4.488817891373802,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 2810
+    },
+    {
+      "epoch": 4.4904153354632586,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2811
+    },
+    {
+      "epoch": 4.492012779552716,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2812
+    },
+    {
+      "epoch": 4.493610223642173,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 2813
+    },
+    {
+      "epoch": 4.49520766773163,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2814
+    },
+    {
+      "epoch": 4.496805111821086,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1667,
+      "step": 2815
+    },
+    {
+      "epoch": 4.498402555910543,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 2816
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 2817
+    },
+    {
+      "epoch": 4.501597444089457,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 2818
+    },
+    {
+      "epoch": 4.503194888178914,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 2819
+    },
+    {
+      "epoch": 4.50479233226837,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 2820
+    },
+    {
+      "epoch": 4.506389776357827,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2821
+    },
+    {
+      "epoch": 4.507987220447284,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 2822
+    },
+    {
+      "epoch": 4.5095846645367414,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 2823
+    },
+    {
+      "epoch": 4.511182108626198,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 2824
+    },
+    {
+      "epoch": 4.512779552715655,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2825
+    },
+    {
+      "epoch": 4.514376996805112,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 2826
+    },
+    {
+      "epoch": 4.515974440894569,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 2827
+    },
+    {
+      "epoch": 4.517571884984026,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 2828
+    },
+    {
+      "epoch": 4.519169329073483,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 2829
+    },
+    {
+      "epoch": 4.520766773162939,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 2830
+    },
+    {
+      "epoch": 4.522364217252396,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2831
+    },
+    {
+      "epoch": 4.523961661341853,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2832
+    },
+    {
+      "epoch": 4.52555910543131,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2833
+    },
+    {
+      "epoch": 4.527156549520766,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 2834
+    },
+    {
+      "epoch": 4.5287539936102235,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 2835
+    },
+    {
+      "epoch": 4.5303514376996805,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 2836
+    },
+    {
+      "epoch": 4.531948881789138,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 2837
+    },
+    {
+      "epoch": 4.533546325878595,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 2838
+    },
+    {
+      "epoch": 4.535143769968051,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2839
+    },
+    {
+      "epoch": 4.536741214057508,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2840
+    },
+    {
+      "epoch": 4.538338658146965,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 2841
+    },
+    {
+      "epoch": 4.539936102236422,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 2842
+    },
+    {
+      "epoch": 4.541533546325878,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 2843
+    },
+    {
+      "epoch": 4.543130990415335,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 2844
+    },
+    {
+      "epoch": 4.544728434504792,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 2845
+    },
+    {
+      "epoch": 4.546325878594249,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 2846
+    },
+    {
+      "epoch": 4.547923322683706,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 2847
+    },
+    {
+      "epoch": 4.549520766773163,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 2848
+    },
+    {
+      "epoch": 4.55111821086262,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 2849
+    },
+    {
+      "epoch": 4.552715654952077,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 2850
+    },
+    {
+      "epoch": 4.554313099041534,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 2851
+    },
+    {
+      "epoch": 4.555910543130991,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 2852
+    },
+    {
+      "epoch": 4.557507987220447,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 2853
+    },
+    {
+      "epoch": 4.559105431309904,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 2854
+    },
+    {
+      "epoch": 4.560702875399361,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 2855
+    },
+    {
+      "epoch": 4.562300319488818,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 2856
+    },
+    {
+      "epoch": 4.563897763578275,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 2857
+    },
+    {
+      "epoch": 4.565495207667731,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 2858
+    },
+    {
+      "epoch": 4.567092651757188,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2859
+    },
+    {
+      "epoch": 4.568690095846645,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2860
+    },
+    {
+      "epoch": 4.5702875399361025,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 2861
+    },
+    {
+      "epoch": 4.571884984025559,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2862
+    },
+    {
+      "epoch": 4.573482428115016,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2863
+    },
+    {
+      "epoch": 4.575079872204473,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1662,
+      "step": 2864
+    },
+    {
+      "epoch": 4.57667731629393,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1665,
+      "step": 2865
+    },
+    {
+      "epoch": 4.578274760383387,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2866
+    },
+    {
+      "epoch": 4.579872204472844,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1576,
+      "step": 2867
+    },
+    {
+      "epoch": 4.5814696485623,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 2868
+    },
+    {
+      "epoch": 4.583067092651757,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 2869
+    },
+    {
+      "epoch": 4.584664536741214,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 2870
+    },
+    {
+      "epoch": 4.586261980830671,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 2871
+    },
+    {
+      "epoch": 4.587859424920127,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2872
+    },
+    {
+      "epoch": 4.5894568690095845,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 2873
+    },
+    {
+      "epoch": 4.5910543130990416,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 2874
+    },
+    {
+      "epoch": 4.592651757188499,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 2875
+    },
+    {
+      "epoch": 4.594249201277956,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 2876
+    },
+    {
+      "epoch": 4.595846645367412,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1642,
+      "step": 2877
+    },
+    {
+      "epoch": 4.597444089456869,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2878
+    },
+    {
+      "epoch": 4.599041533546326,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 2879
+    },
+    {
+      "epoch": 4.600638977635783,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 2880
+    },
+    {
+      "epoch": 4.602236421725239,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2881
+    },
+    {
+      "epoch": 4.603833865814696,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 2882
+    },
+    {
+      "epoch": 4.605431309904153,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 2883
+    },
+    {
+      "epoch": 4.60702875399361,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 2884
+    },
+    {
+      "epoch": 4.608626198083067,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 2885
+    },
+    {
+      "epoch": 4.6102236421725244,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 2886
+    },
+    {
+      "epoch": 4.611821086261981,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 2887
+    },
+    {
+      "epoch": 4.613418530351438,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1638,
+      "step": 2888
+    },
+    {
+      "epoch": 4.615015974440895,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1694,
+      "step": 2889
+    },
+    {
+      "epoch": 4.616613418530352,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 2890
+    },
+    {
+      "epoch": 4.618210862619808,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 2891
+    },
+    {
+      "epoch": 4.619808306709265,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2892
+    },
+    {
+      "epoch": 4.621405750798722,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 2893
+    },
+    {
+      "epoch": 4.623003194888179,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 2894
+    },
+    {
+      "epoch": 4.624600638977636,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 2895
+    },
+    {
+      "epoch": 4.626198083067092,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 2896
+    },
+    {
+      "epoch": 4.627795527156549,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 2897
+    },
+    {
+      "epoch": 4.6293929712460065,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2898
+    },
+    {
+      "epoch": 4.6309904153354635,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 2899
+    },
+    {
+      "epoch": 4.63258785942492,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 2900
+    },
+    {
+      "epoch": 4.634185303514377,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 2901
+    },
+    {
+      "epoch": 4.635782747603834,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 2902
+    },
+    {
+      "epoch": 4.637380191693291,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 2903
+    },
+    {
+      "epoch": 4.638977635782748,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 2904
+    },
+    {
+      "epoch": 4.640575079872205,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 2905
+    },
+    {
+      "epoch": 4.642172523961661,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 2906
+    },
+    {
+      "epoch": 4.643769968051118,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1603,
+      "step": 2907
+    },
+    {
+      "epoch": 4.645367412140575,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 2908
+    },
+    {
+      "epoch": 4.646964856230032,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1679,
+      "step": 2909
+    },
+    {
+      "epoch": 4.6485623003194885,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 2910
+    },
+    {
+      "epoch": 4.6501597444089455,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 2911
+    },
+    {
+      "epoch": 4.651757188498403,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 2912
+    },
+    {
+      "epoch": 4.65335463258786,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 2913
+    },
+    {
+      "epoch": 4.654952076677317,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2914
+    },
+    {
+      "epoch": 4.656549520766773,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2915
+    },
+    {
+      "epoch": 4.65814696485623,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2916
+    },
+    {
+      "epoch": 4.659744408945687,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 2917
+    },
+    {
+      "epoch": 4.661341853035144,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2918
+    },
+    {
+      "epoch": 4.6629392971246,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.166,
+      "step": 2919
+    },
+    {
+      "epoch": 4.664536741214057,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 2920
+    },
+    {
+      "epoch": 4.666134185303514,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 2921
+    },
+    {
+      "epoch": 4.667731629392971,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 2922
+    },
+    {
+      "epoch": 4.669329073482428,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 2923
+    },
+    {
+      "epoch": 4.6709265175718855,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1491,
+      "step": 2924
+    },
+    {
+      "epoch": 4.672523961661342,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 2925
+    },
+    {
+      "epoch": 4.674121405750799,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 2926
+    },
+    {
+      "epoch": 4.675718849840256,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 2927
+    },
+    {
+      "epoch": 4.677316293929713,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2928
+    },
+    {
+      "epoch": 4.678913738019169,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1643,
+      "step": 2929
+    },
+    {
+      "epoch": 4.680511182108626,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 2930
+    },
+    {
+      "epoch": 4.682108626198083,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 2931
+    },
+    {
+      "epoch": 4.68370607028754,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2932
+    },
+    {
+      "epoch": 4.685303514376997,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1592,
+      "step": 2933
+    },
+    {
+      "epoch": 4.686900958466453,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2934
+    },
+    {
+      "epoch": 4.68849840255591,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 2935
+    },
+    {
+      "epoch": 4.6900958466453675,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 2936
+    },
+    {
+      "epoch": 4.6916932907348246,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 2937
+    },
+    {
+      "epoch": 4.693290734824281,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 2938
+    },
+    {
+      "epoch": 4.694888178913738,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2939
+    },
+    {
+      "epoch": 4.696485623003195,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 2940
+    },
+    {
+      "epoch": 4.698083067092652,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 2941
+    },
+    {
+      "epoch": 4.699680511182109,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 2942
+    },
+    {
+      "epoch": 4.701277955271565,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 2943
+    },
+    {
+      "epoch": 4.702875399361022,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2944
+    },
+    {
+      "epoch": 4.704472843450479,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 2945
+    },
+    {
+      "epoch": 4.706070287539936,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 2946
+    },
+    {
+      "epoch": 4.707667731629393,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 2947
+    },
+    {
+      "epoch": 4.7092651757188495,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 2948
+    },
+    {
+      "epoch": 4.710862619808307,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 2949
+    },
+    {
+      "epoch": 4.712460063897764,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 2950
+    },
+    {
+      "epoch": 4.714057507987221,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 2951
+    },
+    {
+      "epoch": 4.715654952076678,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 2952
+    },
+    {
+      "epoch": 4.717252396166134,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 2953
+    },
+    {
+      "epoch": 4.718849840255591,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2954
+    },
+    {
+      "epoch": 4.720447284345048,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 2955
+    },
+    {
+      "epoch": 4.722044728434505,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 2956
+    },
+    {
+      "epoch": 4.723642172523961,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2957
+    },
+    {
+      "epoch": 4.725239616613418,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 2958
+    },
+    {
+      "epoch": 4.726837060702875,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 2959
+    },
+    {
+      "epoch": 4.728434504792332,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 2960
+    },
+    {
+      "epoch": 4.7300319488817895,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 2961
+    },
+    {
+      "epoch": 4.731629392971246,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 2962
+    },
+    {
+      "epoch": 4.733226837060703,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 2963
+    },
+    {
+      "epoch": 4.73482428115016,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 2964
+    },
+    {
+      "epoch": 4.736421725239617,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 2965
+    },
+    {
+      "epoch": 4.738019169329074,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 2966
+    },
+    {
+      "epoch": 4.73961661341853,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 2967
+    },
+    {
+      "epoch": 4.741214057507987,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 2968
+    },
+    {
+      "epoch": 4.742811501597444,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 2969
+    },
+    {
+      "epoch": 4.744408945686901,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 2970
+    },
+    {
+      "epoch": 4.746006389776358,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 2971
+    },
+    {
+      "epoch": 4.747603833865814,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 2972
+    },
+    {
+      "epoch": 4.7492012779552715,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 2973
+    },
+    {
+      "epoch": 4.7507987220447285,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 2974
+    },
+    {
+      "epoch": 4.752396166134186,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 2975
+    },
+    {
+      "epoch": 4.753993610223642,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 2976
+    },
+    {
+      "epoch": 4.755591054313099,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1603,
+      "step": 2977
+    },
+    {
+      "epoch": 4.757188498402556,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 2978
+    },
+    {
+      "epoch": 4.758785942492013,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 2979
+    },
+    {
+      "epoch": 4.76038338658147,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 2980
+    },
+    {
+      "epoch": 4.761980830670926,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 2981
+    },
+    {
+      "epoch": 4.763578274760383,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 2982
+    },
+    {
+      "epoch": 4.76517571884984,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 2983
+    },
+    {
+      "epoch": 4.766773162939297,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 2984
+    },
+    {
+      "epoch": 4.768370607028754,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 2985
+    },
+    {
+      "epoch": 4.7699680511182105,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 2986
+    },
+    {
+      "epoch": 4.771565495207668,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 2987
+    },
+    {
+      "epoch": 4.773162939297125,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 2988
+    },
+    {
+      "epoch": 4.774760383386582,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 2989
+    },
+    {
+      "epoch": 4.776357827476039,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 2990
+    },
+    {
+      "epoch": 4.777955271565495,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.162,
+      "step": 2991
+    },
+    {
+      "epoch": 4.779552715654952,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 2992
+    },
+    {
+      "epoch": 4.781150159744409,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 2993
+    },
+    {
+      "epoch": 4.782747603833866,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 2994
+    },
+    {
+      "epoch": 4.784345047923322,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 2995
+    },
+    {
+      "epoch": 4.785942492012779,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 2996
+    },
+    {
+      "epoch": 4.787539936102236,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 2997
+    },
+    {
+      "epoch": 4.789137380191693,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 2998
+    },
+    {
+      "epoch": 4.7907348242811505,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 2999
+    },
+    {
+      "epoch": 4.792332268370607,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3000
+    },
+    {
+      "epoch": 4.793929712460064,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 3001
+    },
+    {
+      "epoch": 4.795527156549521,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 3002
+    },
+    {
+      "epoch": 4.797124600638978,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 3003
+    },
+    {
+      "epoch": 4.798722044728435,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1704,
+      "step": 3004
+    },
+    {
+      "epoch": 4.800319488817891,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3005
+    },
+    {
+      "epoch": 4.801916932907348,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 3006
+    },
+    {
+      "epoch": 4.803514376996805,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 3007
+    },
+    {
+      "epoch": 4.805111821086262,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 3008
+    },
+    {
+      "epoch": 4.806709265175719,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1639,
+      "step": 3009
+    },
+    {
+      "epoch": 4.8083067092651754,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3010
+    },
+    {
+      "epoch": 4.8099041533546325,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 3011
+    },
+    {
+      "epoch": 4.81150159744409,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 3012
+    },
+    {
+      "epoch": 4.813099041533547,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 3013
+    },
+    {
+      "epoch": 4.814696485623003,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 3014
+    },
+    {
+      "epoch": 4.81629392971246,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3015
+    },
+    {
+      "epoch": 4.817891373801917,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 3016
+    },
+    {
+      "epoch": 4.819488817891374,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1592,
+      "step": 3017
+    },
+    {
+      "epoch": 4.821086261980831,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 3018
+    },
+    {
+      "epoch": 4.822683706070287,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3019
+    },
+    {
+      "epoch": 4.824281150159744,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 3020
+    },
+    {
+      "epoch": 4.825878594249201,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 3021
+    },
+    {
+      "epoch": 4.827476038338658,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 3022
+    },
+    {
+      "epoch": 4.8290734824281145,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 3023
+    },
+    {
+      "epoch": 4.830670926517572,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 3024
+    },
+    {
+      "epoch": 4.832268370607029,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 3025
+    },
+    {
+      "epoch": 4.833865814696486,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 3026
+    },
+    {
+      "epoch": 4.835463258785943,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 3027
+    },
+    {
+      "epoch": 4.8370607028754,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 3028
+    },
+    {
+      "epoch": 4.838658146964856,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 3029
+    },
+    {
+      "epoch": 4.840255591054313,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 3030
+    },
+    {
+      "epoch": 4.84185303514377,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 3031
+    },
+    {
+      "epoch": 4.843450479233227,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 3032
+    },
+    {
+      "epoch": 4.845047923322683,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 3033
+    },
+    {
+      "epoch": 4.84664536741214,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3034
+    },
+    {
+      "epoch": 4.848242811501597,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3035
+    },
+    {
+      "epoch": 4.8498402555910545,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 3036
+    },
+    {
+      "epoch": 4.8514376996805115,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3037
+    },
+    {
+      "epoch": 4.853035143769968,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 3038
+    },
+    {
+      "epoch": 4.854632587859425,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3039
+    },
+    {
+      "epoch": 4.856230031948882,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 3040
+    },
+    {
+      "epoch": 4.857827476038339,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 3041
+    },
+    {
+      "epoch": 4.859424920127795,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 3042
+    },
+    {
+      "epoch": 4.861022364217252,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 3043
+    },
+    {
+      "epoch": 4.862619808306709,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 3044
+    },
+    {
+      "epoch": 4.864217252396166,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 3045
+    },
+    {
+      "epoch": 4.865814696485623,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 3046
+    },
+    {
+      "epoch": 4.86741214057508,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 3047
+    },
+    {
+      "epoch": 4.8690095846645365,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 3048
+    },
+    {
+      "epoch": 4.8706070287539935,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 3049
+    },
+    {
+      "epoch": 4.872204472843451,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3050
+    },
+    {
+      "epoch": 4.873801916932908,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 3051
+    },
+    {
+      "epoch": 4.875399361022364,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3052
+    },
+    {
+      "epoch": 4.876996805111821,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 3053
+    },
+    {
+      "epoch": 4.878594249201278,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 3054
+    },
+    {
+      "epoch": 4.880191693290735,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3055
+    },
+    {
+      "epoch": 4.881789137380192,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1709,
+      "step": 3056
+    },
+    {
+      "epoch": 4.883386581469648,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 3057
+    },
+    {
+      "epoch": 4.884984025559105,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3058
+    },
+    {
+      "epoch": 4.886581469648562,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3059
+    },
+    {
+      "epoch": 4.888178913738019,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3060
+    },
+    {
+      "epoch": 4.8897763578274756,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3061
+    },
+    {
+      "epoch": 4.891373801916933,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 3062
+    },
+    {
+      "epoch": 4.89297124600639,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1622,
+      "step": 3063
+    },
+    {
+      "epoch": 4.894568690095847,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 3064
+    },
+    {
+      "epoch": 4.896166134185304,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 3065
+    },
+    {
+      "epoch": 4.897763578274761,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 3066
+    },
+    {
+      "epoch": 4.899361022364217,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 3067
+    },
+    {
+      "epoch": 4.900958466453674,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3068
+    },
+    {
+      "epoch": 4.902555910543131,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3069
+    },
+    {
+      "epoch": 4.904153354632588,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3070
+    },
+    {
+      "epoch": 4.905750798722044,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3071
+    },
+    {
+      "epoch": 4.907348242811501,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 3072
+    },
+    {
+      "epoch": 4.9089456869009584,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 3073
+    },
+    {
+      "epoch": 4.9105431309904155,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3074
+    },
+    {
+      "epoch": 4.912140575079873,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 3075
+    },
+    {
+      "epoch": 4.913738019169329,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 3076
+    },
+    {
+      "epoch": 4.915335463258786,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3077
+    },
+    {
+      "epoch": 4.916932907348243,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 3078
+    },
+    {
+      "epoch": 4.9185303514377,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 3079
+    },
+    {
+      "epoch": 4.920127795527156,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 3080
+    },
+    {
+      "epoch": 4.921725239616613,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 3081
+    },
+    {
+      "epoch": 4.92332268370607,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 3082
+    },
+    {
+      "epoch": 4.924920127795527,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 3083
+    },
+    {
+      "epoch": 4.926517571884984,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 3084
+    },
+    {
+      "epoch": 4.928115015974441,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3085
+    },
+    {
+      "epoch": 4.9297124600638975,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 3086
+    },
+    {
+      "epoch": 4.931309904153355,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 3087
+    },
+    {
+      "epoch": 4.932907348242812,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1595,
+      "step": 3088
+    },
+    {
+      "epoch": 4.934504792332269,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 3089
+    },
+    {
+      "epoch": 4.936102236421725,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 3090
+    },
+    {
+      "epoch": 4.937699680511182,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 3091
+    },
+    {
+      "epoch": 4.939297124600639,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 3092
+    },
+    {
+      "epoch": 4.940894568690096,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1621,
+      "step": 3093
+    },
+    {
+      "epoch": 4.942492012779553,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3094
+    },
+    {
+      "epoch": 4.944089456869009,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3095
+    },
+    {
+      "epoch": 4.945686900958466,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3096
+    },
+    {
+      "epoch": 4.947284345047923,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3097
+    },
+    {
+      "epoch": 4.94888178913738,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3098
+    },
+    {
+      "epoch": 4.950479233226837,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3099
+    },
+    {
+      "epoch": 4.952076677316294,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 3100
+    },
+    {
+      "epoch": 4.953674121405751,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 3101
+    },
+    {
+      "epoch": 4.955271565495208,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 3102
+    },
+    {
+      "epoch": 4.956869009584665,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3103
+    },
+    {
+      "epoch": 4.958466453674122,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3104
+    },
+    {
+      "epoch": 4.960063897763578,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 3105
+    },
+    {
+      "epoch": 4.961661341853035,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3106
+    },
+    {
+      "epoch": 4.963258785942492,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 3107
+    },
+    {
+      "epoch": 4.964856230031949,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 3108
+    },
+    {
+      "epoch": 4.966453674121405,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 3109
+    },
+    {
+      "epoch": 4.968051118210862,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 3110
+    },
+    {
+      "epoch": 4.9696485623003195,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 3111
+    },
+    {
+      "epoch": 4.9712460063897765,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 3112
+    },
+    {
+      "epoch": 4.972843450479234,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 3113
+    },
+    {
+      "epoch": 4.97444089456869,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 3114
+    },
+    {
+      "epoch": 4.976038338658147,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 3115
+    },
+    {
+      "epoch": 4.977635782747604,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 3116
+    },
+    {
+      "epoch": 4.979233226837061,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3117
+    },
+    {
+      "epoch": 4.980830670926517,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 3118
+    },
+    {
+      "epoch": 4.982428115015974,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3119
+    },
+    {
+      "epoch": 4.984025559105431,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 3120
+    },
+    {
+      "epoch": 4.985623003194888,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3121
+    },
+    {
+      "epoch": 4.987220447284345,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3122
+    },
+    {
+      "epoch": 4.988817891373802,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3123
+    },
+    {
+      "epoch": 4.9904153354632586,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1563,
+      "step": 3124
+    },
+    {
+      "epoch": 4.992012779552716,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3125
+    },
+    {
+      "epoch": 4.993610223642173,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 3126
+    },
+    {
+      "epoch": 4.99520766773163,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 3127
+    },
+    {
+      "epoch": 4.996805111821086,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3128
+    },
+    {
+      "epoch": 4.998402555910543,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3129
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3130
+    },
+    {
+      "epoch": 5.001597444089457,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 3131
+    },
+    {
+      "epoch": 5.003194888178914,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 3132
+    },
+    {
+      "epoch": 5.00479233226837,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3133
+    },
+    {
+      "epoch": 5.006389776357827,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 3134
+    },
+    {
+      "epoch": 5.007987220447284,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 3135
+    },
+    {
+      "epoch": 5.0095846645367414,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 3136
+    },
+    {
+      "epoch": 5.0111821086261985,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 3137
+    },
+    {
+      "epoch": 5.012779552715655,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3138
+    },
+    {
+      "epoch": 5.014376996805112,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 3139
+    },
+    {
+      "epoch": 5.015974440894569,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3140
+    },
+    {
+      "epoch": 5.017571884984026,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3141
+    },
+    {
+      "epoch": 5.019169329073482,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 3142
+    },
+    {
+      "epoch": 5.020766773162939,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 3143
+    },
+    {
+      "epoch": 5.022364217252396,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 3144
+    },
+    {
+      "epoch": 5.023961661341853,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.166,
+      "step": 3145
+    },
+    {
+      "epoch": 5.02555910543131,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 3146
+    },
+    {
+      "epoch": 5.027156549520766,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 3147
+    },
+    {
+      "epoch": 5.0287539936102235,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3148
+    },
+    {
+      "epoch": 5.0303514376996805,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3149
+    },
+    {
+      "epoch": 5.031948881789138,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 3150
+    },
+    {
+      "epoch": 5.033546325878595,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 3151
+    },
+    {
+      "epoch": 5.035143769968051,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3152
+    },
+    {
+      "epoch": 5.036741214057508,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 3153
+    },
+    {
+      "epoch": 5.038338658146965,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3154
+    },
+    {
+      "epoch": 5.039936102236422,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 3155
+    },
+    {
+      "epoch": 5.041533546325879,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3156
+    },
+    {
+      "epoch": 5.043130990415335,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 3157
+    },
+    {
+      "epoch": 5.044728434504792,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 3158
+    },
+    {
+      "epoch": 5.046325878594249,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3159
+    },
+    {
+      "epoch": 5.047923322683706,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3160
+    },
+    {
+      "epoch": 5.0495207667731625,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3161
+    },
+    {
+      "epoch": 5.05111821086262,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 3162
+    },
+    {
+      "epoch": 5.052715654952077,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 3163
+    },
+    {
+      "epoch": 5.054313099041534,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3164
+    },
+    {
+      "epoch": 5.055910543130991,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3165
+    },
+    {
+      "epoch": 5.057507987220447,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 3166
+    },
+    {
+      "epoch": 5.059105431309904,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 3167
+    },
+    {
+      "epoch": 5.060702875399361,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 3168
+    },
+    {
+      "epoch": 5.062300319488818,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3169
+    },
+    {
+      "epoch": 5.063897763578275,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 3170
+    },
+    {
+      "epoch": 5.065495207667731,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 3171
+    },
+    {
+      "epoch": 5.067092651757188,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3172
+    },
+    {
+      "epoch": 5.068690095846645,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3173
+    },
+    {
+      "epoch": 5.0702875399361025,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 3174
+    },
+    {
+      "epoch": 5.0718849840255595,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3175
+    },
+    {
+      "epoch": 5.073482428115016,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 3176
+    },
+    {
+      "epoch": 5.075079872204473,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 3177
+    },
+    {
+      "epoch": 5.07667731629393,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3178
+    },
+    {
+      "epoch": 5.078274760383387,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 3179
+    },
+    {
+      "epoch": 5.079872204472843,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3180
+    },
+    {
+      "epoch": 5.0814696485623,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 3181
+    },
+    {
+      "epoch": 5.083067092651757,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 3182
+    },
+    {
+      "epoch": 5.084664536741214,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3183
+    },
+    {
+      "epoch": 5.086261980830671,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 3184
+    },
+    {
+      "epoch": 5.087859424920127,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 3185
+    },
+    {
+      "epoch": 5.0894568690095845,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 3186
+    },
+    {
+      "epoch": 5.0910543130990416,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 3187
+    },
+    {
+      "epoch": 5.092651757188499,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 3188
+    },
+    {
+      "epoch": 5.094249201277956,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1727,
+      "step": 3189
+    },
+    {
+      "epoch": 5.095846645367412,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 3190
+    },
+    {
+      "epoch": 5.097444089456869,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 3191
+    },
+    {
+      "epoch": 5.099041533546326,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 3192
+    },
+    {
+      "epoch": 5.100638977635783,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 3193
+    },
+    {
+      "epoch": 5.102236421725239,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 3194
+    },
+    {
+      "epoch": 5.103833865814696,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3195
+    },
+    {
+      "epoch": 5.105431309904153,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 3196
+    },
+    {
+      "epoch": 5.10702875399361,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3197
+    },
+    {
+      "epoch": 5.108626198083067,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 3198
+    },
+    {
+      "epoch": 5.110223642172524,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 3199
+    },
+    {
+      "epoch": 5.111821086261981,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 3200
+    },
+    {
+      "epoch": 5.113418530351438,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3201
+    },
+    {
+      "epoch": 5.115015974440895,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 3202
+    },
+    {
+      "epoch": 5.116613418530352,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 3203
+    },
+    {
+      "epoch": 5.118210862619808,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 3204
+    },
+    {
+      "epoch": 5.119808306709265,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 3205
+    },
+    {
+      "epoch": 5.121405750798722,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 3206
+    },
+    {
+      "epoch": 5.123003194888179,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3207
+    },
+    {
+      "epoch": 5.124600638977636,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1491,
+      "step": 3208
+    },
+    {
+      "epoch": 5.126198083067092,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3209
+    },
+    {
+      "epoch": 5.127795527156549,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3210
+    },
+    {
+      "epoch": 5.1293929712460065,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 3211
+    },
+    {
+      "epoch": 5.1309904153354635,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 3212
+    },
+    {
+      "epoch": 5.13258785942492,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1623,
+      "step": 3213
+    },
+    {
+      "epoch": 5.134185303514377,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 3214
+    },
+    {
+      "epoch": 5.135782747603834,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3215
+    },
+    {
+      "epoch": 5.137380191693291,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3216
+    },
+    {
+      "epoch": 5.138977635782748,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 3217
+    },
+    {
+      "epoch": 5.140575079872204,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 3218
+    },
+    {
+      "epoch": 5.142172523961661,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 3219
+    },
+    {
+      "epoch": 5.143769968051118,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3220
+    },
+    {
+      "epoch": 5.145367412140575,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 3221
+    },
+    {
+      "epoch": 5.146964856230032,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 3222
+    },
+    {
+      "epoch": 5.1485623003194885,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 3223
+    },
+    {
+      "epoch": 5.1501597444089455,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 3224
+    },
+    {
+      "epoch": 5.151757188498403,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 3225
+    },
+    {
+      "epoch": 5.15335463258786,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 3226
+    },
+    {
+      "epoch": 5.154952076677317,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3227
+    },
+    {
+      "epoch": 5.156549520766773,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 3228
+    },
+    {
+      "epoch": 5.15814696485623,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 3229
+    },
+    {
+      "epoch": 5.159744408945687,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3230
+    },
+    {
+      "epoch": 5.161341853035144,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3231
+    },
+    {
+      "epoch": 5.1629392971246,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1631,
+      "step": 3232
+    },
+    {
+      "epoch": 5.164536741214057,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 3233
+    },
+    {
+      "epoch": 5.166134185303514,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 3234
+    },
+    {
+      "epoch": 5.167731629392971,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3235
+    },
+    {
+      "epoch": 5.169329073482428,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3236
+    },
+    {
+      "epoch": 5.170926517571885,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3237
+    },
+    {
+      "epoch": 5.172523961661342,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3238
+    },
+    {
+      "epoch": 5.174121405750799,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 3239
+    },
+    {
+      "epoch": 5.175718849840256,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 3240
+    },
+    {
+      "epoch": 5.177316293929713,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 3241
+    },
+    {
+      "epoch": 5.178913738019169,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1699,
+      "step": 3242
+    },
+    {
+      "epoch": 5.180511182108626,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 3243
+    },
+    {
+      "epoch": 5.182108626198083,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3244
+    },
+    {
+      "epoch": 5.18370607028754,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1652,
+      "step": 3245
+    },
+    {
+      "epoch": 5.185303514376997,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3246
+    },
+    {
+      "epoch": 5.186900958466453,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 3247
+    },
+    {
+      "epoch": 5.18849840255591,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1649,
+      "step": 3248
+    },
+    {
+      "epoch": 5.1900958466453675,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3249
+    },
+    {
+      "epoch": 5.1916932907348246,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3250
+    },
+    {
+      "epoch": 5.193290734824281,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 3251
+    },
+    {
+      "epoch": 5.194888178913738,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 3252
+    },
+    {
+      "epoch": 5.196485623003195,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 3253
+    },
+    {
+      "epoch": 5.198083067092652,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 3254
+    },
+    {
+      "epoch": 5.199680511182109,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3255
+    },
+    {
+      "epoch": 5.201277955271565,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3256
+    },
+    {
+      "epoch": 5.202875399361022,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3257
+    },
+    {
+      "epoch": 5.204472843450479,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 3258
+    },
+    {
+      "epoch": 5.206070287539936,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 3259
+    },
+    {
+      "epoch": 5.207667731629393,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 3260
+    },
+    {
+      "epoch": 5.2092651757188495,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 3261
+    },
+    {
+      "epoch": 5.210862619808307,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.158,
+      "step": 3262
+    },
+    {
+      "epoch": 5.212460063897764,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 3263
+    },
+    {
+      "epoch": 5.214057507987221,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3264
+    },
+    {
+      "epoch": 5.215654952076678,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.157,
+      "step": 3265
+    },
+    {
+      "epoch": 5.217252396166134,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 3266
+    },
+    {
+      "epoch": 5.218849840255591,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 3267
+    },
+    {
+      "epoch": 5.220447284345048,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3268
+    },
+    {
+      "epoch": 5.222044728434505,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3269
+    },
+    {
+      "epoch": 5.223642172523961,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 3270
+    },
+    {
+      "epoch": 5.225239616613418,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1615,
+      "step": 3271
+    },
+    {
+      "epoch": 5.226837060702875,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 3272
+    },
+    {
+      "epoch": 5.228434504792332,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3273
+    },
+    {
+      "epoch": 5.2300319488817895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 3274
+    },
+    {
+      "epoch": 5.231629392971246,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3275
+    },
+    {
+      "epoch": 5.233226837060703,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.163,
+      "step": 3276
+    },
+    {
+      "epoch": 5.23482428115016,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 3277
+    },
+    {
+      "epoch": 5.236421725239617,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 3278
+    },
+    {
+      "epoch": 5.238019169329074,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3279
+    },
+    {
+      "epoch": 5.23961661341853,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3280
+    },
+    {
+      "epoch": 5.241214057507987,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3281
+    },
+    {
+      "epoch": 5.242811501597444,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 3282
+    },
+    {
+      "epoch": 5.244408945686901,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1627,
+      "step": 3283
+    },
+    {
+      "epoch": 5.246006389776358,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3284
+    },
+    {
+      "epoch": 5.247603833865814,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3285
+    },
+    {
+      "epoch": 5.2492012779552715,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3286
+    },
+    {
+      "epoch": 5.2507987220447285,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3287
+    },
+    {
+      "epoch": 5.252396166134186,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 3288
+    },
+    {
+      "epoch": 5.253993610223642,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3289
+    },
+    {
+      "epoch": 5.255591054313099,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 3290
+    },
+    {
+      "epoch": 5.257188498402556,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3291
+    },
+    {
+      "epoch": 5.258785942492013,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 3292
+    },
+    {
+      "epoch": 5.26038338658147,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 3293
+    },
+    {
+      "epoch": 5.261980830670926,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 3294
+    },
+    {
+      "epoch": 5.263578274760383,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3295
+    },
+    {
+      "epoch": 5.26517571884984,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 3296
+    },
+    {
+      "epoch": 5.266773162939297,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1598,
+      "step": 3297
+    },
+    {
+      "epoch": 5.268370607028754,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 3298
+    },
+    {
+      "epoch": 5.2699680511182105,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3299
+    },
+    {
+      "epoch": 5.271565495207668,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 3300
+    },
+    {
+      "epoch": 5.273162939297125,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 3301
+    },
+    {
+      "epoch": 5.274760383386582,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 3302
+    },
+    {
+      "epoch": 5.276357827476039,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 3303
+    },
+    {
+      "epoch": 5.277955271565495,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 3304
+    },
+    {
+      "epoch": 5.279552715654952,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1636,
+      "step": 3305
+    },
+    {
+      "epoch": 5.281150159744409,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 3306
+    },
+    {
+      "epoch": 5.282747603833866,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 3307
+    },
+    {
+      "epoch": 5.284345047923322,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3308
+    },
+    {
+      "epoch": 5.285942492012779,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3309
+    },
+    {
+      "epoch": 5.287539936102236,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 3310
+    },
+    {
+      "epoch": 5.289137380191693,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1592,
+      "step": 3311
+    },
+    {
+      "epoch": 5.2907348242811505,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3312
+    },
+    {
+      "epoch": 5.292332268370607,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 3313
+    },
+    {
+      "epoch": 5.293929712460064,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 3314
+    },
+    {
+      "epoch": 5.295527156549521,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 3315
+    },
+    {
+      "epoch": 5.297124600638978,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 3316
+    },
+    {
+      "epoch": 5.298722044728435,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3317
+    },
+    {
+      "epoch": 5.300319488817891,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3318
+    },
+    {
+      "epoch": 5.301916932907348,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3319
+    },
+    {
+      "epoch": 5.303514376996805,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 3320
+    },
+    {
+      "epoch": 5.305111821086262,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 3321
+    },
+    {
+      "epoch": 5.306709265175719,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 3322
+    },
+    {
+      "epoch": 5.3083067092651754,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3323
+    },
+    {
+      "epoch": 5.3099041533546325,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3324
+    },
+    {
+      "epoch": 5.31150159744409,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3325
+    },
+    {
+      "epoch": 5.313099041533547,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3326
+    },
+    {
+      "epoch": 5.314696485623003,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3327
+    },
+    {
+      "epoch": 5.31629392971246,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3328
+    },
+    {
+      "epoch": 5.317891373801917,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 3329
+    },
+    {
+      "epoch": 5.319488817891374,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1539,
+      "step": 3330
+    },
+    {
+      "epoch": 5.321086261980831,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3331
+    },
+    {
+      "epoch": 5.322683706070287,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 3332
+    },
+    {
+      "epoch": 5.324281150159744,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3333
+    },
+    {
+      "epoch": 5.325878594249201,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3334
+    },
+    {
+      "epoch": 5.327476038338658,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3335
+    },
+    {
+      "epoch": 5.329073482428115,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3336
+    },
+    {
+      "epoch": 5.330670926517572,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3337
+    },
+    {
+      "epoch": 5.332268370607029,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 3338
+    },
+    {
+      "epoch": 5.333865814696486,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3339
+    },
+    {
+      "epoch": 5.335463258785943,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3340
+    },
+    {
+      "epoch": 5.3370607028754,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 3341
+    },
+    {
+      "epoch": 5.338658146964856,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3342
+    },
+    {
+      "epoch": 5.340255591054313,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 3343
+    },
+    {
+      "epoch": 5.34185303514377,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 3344
+    },
+    {
+      "epoch": 5.343450479233227,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 3345
+    },
+    {
+      "epoch": 5.345047923322683,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 3346
+    },
+    {
+      "epoch": 5.34664536741214,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3347
+    },
+    {
+      "epoch": 5.348242811501597,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3348
+    },
+    {
+      "epoch": 5.3498402555910545,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3349
+    },
+    {
+      "epoch": 5.3514376996805115,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 3350
+    },
+    {
+      "epoch": 5.353035143769968,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 3351
+    },
+    {
+      "epoch": 5.354632587859425,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 3352
+    },
+    {
+      "epoch": 5.356230031948882,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 3353
+    },
+    {
+      "epoch": 5.357827476038339,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 3354
+    },
+    {
+      "epoch": 5.359424920127796,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3355
+    },
+    {
+      "epoch": 5.361022364217252,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 3356
+    },
+    {
+      "epoch": 5.362619808306709,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 3357
+    },
+    {
+      "epoch": 5.364217252396166,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3358
+    },
+    {
+      "epoch": 5.365814696485623,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3359
+    },
+    {
+      "epoch": 5.36741214057508,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 3360
+    },
+    {
+      "epoch": 5.3690095846645365,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 3361
+    },
+    {
+      "epoch": 5.3706070287539935,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 3362
+    },
+    {
+      "epoch": 5.372204472843451,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 3363
+    },
+    {
+      "epoch": 5.373801916932908,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 3364
+    },
+    {
+      "epoch": 5.375399361022364,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 3365
+    },
+    {
+      "epoch": 5.376996805111821,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1589,
+      "step": 3366
+    },
+    {
+      "epoch": 5.378594249201278,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3367
+    },
+    {
+      "epoch": 5.380191693290735,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 3368
+    },
+    {
+      "epoch": 5.381789137380192,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 3369
+    },
+    {
+      "epoch": 5.383386581469648,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3370
+    },
+    {
+      "epoch": 5.384984025559105,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3371
+    },
+    {
+      "epoch": 5.386581469648562,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3372
+    },
+    {
+      "epoch": 5.388178913738019,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 3373
+    },
+    {
+      "epoch": 5.389776357827476,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3374
+    },
+    {
+      "epoch": 5.391373801916933,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 3375
+    },
+    {
+      "epoch": 5.39297124600639,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1638,
+      "step": 3376
+    },
+    {
+      "epoch": 5.394568690095847,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 3377
+    },
+    {
+      "epoch": 5.396166134185304,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3378
+    },
+    {
+      "epoch": 5.397763578274761,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 3379
+    },
+    {
+      "epoch": 5.399361022364217,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 3380
+    },
+    {
+      "epoch": 5.400958466453674,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 3381
+    },
+    {
+      "epoch": 5.402555910543131,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1595,
+      "step": 3382
+    },
+    {
+      "epoch": 5.404153354632588,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 3383
+    },
+    {
+      "epoch": 5.405750798722044,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3384
+    },
+    {
+      "epoch": 5.407348242811501,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3385
+    },
+    {
+      "epoch": 5.4089456869009584,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3386
+    },
+    {
+      "epoch": 5.4105431309904155,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 3387
+    },
+    {
+      "epoch": 5.412140575079873,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 3388
+    },
+    {
+      "epoch": 5.413738019169329,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1607,
+      "step": 3389
+    },
+    {
+      "epoch": 5.415335463258786,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3390
+    },
+    {
+      "epoch": 5.416932907348243,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 3391
+    },
+    {
+      "epoch": 5.4185303514377,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 3392
+    },
+    {
+      "epoch": 5.420127795527157,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 3393
+    },
+    {
+      "epoch": 5.421725239616613,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 3394
+    },
+    {
+      "epoch": 5.42332268370607,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3395
+    },
+    {
+      "epoch": 5.424920127795527,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 3396
+    },
+    {
+      "epoch": 5.426517571884984,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 3397
+    },
+    {
+      "epoch": 5.428115015974441,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 3398
+    },
+    {
+      "epoch": 5.4297124600638975,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 3399
+    },
+    {
+      "epoch": 5.431309904153355,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3400
+    },
+    {
+      "epoch": 5.432907348242812,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3401
+    },
+    {
+      "epoch": 5.434504792332269,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3402
+    },
+    {
+      "epoch": 5.436102236421725,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 3403
+    },
+    {
+      "epoch": 5.437699680511182,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3404
+    },
+    {
+      "epoch": 5.439297124600639,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 3405
+    },
+    {
+      "epoch": 5.440894568690096,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 3406
+    },
+    {
+      "epoch": 5.442492012779553,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3407
+    },
+    {
+      "epoch": 5.444089456869009,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3408
+    },
+    {
+      "epoch": 5.445686900958466,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 3409
+    },
+    {
+      "epoch": 5.447284345047923,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3410
+    },
+    {
+      "epoch": 5.44888178913738,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3411
+    },
+    {
+      "epoch": 5.4504792332268375,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 3412
+    },
+    {
+      "epoch": 5.452076677316294,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3413
+    },
+    {
+      "epoch": 5.453674121405751,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 3414
+    },
+    {
+      "epoch": 5.455271565495208,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3415
+    },
+    {
+      "epoch": 5.456869009584665,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 3416
+    },
+    {
+      "epoch": 5.458466453674122,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3417
+    },
+    {
+      "epoch": 5.460063897763578,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3418
+    },
+    {
+      "epoch": 5.461661341853035,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 3419
+    },
+    {
+      "epoch": 5.463258785942492,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3420
+    },
+    {
+      "epoch": 5.464856230031949,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 3421
+    },
+    {
+      "epoch": 5.466453674121405,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 3422
+    },
+    {
+      "epoch": 5.468051118210862,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 3423
+    },
+    {
+      "epoch": 5.4696485623003195,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 3424
+    },
+    {
+      "epoch": 5.4712460063897765,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 3425
+    },
+    {
+      "epoch": 5.472843450479234,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.16,
+      "step": 3426
+    },
+    {
+      "epoch": 5.47444089456869,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 3427
+    },
+    {
+      "epoch": 5.476038338658147,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 3428
+    },
+    {
+      "epoch": 5.477635782747604,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3429
+    },
+    {
+      "epoch": 5.479233226837061,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 3430
+    },
+    {
+      "epoch": 5.480830670926517,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3431
+    },
+    {
+      "epoch": 5.482428115015974,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 3432
+    },
+    {
+      "epoch": 5.484025559105431,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3433
+    },
+    {
+      "epoch": 5.485623003194888,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 3434
+    },
+    {
+      "epoch": 5.487220447284345,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 3435
+    },
+    {
+      "epoch": 5.488817891373802,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 3436
+    },
+    {
+      "epoch": 5.4904153354632586,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3437
+    },
+    {
+      "epoch": 5.492012779552716,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3438
+    },
+    {
+      "epoch": 5.493610223642173,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 3439
+    },
+    {
+      "epoch": 5.49520766773163,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3440
+    },
+    {
+      "epoch": 5.496805111821086,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1613,
+      "step": 3441
+    },
+    {
+      "epoch": 5.498402555910543,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 3442
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 3443
+    },
+    {
+      "epoch": 5.501597444089457,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 3444
+    },
+    {
+      "epoch": 5.503194888178914,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3445
+    },
+    {
+      "epoch": 5.50479233226837,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1625,
+      "step": 3446
+    },
+    {
+      "epoch": 5.506389776357827,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 3447
+    },
+    {
+      "epoch": 5.507987220447284,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 3448
+    },
+    {
+      "epoch": 5.5095846645367414,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3449
+    },
+    {
+      "epoch": 5.511182108626198,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 3450
+    },
+    {
+      "epoch": 5.512779552715655,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 3451
+    },
+    {
+      "epoch": 5.514376996805112,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3452
+    },
+    {
+      "epoch": 5.515974440894569,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3453
+    },
+    {
+      "epoch": 5.517571884984026,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3454
+    },
+    {
+      "epoch": 5.519169329073483,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 3455
+    },
+    {
+      "epoch": 5.520766773162939,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1717,
+      "step": 3456
+    },
+    {
+      "epoch": 5.522364217252396,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 3457
+    },
+    {
+      "epoch": 5.523961661341853,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 3458
+    },
+    {
+      "epoch": 5.52555910543131,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3459
+    },
+    {
+      "epoch": 5.527156549520766,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 3460
+    },
+    {
+      "epoch": 5.5287539936102235,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3461
+    },
+    {
+      "epoch": 5.5303514376996805,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3462
+    },
+    {
+      "epoch": 5.531948881789138,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3463
+    },
+    {
+      "epoch": 5.533546325878595,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3464
+    },
+    {
+      "epoch": 5.535143769968051,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 3465
+    },
+    {
+      "epoch": 5.536741214057508,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.156,
+      "step": 3466
+    },
+    {
+      "epoch": 5.538338658146965,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 3467
+    },
+    {
+      "epoch": 5.539936102236422,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3468
+    },
+    {
+      "epoch": 5.541533546325878,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3469
+    },
+    {
+      "epoch": 5.543130990415335,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3470
+    },
+    {
+      "epoch": 5.544728434504792,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 3471
+    },
+    {
+      "epoch": 5.546325878594249,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 3472
+    },
+    {
+      "epoch": 5.547923322683706,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 3473
+    },
+    {
+      "epoch": 5.549520766773163,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 3474
+    },
+    {
+      "epoch": 5.55111821086262,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 3475
+    },
+    {
+      "epoch": 5.552715654952077,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3476
+    },
+    {
+      "epoch": 5.554313099041534,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 3477
+    },
+    {
+      "epoch": 5.555910543130991,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3478
+    },
+    {
+      "epoch": 5.557507987220447,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 3479
+    },
+    {
+      "epoch": 5.559105431309904,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 3480
+    },
+    {
+      "epoch": 5.560702875399361,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 3481
+    },
+    {
+      "epoch": 5.562300319488818,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1576,
+      "step": 3482
+    },
+    {
+      "epoch": 5.563897763578275,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 3483
+    },
+    {
+      "epoch": 5.565495207667731,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3484
+    },
+    {
+      "epoch": 5.567092651757188,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 3485
+    },
+    {
+      "epoch": 5.568690095846645,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 3486
+    },
+    {
+      "epoch": 5.5702875399361025,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3487
+    },
+    {
+      "epoch": 5.571884984025559,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 3488
+    },
+    {
+      "epoch": 5.573482428115016,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3489
+    },
+    {
+      "epoch": 5.575079872204473,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 3490
+    },
+    {
+      "epoch": 5.57667731629393,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 3491
+    },
+    {
+      "epoch": 5.578274760383387,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 3492
+    },
+    {
+      "epoch": 5.579872204472844,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 3493
+    },
+    {
+      "epoch": 5.5814696485623,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1682,
+      "step": 3494
+    },
+    {
+      "epoch": 5.583067092651757,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 3495
+    },
+    {
+      "epoch": 5.584664536741214,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 3496
+    },
+    {
+      "epoch": 5.586261980830671,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3497
+    },
+    {
+      "epoch": 5.587859424920127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 3498
+    },
+    {
+      "epoch": 5.5894568690095845,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 3499
+    },
+    {
+      "epoch": 5.5910543130990416,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3500
+    },
+    {
+      "epoch": 5.592651757188499,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 3501
+    },
+    {
+      "epoch": 5.594249201277956,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3502
+    },
+    {
+      "epoch": 5.595846645367412,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 3503
+    },
+    {
+      "epoch": 5.597444089456869,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3504
+    },
+    {
+      "epoch": 5.599041533546326,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 3505
+    },
+    {
+      "epoch": 5.600638977635783,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 3506
+    },
+    {
+      "epoch": 5.602236421725239,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3507
+    },
+    {
+      "epoch": 5.603833865814696,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 3508
+    },
+    {
+      "epoch": 5.605431309904153,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 3509
+    },
+    {
+      "epoch": 5.60702875399361,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 3510
+    },
+    {
+      "epoch": 5.608626198083067,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 3511
+    },
+    {
+      "epoch": 5.6102236421725244,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3512
+    },
+    {
+      "epoch": 5.611821086261981,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3513
+    },
+    {
+      "epoch": 5.613418530351438,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 3514
+    },
+    {
+      "epoch": 5.615015974440895,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3515
+    },
+    {
+      "epoch": 5.616613418530352,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 3516
+    },
+    {
+      "epoch": 5.618210862619808,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 3517
+    },
+    {
+      "epoch": 5.619808306709265,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 3518
+    },
+    {
+      "epoch": 5.621405750798722,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1602,
+      "step": 3519
+    },
+    {
+      "epoch": 5.623003194888179,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3520
+    },
+    {
+      "epoch": 5.624600638977636,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 3521
+    },
+    {
+      "epoch": 5.626198083067092,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 3522
+    },
+    {
+      "epoch": 5.627795527156549,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 3523
+    },
+    {
+      "epoch": 5.6293929712460065,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3524
+    },
+    {
+      "epoch": 5.6309904153354635,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3525
+    },
+    {
+      "epoch": 5.63258785942492,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 3526
+    },
+    {
+      "epoch": 5.634185303514377,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 3527
+    },
+    {
+      "epoch": 5.635782747603834,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 3528
+    },
+    {
+      "epoch": 5.637380191693291,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3529
+    },
+    {
+      "epoch": 5.638977635782748,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 3530
+    },
+    {
+      "epoch": 5.640575079872205,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 3531
+    },
+    {
+      "epoch": 5.642172523961661,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3532
+    },
+    {
+      "epoch": 5.643769968051118,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 3533
+    },
+    {
+      "epoch": 5.645367412140575,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3534
+    },
+    {
+      "epoch": 5.646964856230032,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 3535
+    },
+    {
+      "epoch": 5.6485623003194885,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 3536
+    },
+    {
+      "epoch": 5.6501597444089455,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3537
+    },
+    {
+      "epoch": 5.651757188498403,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 3538
+    },
+    {
+      "epoch": 5.65335463258786,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 3539
+    },
+    {
+      "epoch": 5.654952076677317,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 3540
+    },
+    {
+      "epoch": 5.656549520766773,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 3541
+    },
+    {
+      "epoch": 5.65814696485623,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3542
+    },
+    {
+      "epoch": 5.659744408945687,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3543
+    },
+    {
+      "epoch": 5.661341853035144,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3544
+    },
+    {
+      "epoch": 5.6629392971246,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 3545
+    },
+    {
+      "epoch": 5.664536741214057,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 3546
+    },
+    {
+      "epoch": 5.666134185303514,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 3547
+    },
+    {
+      "epoch": 5.667731629392971,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3548
+    },
+    {
+      "epoch": 5.669329073482428,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 3549
+    },
+    {
+      "epoch": 5.6709265175718855,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 3550
+    },
+    {
+      "epoch": 5.672523961661342,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3551
+    },
+    {
+      "epoch": 5.674121405750799,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 3552
+    },
+    {
+      "epoch": 5.675718849840256,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 3553
+    },
+    {
+      "epoch": 5.677316293929713,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 3554
+    },
+    {
+      "epoch": 5.678913738019169,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 3555
+    },
+    {
+      "epoch": 5.680511182108626,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 3556
+    },
+    {
+      "epoch": 5.682108626198083,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 3557
+    },
+    {
+      "epoch": 5.68370607028754,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 3558
+    },
+    {
+      "epoch": 5.685303514376997,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3559
+    },
+    {
+      "epoch": 5.686900958466453,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3560
+    },
+    {
+      "epoch": 5.68849840255591,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 3561
+    },
+    {
+      "epoch": 5.6900958466453675,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3562
+    },
+    {
+      "epoch": 5.6916932907348246,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 3563
+    },
+    {
+      "epoch": 5.693290734824281,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 3564
+    },
+    {
+      "epoch": 5.694888178913738,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3565
+    },
+    {
+      "epoch": 5.696485623003195,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3566
+    },
+    {
+      "epoch": 5.698083067092652,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 3567
+    },
+    {
+      "epoch": 5.699680511182109,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 3568
+    },
+    {
+      "epoch": 5.701277955271565,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 3569
+    },
+    {
+      "epoch": 5.702875399361022,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3570
+    },
+    {
+      "epoch": 5.704472843450479,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 3571
+    },
+    {
+      "epoch": 5.706070287539936,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 3572
+    },
+    {
+      "epoch": 5.707667731629393,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 3573
+    },
+    {
+      "epoch": 5.7092651757188495,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3574
+    },
+    {
+      "epoch": 5.710862619808307,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 3575
+    },
+    {
+      "epoch": 5.712460063897764,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3576
+    },
+    {
+      "epoch": 5.714057507987221,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 3577
+    },
+    {
+      "epoch": 5.715654952076678,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 3578
+    },
+    {
+      "epoch": 5.717252396166134,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3579
+    },
+    {
+      "epoch": 5.718849840255591,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3580
+    },
+    {
+      "epoch": 5.720447284345048,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 3581
+    },
+    {
+      "epoch": 5.722044728434505,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3582
+    },
+    {
+      "epoch": 5.723642172523961,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 3583
+    },
+    {
+      "epoch": 5.725239616613418,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 3584
+    },
+    {
+      "epoch": 5.726837060702875,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 3585
+    },
+    {
+      "epoch": 5.728434504792332,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 3586
+    },
+    {
+      "epoch": 5.7300319488817895,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 3587
+    },
+    {
+      "epoch": 5.731629392971246,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3588
+    },
+    {
+      "epoch": 5.733226837060703,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 3589
+    },
+    {
+      "epoch": 5.73482428115016,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 3590
+    },
+    {
+      "epoch": 5.736421725239617,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3591
+    },
+    {
+      "epoch": 5.738019169329074,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 3592
+    },
+    {
+      "epoch": 5.73961661341853,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3593
+    },
+    {
+      "epoch": 5.741214057507987,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 3594
+    },
+    {
+      "epoch": 5.742811501597444,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1619,
+      "step": 3595
+    },
+    {
+      "epoch": 5.744408945686901,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3596
+    },
+    {
+      "epoch": 5.746006389776358,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 3597
+    },
+    {
+      "epoch": 5.747603833865814,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 3598
+    },
+    {
+      "epoch": 5.7492012779552715,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 3599
+    },
+    {
+      "epoch": 5.7507987220447285,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 3600
+    },
+    {
+      "epoch": 5.752396166134186,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3601
+    },
+    {
+      "epoch": 5.753993610223642,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3602
+    },
+    {
+      "epoch": 5.755591054313099,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3603
+    },
+    {
+      "epoch": 5.757188498402556,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 3604
+    },
+    {
+      "epoch": 5.758785942492013,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 3605
+    },
+    {
+      "epoch": 5.76038338658147,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3606
+    },
+    {
+      "epoch": 5.761980830670926,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 3607
+    },
+    {
+      "epoch": 5.763578274760383,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1539,
+      "step": 3608
+    },
+    {
+      "epoch": 5.76517571884984,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 3609
+    },
+    {
+      "epoch": 5.766773162939297,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1539,
+      "step": 3610
+    },
+    {
+      "epoch": 5.768370607028754,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 3611
+    },
+    {
+      "epoch": 5.7699680511182105,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3612
+    },
+    {
+      "epoch": 5.771565495207668,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 3613
+    },
+    {
+      "epoch": 5.773162939297125,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 3614
+    },
+    {
+      "epoch": 5.774760383386582,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 3615
+    },
+    {
+      "epoch": 5.776357827476039,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 3616
+    },
+    {
+      "epoch": 5.777955271565495,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.156,
+      "step": 3617
+    },
+    {
+      "epoch": 5.779552715654952,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 3618
+    },
+    {
+      "epoch": 5.781150159744409,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3619
+    },
+    {
+      "epoch": 5.782747603833866,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3620
+    },
+    {
+      "epoch": 5.784345047923322,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 3621
+    },
+    {
+      "epoch": 5.785942492012779,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 3622
+    },
+    {
+      "epoch": 5.787539936102236,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3623
+    },
+    {
+      "epoch": 5.789137380191693,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 3624
+    },
+    {
+      "epoch": 5.7907348242811505,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 3625
+    },
+    {
+      "epoch": 5.792332268370607,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 3626
+    },
+    {
+      "epoch": 5.793929712460064,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 3627
+    },
+    {
+      "epoch": 5.795527156549521,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3628
+    },
+    {
+      "epoch": 5.797124600638978,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 3629
+    },
+    {
+      "epoch": 5.798722044728435,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 3630
+    },
+    {
+      "epoch": 5.800319488817891,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3631
+    },
+    {
+      "epoch": 5.801916932907348,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3632
+    },
+    {
+      "epoch": 5.803514376996805,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 3633
+    },
+    {
+      "epoch": 5.805111821086262,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 3634
+    },
+    {
+      "epoch": 5.806709265175719,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3635
+    },
+    {
+      "epoch": 5.8083067092651754,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3636
+    },
+    {
+      "epoch": 5.8099041533546325,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 3637
+    },
+    {
+      "epoch": 5.81150159744409,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3638
+    },
+    {
+      "epoch": 5.813099041533547,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 3639
+    },
+    {
+      "epoch": 5.814696485623003,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 3640
+    },
+    {
+      "epoch": 5.81629392971246,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 3641
+    },
+    {
+      "epoch": 5.817891373801917,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 3642
+    },
+    {
+      "epoch": 5.819488817891374,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 3643
+    },
+    {
+      "epoch": 5.821086261980831,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 3644
+    },
+    {
+      "epoch": 5.822683706070287,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3645
+    },
+    {
+      "epoch": 5.824281150159744,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3646
+    },
+    {
+      "epoch": 5.825878594249201,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 3647
+    },
+    {
+      "epoch": 5.827476038338658,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 3648
+    },
+    {
+      "epoch": 5.8290734824281145,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 3649
+    },
+    {
+      "epoch": 5.830670926517572,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 3650
+    },
+    {
+      "epoch": 5.832268370607029,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 3651
+    },
+    {
+      "epoch": 5.833865814696486,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3652
+    },
+    {
+      "epoch": 5.835463258785943,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3653
+    },
+    {
+      "epoch": 5.8370607028754,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3654
+    },
+    {
+      "epoch": 5.838658146964856,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 3655
+    },
+    {
+      "epoch": 5.840255591054313,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3656
+    },
+    {
+      "epoch": 5.84185303514377,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3657
+    },
+    {
+      "epoch": 5.843450479233227,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3658
+    },
+    {
+      "epoch": 5.845047923322683,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 3659
+    },
+    {
+      "epoch": 5.84664536741214,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3660
+    },
+    {
+      "epoch": 5.848242811501597,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3661
+    },
+    {
+      "epoch": 5.8498402555910545,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3662
+    },
+    {
+      "epoch": 5.8514376996805115,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3663
+    },
+    {
+      "epoch": 5.853035143769968,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 3664
+    },
+    {
+      "epoch": 5.854632587859425,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 3665
+    },
+    {
+      "epoch": 5.856230031948882,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3666
+    },
+    {
+      "epoch": 5.857827476038339,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 3667
+    },
+    {
+      "epoch": 5.859424920127795,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3668
+    },
+    {
+      "epoch": 5.861022364217252,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 3669
+    },
+    {
+      "epoch": 5.862619808306709,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3670
+    },
+    {
+      "epoch": 5.864217252396166,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3671
+    },
+    {
+      "epoch": 5.865814696485623,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3672
+    },
+    {
+      "epoch": 5.86741214057508,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 3673
+    },
+    {
+      "epoch": 5.8690095846645365,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 3674
+    },
+    {
+      "epoch": 5.8706070287539935,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 3675
+    },
+    {
+      "epoch": 5.872204472843451,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 3676
+    },
+    {
+      "epoch": 5.873801916932908,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 3677
+    },
+    {
+      "epoch": 5.875399361022364,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.156,
+      "step": 3678
+    },
+    {
+      "epoch": 5.876996805111821,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 3679
+    },
+    {
+      "epoch": 5.878594249201278,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3680
+    },
+    {
+      "epoch": 5.880191693290735,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 3681
+    },
+    {
+      "epoch": 5.881789137380192,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 3682
+    },
+    {
+      "epoch": 5.883386581469648,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3683
+    },
+    {
+      "epoch": 5.884984025559105,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 3684
+    },
+    {
+      "epoch": 5.886581469648562,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3685
+    },
+    {
+      "epoch": 5.888178913738019,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 3686
+    },
+    {
+      "epoch": 5.8897763578274756,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3687
+    },
+    {
+      "epoch": 5.891373801916933,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 3688
+    },
+    {
+      "epoch": 5.89297124600639,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3689
+    },
+    {
+      "epoch": 5.894568690095847,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3690
+    },
+    {
+      "epoch": 5.896166134185304,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 3691
+    },
+    {
+      "epoch": 5.897763578274761,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 3692
+    },
+    {
+      "epoch": 5.899361022364217,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 3693
+    },
+    {
+      "epoch": 5.900958466453674,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 3694
+    },
+    {
+      "epoch": 5.902555910543131,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 3695
+    },
+    {
+      "epoch": 5.904153354632588,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1555,
+      "step": 3696
+    },
+    {
+      "epoch": 5.905750798722044,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 3697
+    },
+    {
+      "epoch": 5.907348242811501,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 3698
+    },
+    {
+      "epoch": 5.9089456869009584,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3699
+    },
+    {
+      "epoch": 5.9105431309904155,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.161,
+      "step": 3700
+    },
+    {
+      "epoch": 5.912140575079873,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 3701
+    },
+    {
+      "epoch": 5.913738019169329,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 3702
+    },
+    {
+      "epoch": 5.915335463258786,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3703
+    },
+    {
+      "epoch": 5.916932907348243,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 3704
+    },
+    {
+      "epoch": 5.9185303514377,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 3705
+    },
+    {
+      "epoch": 5.920127795527156,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 3706
+    },
+    {
+      "epoch": 5.921725239616613,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 3707
+    },
+    {
+      "epoch": 5.92332268370607,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 3708
+    },
+    {
+      "epoch": 5.924920127795527,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 3709
+    },
+    {
+      "epoch": 5.926517571884984,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 3710
+    },
+    {
+      "epoch": 5.928115015974441,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3711
+    },
+    {
+      "epoch": 5.9297124600638975,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3712
+    },
+    {
+      "epoch": 5.931309904153355,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 3713
+    },
+    {
+      "epoch": 5.932907348242812,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3714
+    },
+    {
+      "epoch": 5.934504792332269,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 3715
+    },
+    {
+      "epoch": 5.936102236421725,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 3716
+    },
+    {
+      "epoch": 5.937699680511182,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 3717
+    },
+    {
+      "epoch": 5.939297124600639,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3718
+    },
+    {
+      "epoch": 5.940894568690096,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3719
+    },
+    {
+      "epoch": 5.942492012779553,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 3720
+    },
+    {
+      "epoch": 5.944089456869009,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 3721
+    },
+    {
+      "epoch": 5.945686900958466,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1576,
+      "step": 3722
+    },
+    {
+      "epoch": 5.947284345047923,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3723
+    },
+    {
+      "epoch": 5.94888178913738,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 3724
+    },
+    {
+      "epoch": 5.950479233226837,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3725
+    },
+    {
+      "epoch": 5.952076677316294,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3726
+    },
+    {
+      "epoch": 5.953674121405751,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 3727
+    },
+    {
+      "epoch": 5.955271565495208,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3728
+    },
+    {
+      "epoch": 5.956869009584665,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 3729
+    },
+    {
+      "epoch": 5.958466453674122,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3730
+    },
+    {
+      "epoch": 5.960063897763578,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 3731
+    },
+    {
+      "epoch": 5.961661341853035,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 3732
+    },
+    {
+      "epoch": 5.963258785942492,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3733
+    },
+    {
+      "epoch": 5.964856230031949,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 3734
+    },
+    {
+      "epoch": 5.966453674121405,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 3735
+    },
+    {
+      "epoch": 5.968051118210862,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 3736
+    },
+    {
+      "epoch": 5.9696485623003195,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3737
+    },
+    {
+      "epoch": 5.9712460063897765,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1551,
+      "step": 3738
+    },
+    {
+      "epoch": 5.972843450479234,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 3739
+    },
+    {
+      "epoch": 5.97444089456869,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 3740
+    },
+    {
+      "epoch": 5.976038338658147,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 3741
+    },
+    {
+      "epoch": 5.977635782747604,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 3742
+    },
+    {
+      "epoch": 5.979233226837061,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 3743
+    },
+    {
+      "epoch": 5.980830670926517,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 3744
+    },
+    {
+      "epoch": 5.982428115015974,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 3745
+    },
+    {
+      "epoch": 5.984025559105431,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3746
+    },
+    {
+      "epoch": 5.985623003194888,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 3747
+    },
+    {
+      "epoch": 5.987220447284345,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3748
+    },
+    {
+      "epoch": 5.988817891373802,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3749
+    },
+    {
+      "epoch": 5.9904153354632586,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 3750
+    },
+    {
+      "epoch": 5.992012779552716,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 3751
+    },
+    {
+      "epoch": 5.993610223642173,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 3752
+    },
+    {
+      "epoch": 5.99520766773163,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 3753
+    },
+    {
+      "epoch": 5.996805111821086,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 3754
+    },
+    {
+      "epoch": 5.998402555910543,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 3755
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 3756
+    },
+    {
+      "epoch": 6.001597444089457,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 3757
+    },
+    {
+      "epoch": 6.003194888178914,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3758
+    },
+    {
+      "epoch": 6.00479233226837,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1608,
+      "step": 3759
+    },
+    {
+      "epoch": 6.006389776357827,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3760
+    },
+    {
+      "epoch": 6.007987220447284,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 3761
+    },
+    {
+      "epoch": 6.0095846645367414,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3762
+    },
+    {
+      "epoch": 6.0111821086261985,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 3763
+    },
+    {
+      "epoch": 6.012779552715655,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 3764
+    },
+    {
+      "epoch": 6.014376996805112,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 3765
+    },
+    {
+      "epoch": 6.015974440894569,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 3766
+    },
+    {
+      "epoch": 6.017571884984026,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 3767
+    },
+    {
+      "epoch": 6.019169329073482,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 3768
+    },
+    {
+      "epoch": 6.020766773162939,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 3769
+    },
+    {
+      "epoch": 6.022364217252396,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 3770
+    },
+    {
+      "epoch": 6.023961661341853,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 3771
+    },
+    {
+      "epoch": 6.02555910543131,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3772
+    },
+    {
+      "epoch": 6.027156549520766,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 3773
+    },
+    {
+      "epoch": 6.0287539936102235,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3774
+    },
+    {
+      "epoch": 6.0303514376996805,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3775
+    },
+    {
+      "epoch": 6.031948881789138,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 3776
+    },
+    {
+      "epoch": 6.033546325878595,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 3777
+    },
+    {
+      "epoch": 6.035143769968051,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 3778
+    },
+    {
+      "epoch": 6.036741214057508,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3779
+    },
+    {
+      "epoch": 6.038338658146965,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 3780
+    },
+    {
+      "epoch": 6.039936102236422,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3781
+    },
+    {
+      "epoch": 6.041533546325879,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 3782
+    },
+    {
+      "epoch": 6.043130990415335,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 3783
+    },
+    {
+      "epoch": 6.044728434504792,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 3784
+    },
+    {
+      "epoch": 6.046325878594249,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3785
+    },
+    {
+      "epoch": 6.047923322683706,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 3786
+    },
+    {
+      "epoch": 6.0495207667731625,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 3787
+    },
+    {
+      "epoch": 6.05111821086262,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 3788
+    },
+    {
+      "epoch": 6.052715654952077,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3789
+    },
+    {
+      "epoch": 6.054313099041534,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 3790
+    },
+    {
+      "epoch": 6.055910543130991,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 3791
+    },
+    {
+      "epoch": 6.057507987220447,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 3792
+    },
+    {
+      "epoch": 6.059105431309904,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 3793
+    },
+    {
+      "epoch": 6.060702875399361,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 3794
+    },
+    {
+      "epoch": 6.062300319488818,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 3795
+    },
+    {
+      "epoch": 6.063897763578275,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 3796
+    },
+    {
+      "epoch": 6.065495207667731,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 3797
+    },
+    {
+      "epoch": 6.067092651757188,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 3798
+    },
+    {
+      "epoch": 6.068690095846645,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 3799
+    },
+    {
+      "epoch": 6.0702875399361025,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3800
+    },
+    {
+      "epoch": 6.0718849840255595,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3801
+    },
+    {
+      "epoch": 6.073482428115016,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3802
+    },
+    {
+      "epoch": 6.075079872204473,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3803
+    },
+    {
+      "epoch": 6.07667731629393,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 3804
+    },
+    {
+      "epoch": 6.078274760383387,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 3805
+    },
+    {
+      "epoch": 6.079872204472843,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 3806
+    },
+    {
+      "epoch": 6.0814696485623,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3807
+    },
+    {
+      "epoch": 6.083067092651757,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 3808
+    },
+    {
+      "epoch": 6.084664536741214,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3809
+    },
+    {
+      "epoch": 6.086261980830671,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3810
+    },
+    {
+      "epoch": 6.087859424920127,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 3811
+    },
+    {
+      "epoch": 6.0894568690095845,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 3812
+    },
+    {
+      "epoch": 6.0910543130990416,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 3813
+    },
+    {
+      "epoch": 6.092651757188499,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 3814
+    },
+    {
+      "epoch": 6.094249201277956,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 3815
+    },
+    {
+      "epoch": 6.095846645367412,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 3816
+    },
+    {
+      "epoch": 6.097444089456869,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 3817
+    },
+    {
+      "epoch": 6.099041533546326,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 3818
+    },
+    {
+      "epoch": 6.100638977635783,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3819
+    },
+    {
+      "epoch": 6.102236421725239,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 3820
+    },
+    {
+      "epoch": 6.103833865814696,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 3821
+    },
+    {
+      "epoch": 6.105431309904153,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3822
+    },
+    {
+      "epoch": 6.10702875399361,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 3823
+    },
+    {
+      "epoch": 6.108626198083067,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 3824
+    },
+    {
+      "epoch": 6.110223642172524,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3825
+    },
+    {
+      "epoch": 6.111821086261981,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 3826
+    },
+    {
+      "epoch": 6.113418530351438,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 3827
+    },
+    {
+      "epoch": 6.115015974440895,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1582,
+      "step": 3828
+    },
+    {
+      "epoch": 6.116613418530352,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3829
+    },
+    {
+      "epoch": 6.118210862619808,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1611,
+      "step": 3830
+    },
+    {
+      "epoch": 6.119808306709265,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 3831
+    },
+    {
+      "epoch": 6.121405750798722,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3832
+    },
+    {
+      "epoch": 6.123003194888179,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 3833
+    },
+    {
+      "epoch": 6.124600638977636,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 3834
+    },
+    {
+      "epoch": 6.126198083067092,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 3835
+    },
+    {
+      "epoch": 6.127795527156549,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3836
+    },
+    {
+      "epoch": 6.1293929712460065,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 3837
+    },
+    {
+      "epoch": 6.1309904153354635,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 3838
+    },
+    {
+      "epoch": 6.13258785942492,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 3839
+    },
+    {
+      "epoch": 6.134185303514377,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 3840
+    },
+    {
+      "epoch": 6.135782747603834,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3841
+    },
+    {
+      "epoch": 6.137380191693291,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 3842
+    },
+    {
+      "epoch": 6.138977635782748,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 3843
+    },
+    {
+      "epoch": 6.140575079872204,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3844
+    },
+    {
+      "epoch": 6.142172523961661,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 3845
+    },
+    {
+      "epoch": 6.143769968051118,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 3846
+    },
+    {
+      "epoch": 6.145367412140575,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 3847
+    },
+    {
+      "epoch": 6.146964856230032,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 3848
+    },
+    {
+      "epoch": 6.1485623003194885,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3849
+    },
+    {
+      "epoch": 6.1501597444089455,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 3850
+    },
+    {
+      "epoch": 6.151757188498403,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 3851
+    },
+    {
+      "epoch": 6.15335463258786,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 3852
+    },
+    {
+      "epoch": 6.154952076677317,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 3853
+    },
+    {
+      "epoch": 6.156549520766773,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 3854
+    },
+    {
+      "epoch": 6.15814696485623,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3855
+    },
+    {
+      "epoch": 6.159744408945687,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 3856
+    },
+    {
+      "epoch": 6.161341853035144,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 3857
+    },
+    {
+      "epoch": 6.1629392971246,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3858
+    },
+    {
+      "epoch": 6.164536741214057,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 3859
+    },
+    {
+      "epoch": 6.166134185303514,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1577,
+      "step": 3860
+    },
+    {
+      "epoch": 6.167731629392971,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 3861
+    },
+    {
+      "epoch": 6.169329073482428,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 3862
+    },
+    {
+      "epoch": 6.170926517571885,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 3863
+    },
+    {
+      "epoch": 6.172523961661342,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3864
+    },
+    {
+      "epoch": 6.174121405750799,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 3865
+    },
+    {
+      "epoch": 6.175718849840256,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 3866
+    },
+    {
+      "epoch": 6.177316293929713,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 3867
+    },
+    {
+      "epoch": 6.178913738019169,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 3868
+    },
+    {
+      "epoch": 6.180511182108626,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3869
+    },
+    {
+      "epoch": 6.182108626198083,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 3870
+    },
+    {
+      "epoch": 6.18370607028754,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 3871
+    },
+    {
+      "epoch": 6.185303514376997,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 3872
+    },
+    {
+      "epoch": 6.186900958466453,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 3873
+    },
+    {
+      "epoch": 6.18849840255591,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 3874
+    },
+    {
+      "epoch": 6.1900958466453675,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 3875
+    },
+    {
+      "epoch": 6.1916932907348246,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3876
+    },
+    {
+      "epoch": 6.193290734824281,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3877
+    },
+    {
+      "epoch": 6.194888178913738,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3878
+    },
+    {
+      "epoch": 6.196485623003195,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 3879
+    },
+    {
+      "epoch": 6.198083067092652,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1539,
+      "step": 3880
+    },
+    {
+      "epoch": 6.199680511182109,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 3881
+    },
+    {
+      "epoch": 6.201277955271565,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 3882
+    },
+    {
+      "epoch": 6.202875399361022,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 3883
+    },
+    {
+      "epoch": 6.204472843450479,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 3884
+    },
+    {
+      "epoch": 6.206070287539936,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 3885
+    },
+    {
+      "epoch": 6.207667731629393,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 3886
+    },
+    {
+      "epoch": 6.2092651757188495,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 3887
+    },
+    {
+      "epoch": 6.210862619808307,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 3888
+    },
+    {
+      "epoch": 6.212460063897764,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 3889
+    },
+    {
+      "epoch": 6.214057507987221,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 3890
+    },
+    {
+      "epoch": 6.215654952076678,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 3891
+    },
+    {
+      "epoch": 6.217252396166134,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3892
+    },
+    {
+      "epoch": 6.218849840255591,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 3893
+    },
+    {
+      "epoch": 6.220447284345048,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1491,
+      "step": 3894
+    },
+    {
+      "epoch": 6.222044728434505,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 3895
+    },
+    {
+      "epoch": 6.223642172523961,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 3896
+    },
+    {
+      "epoch": 6.225239616613418,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 3897
+    },
+    {
+      "epoch": 6.226837060702875,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3898
+    },
+    {
+      "epoch": 6.228434504792332,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3899
+    },
+    {
+      "epoch": 6.2300319488817895,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 3900
+    },
+    {
+      "epoch": 6.231629392971246,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 3901
+    },
+    {
+      "epoch": 6.233226837060703,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 3902
+    },
+    {
+      "epoch": 6.23482428115016,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 3903
+    },
+    {
+      "epoch": 6.236421725239617,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 3904
+    },
+    {
+      "epoch": 6.238019169329074,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 3905
+    },
+    {
+      "epoch": 6.23961661341853,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 3906
+    },
+    {
+      "epoch": 6.241214057507987,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3907
+    },
+    {
+      "epoch": 6.242811501597444,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 3908
+    },
+    {
+      "epoch": 6.244408945686901,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 3909
+    },
+    {
+      "epoch": 6.246006389776358,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3910
+    },
+    {
+      "epoch": 6.247603833865814,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3911
+    },
+    {
+      "epoch": 6.2492012779552715,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 3912
+    },
+    {
+      "epoch": 6.2507987220447285,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 3913
+    },
+    {
+      "epoch": 6.252396166134186,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 3914
+    },
+    {
+      "epoch": 6.253993610223642,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 3915
+    },
+    {
+      "epoch": 6.255591054313099,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3916
+    },
+    {
+      "epoch": 6.257188498402556,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 3917
+    },
+    {
+      "epoch": 6.258785942492013,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3918
+    },
+    {
+      "epoch": 6.26038338658147,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 3919
+    },
+    {
+      "epoch": 6.261980830670926,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 3920
+    },
+    {
+      "epoch": 6.263578274760383,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 3921
+    },
+    {
+      "epoch": 6.26517571884984,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 3922
+    },
+    {
+      "epoch": 6.266773162939297,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 3923
+    },
+    {
+      "epoch": 6.268370607028754,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 3924
+    },
+    {
+      "epoch": 6.2699680511182105,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 3925
+    },
+    {
+      "epoch": 6.271565495207668,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 3926
+    },
+    {
+      "epoch": 6.273162939297125,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3927
+    },
+    {
+      "epoch": 6.274760383386582,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 3928
+    },
+    {
+      "epoch": 6.276357827476039,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 3929
+    },
+    {
+      "epoch": 6.277955271565495,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 3930
+    },
+    {
+      "epoch": 6.279552715654952,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 3931
+    },
+    {
+      "epoch": 6.281150159744409,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 3932
+    },
+    {
+      "epoch": 6.282747603833866,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 3933
+    },
+    {
+      "epoch": 6.284345047923322,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3934
+    },
+    {
+      "epoch": 6.285942492012779,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 3935
+    },
+    {
+      "epoch": 6.287539936102236,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3936
+    },
+    {
+      "epoch": 6.289137380191693,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 3937
+    },
+    {
+      "epoch": 6.2907348242811505,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 3938
+    },
+    {
+      "epoch": 6.292332268370607,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 3939
+    },
+    {
+      "epoch": 6.293929712460064,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 3940
+    },
+    {
+      "epoch": 6.295527156549521,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 3941
+    },
+    {
+      "epoch": 6.297124600638978,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 3942
+    },
+    {
+      "epoch": 6.298722044728435,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 3943
+    },
+    {
+      "epoch": 6.300319488817891,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1588,
+      "step": 3944
+    },
+    {
+      "epoch": 6.301916932907348,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 3945
+    },
+    {
+      "epoch": 6.303514376996805,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 3946
+    },
+    {
+      "epoch": 6.305111821086262,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3947
+    },
+    {
+      "epoch": 6.306709265175719,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 3948
+    },
+    {
+      "epoch": 6.3083067092651754,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 3949
+    },
+    {
+      "epoch": 6.3099041533546325,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 3950
+    },
+    {
+      "epoch": 6.31150159744409,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 3951
+    },
+    {
+      "epoch": 6.313099041533547,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 3952
+    },
+    {
+      "epoch": 6.314696485623003,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 3953
+    },
+    {
+      "epoch": 6.31629392971246,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 3954
+    },
+    {
+      "epoch": 6.317891373801917,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 3955
+    },
+    {
+      "epoch": 6.319488817891374,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 3956
+    },
+    {
+      "epoch": 6.321086261980831,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 3957
+    },
+    {
+      "epoch": 6.322683706070287,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 3958
+    },
+    {
+      "epoch": 6.324281150159744,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 3959
+    },
+    {
+      "epoch": 6.325878594249201,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 3960
+    },
+    {
+      "epoch": 6.327476038338658,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 3961
+    },
+    {
+      "epoch": 6.329073482428115,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 3962
+    },
+    {
+      "epoch": 6.330670926517572,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 3963
+    },
+    {
+      "epoch": 6.332268370607029,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 3964
+    },
+    {
+      "epoch": 6.333865814696486,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 3965
+    },
+    {
+      "epoch": 6.335463258785943,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3966
+    },
+    {
+      "epoch": 6.3370607028754,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3967
+    },
+    {
+      "epoch": 6.338658146964856,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 3968
+    },
+    {
+      "epoch": 6.340255591054313,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 3969
+    },
+    {
+      "epoch": 6.34185303514377,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 3970
+    },
+    {
+      "epoch": 6.343450479233227,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 3971
+    },
+    {
+      "epoch": 6.345047923322683,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 3972
+    },
+    {
+      "epoch": 6.34664536741214,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3973
+    },
+    {
+      "epoch": 6.348242811501597,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 3974
+    },
+    {
+      "epoch": 6.3498402555910545,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 3975
+    },
+    {
+      "epoch": 6.3514376996805115,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 3976
+    },
+    {
+      "epoch": 6.353035143769968,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 3977
+    },
+    {
+      "epoch": 6.354632587859425,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 3978
+    },
+    {
+      "epoch": 6.356230031948882,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 3979
+    },
+    {
+      "epoch": 6.357827476038339,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 3980
+    },
+    {
+      "epoch": 6.359424920127796,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1637,
+      "step": 3981
+    },
+    {
+      "epoch": 6.361022364217252,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 3982
+    },
+    {
+      "epoch": 6.362619808306709,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 3983
+    },
+    {
+      "epoch": 6.364217252396166,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 3984
+    },
+    {
+      "epoch": 6.365814696485623,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 3985
+    },
+    {
+      "epoch": 6.36741214057508,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 3986
+    },
+    {
+      "epoch": 6.3690095846645365,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 3987
+    },
+    {
+      "epoch": 6.3706070287539935,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 3988
+    },
+    {
+      "epoch": 6.372204472843451,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 3989
+    },
+    {
+      "epoch": 6.373801916932908,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 3990
+    },
+    {
+      "epoch": 6.375399361022364,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 3991
+    },
+    {
+      "epoch": 6.376996805111821,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 3992
+    },
+    {
+      "epoch": 6.378594249201278,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 3993
+    },
+    {
+      "epoch": 6.380191693290735,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 3994
+    },
+    {
+      "epoch": 6.381789137380192,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 3995
+    },
+    {
+      "epoch": 6.383386581469648,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 3996
+    },
+    {
+      "epoch": 6.384984025559105,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 3997
+    },
+    {
+      "epoch": 6.386581469648562,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 3998
+    },
+    {
+      "epoch": 6.388178913738019,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 3999
+    },
+    {
+      "epoch": 6.389776357827476,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 4000
+    },
+    {
+      "epoch": 6.391373801916933,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 4001
+    },
+    {
+      "epoch": 6.39297124600639,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4002
+    },
+    {
+      "epoch": 6.394568690095847,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.155,
+      "step": 4003
+    },
+    {
+      "epoch": 6.396166134185304,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 4004
+    },
+    {
+      "epoch": 6.397763578274761,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1567,
+      "step": 4005
+    },
+    {
+      "epoch": 6.399361022364217,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4006
+    },
+    {
+      "epoch": 6.400958466453674,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4007
+    },
+    {
+      "epoch": 6.402555910543131,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4008
+    },
+    {
+      "epoch": 6.404153354632588,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 4009
+    },
+    {
+      "epoch": 6.405750798722044,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4010
+    },
+    {
+      "epoch": 6.407348242811501,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4011
+    },
+    {
+      "epoch": 6.4089456869009584,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 4012
+    },
+    {
+      "epoch": 6.4105431309904155,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 4013
+    },
+    {
+      "epoch": 6.412140575079873,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 4014
+    },
+    {
+      "epoch": 6.413738019169329,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 4015
+    },
+    {
+      "epoch": 6.415335463258786,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 4016
+    },
+    {
+      "epoch": 6.416932907348243,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4017
+    },
+    {
+      "epoch": 6.4185303514377,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 4018
+    },
+    {
+      "epoch": 6.420127795527157,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 4019
+    },
+    {
+      "epoch": 6.421725239616613,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1491,
+      "step": 4020
+    },
+    {
+      "epoch": 6.42332268370607,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 4021
+    },
+    {
+      "epoch": 6.424920127795527,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4022
+    },
+    {
+      "epoch": 6.426517571884984,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4023
+    },
+    {
+      "epoch": 6.428115015974441,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4024
+    },
+    {
+      "epoch": 6.4297124600638975,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4025
+    },
+    {
+      "epoch": 6.431309904153355,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4026
+    },
+    {
+      "epoch": 6.432907348242812,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4027
+    },
+    {
+      "epoch": 6.434504792332269,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4028
+    },
+    {
+      "epoch": 6.436102236421725,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 4029
+    },
+    {
+      "epoch": 6.437699680511182,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4030
+    },
+    {
+      "epoch": 6.439297124600639,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0005,
+      "loss": 1.155,
+      "step": 4031
+    },
+    {
+      "epoch": 6.440894568690096,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 4032
+    },
+    {
+      "epoch": 6.442492012779553,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4033
+    },
+    {
+      "epoch": 6.444089456869009,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 4034
+    },
+    {
+      "epoch": 6.445686900958466,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 4035
+    },
+    {
+      "epoch": 6.447284345047923,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4036
+    },
+    {
+      "epoch": 6.44888178913738,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1543,
+      "step": 4037
+    },
+    {
+      "epoch": 6.4504792332268375,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 4038
+    },
+    {
+      "epoch": 6.452076677316294,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 4039
+    },
+    {
+      "epoch": 6.453674121405751,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1597,
+      "step": 4040
+    },
+    {
+      "epoch": 6.455271565495208,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 4041
+    },
+    {
+      "epoch": 6.456869009584665,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4042
+    },
+    {
+      "epoch": 6.458466453674122,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 4043
+    },
+    {
+      "epoch": 6.460063897763578,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1599,
+      "step": 4044
+    },
+    {
+      "epoch": 6.461661341853035,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4045
+    },
+    {
+      "epoch": 6.463258785942492,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 4046
+    },
+    {
+      "epoch": 6.464856230031949,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4047
+    },
+    {
+      "epoch": 6.466453674121405,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 4048
+    },
+    {
+      "epoch": 6.468051118210862,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4049
+    },
+    {
+      "epoch": 6.4696485623003195,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4050
+    },
+    {
+      "epoch": 6.4712460063897765,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4051
+    },
+    {
+      "epoch": 6.472843450479234,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 4052
+    },
+    {
+      "epoch": 6.47444089456869,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4053
+    },
+    {
+      "epoch": 6.476038338658147,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1594,
+      "step": 4054
+    },
+    {
+      "epoch": 6.477635782747604,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 4055
+    },
+    {
+      "epoch": 6.479233226837061,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4056
+    },
+    {
+      "epoch": 6.480830670926517,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 4057
+    },
+    {
+      "epoch": 6.482428115015974,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4058
+    },
+    {
+      "epoch": 6.484025559105431,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 4059
+    },
+    {
+      "epoch": 6.485623003194888,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 4060
+    },
+    {
+      "epoch": 6.487220447284345,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4061
+    },
+    {
+      "epoch": 6.488817891373802,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 4062
+    },
+    {
+      "epoch": 6.4904153354632586,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 4063
+    },
+    {
+      "epoch": 6.492012779552716,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 4064
+    },
+    {
+      "epoch": 6.493610223642173,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1569,
+      "step": 4065
+    },
+    {
+      "epoch": 6.49520766773163,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 4066
+    },
+    {
+      "epoch": 6.496805111821086,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 4067
+    },
+    {
+      "epoch": 6.498402555910543,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 4068
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 4069
+    },
+    {
+      "epoch": 6.501597444089457,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4070
+    },
+    {
+      "epoch": 6.503194888178914,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4071
+    },
+    {
+      "epoch": 6.50479233226837,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 4072
+    },
+    {
+      "epoch": 6.506389776357827,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 4073
+    },
+    {
+      "epoch": 6.507987220447284,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4074
+    },
+    {
+      "epoch": 6.5095846645367414,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4075
+    },
+    {
+      "epoch": 6.511182108626198,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 4076
+    },
+    {
+      "epoch": 6.512779552715655,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 4077
+    },
+    {
+      "epoch": 6.514376996805112,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4078
+    },
+    {
+      "epoch": 6.515974440894569,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 4079
+    },
+    {
+      "epoch": 6.517571884984026,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4080
+    },
+    {
+      "epoch": 6.519169329073483,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1538,
+      "step": 4081
+    },
+    {
+      "epoch": 6.520766773162939,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 4082
+    },
+    {
+      "epoch": 6.522364217252396,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4083
+    },
+    {
+      "epoch": 6.523961661341853,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4084
+    },
+    {
+      "epoch": 6.52555910543131,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 4085
+    },
+    {
+      "epoch": 6.527156549520766,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4086
+    },
+    {
+      "epoch": 6.5287539936102235,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4087
+    },
+    {
+      "epoch": 6.5303514376996805,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4088
+    },
+    {
+      "epoch": 6.531948881789138,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 4089
+    },
+    {
+      "epoch": 6.533546325878595,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4090
+    },
+    {
+      "epoch": 6.535143769968051,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4091
+    },
+    {
+      "epoch": 6.536741214057508,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 4092
+    },
+    {
+      "epoch": 6.538338658146965,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4093
+    },
+    {
+      "epoch": 6.539936102236422,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4094
+    },
+    {
+      "epoch": 6.541533546325878,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4095
+    },
+    {
+      "epoch": 6.543130990415335,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 4096
+    },
+    {
+      "epoch": 6.544728434504792,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 4097
+    },
+    {
+      "epoch": 6.546325878594249,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4098
+    },
+    {
+      "epoch": 6.547923322683706,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 4099
+    },
+    {
+      "epoch": 6.549520766773163,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 4100
+    },
+    {
+      "epoch": 6.55111821086262,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 4101
+    },
+    {
+      "epoch": 6.552715654952077,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4102
+    },
+    {
+      "epoch": 6.554313099041534,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4103
+    },
+    {
+      "epoch": 6.555910543130991,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 4104
+    },
+    {
+      "epoch": 6.557507987220447,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 4105
+    },
+    {
+      "epoch": 6.559105431309904,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 4106
+    },
+    {
+      "epoch": 6.560702875399361,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4107
+    },
+    {
+      "epoch": 6.562300319488818,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 4108
+    },
+    {
+      "epoch": 6.563897763578275,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4109
+    },
+    {
+      "epoch": 6.565495207667731,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4110
+    },
+    {
+      "epoch": 6.567092651757188,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4111
+    },
+    {
+      "epoch": 6.568690095846645,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4112
+    },
+    {
+      "epoch": 6.5702875399361025,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1576,
+      "step": 4113
+    },
+    {
+      "epoch": 6.571884984025559,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 4114
+    },
+    {
+      "epoch": 6.573482428115016,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4115
+    },
+    {
+      "epoch": 6.575079872204473,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4116
+    },
+    {
+      "epoch": 6.57667731629393,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 4117
+    },
+    {
+      "epoch": 6.578274760383387,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 4118
+    },
+    {
+      "epoch": 6.579872204472844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 4119
+    },
+    {
+      "epoch": 6.5814696485623,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 4120
+    },
+    {
+      "epoch": 6.583067092651757,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4121
+    },
+    {
+      "epoch": 6.584664536741214,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4122
+    },
+    {
+      "epoch": 6.586261980830671,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 4123
+    },
+    {
+      "epoch": 6.587859424920127,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 4124
+    },
+    {
+      "epoch": 6.5894568690095845,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 4125
+    },
+    {
+      "epoch": 6.5910543130990416,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 4126
+    },
+    {
+      "epoch": 6.592651757188499,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4127
+    },
+    {
+      "epoch": 6.594249201277956,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4128
+    },
+    {
+      "epoch": 6.595846645367412,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 4129
+    },
+    {
+      "epoch": 6.597444089456869,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 4130
+    },
+    {
+      "epoch": 6.599041533546326,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 4131
+    },
+    {
+      "epoch": 6.600638977635783,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4132
+    },
+    {
+      "epoch": 6.602236421725239,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 4133
+    },
+    {
+      "epoch": 6.603833865814696,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4134
+    },
+    {
+      "epoch": 6.605431309904153,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 4135
+    },
+    {
+      "epoch": 6.60702875399361,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 4136
+    },
+    {
+      "epoch": 6.608626198083067,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 4137
+    },
+    {
+      "epoch": 6.6102236421725244,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 4138
+    },
+    {
+      "epoch": 6.611821086261981,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 4139
+    },
+    {
+      "epoch": 6.613418530351438,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 4140
+    },
+    {
+      "epoch": 6.615015974440895,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4141
+    },
+    {
+      "epoch": 6.616613418530352,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4142
+    },
+    {
+      "epoch": 6.618210862619808,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4143
+    },
+    {
+      "epoch": 6.619808306709265,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4144
+    },
+    {
+      "epoch": 6.621405750798722,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1566,
+      "step": 4145
+    },
+    {
+      "epoch": 6.623003194888179,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 4146
+    },
+    {
+      "epoch": 6.624600638977636,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 4147
+    },
+    {
+      "epoch": 6.626198083067092,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 4148
+    },
+    {
+      "epoch": 6.627795527156549,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 4149
+    },
+    {
+      "epoch": 6.6293929712460065,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 4150
+    },
+    {
+      "epoch": 6.6309904153354635,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 4151
+    },
+    {
+      "epoch": 6.63258785942492,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 4152
+    },
+    {
+      "epoch": 6.634185303514377,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4153
+    },
+    {
+      "epoch": 6.635782747603834,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4154
+    },
+    {
+      "epoch": 6.637380191693291,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4155
+    },
+    {
+      "epoch": 6.638977635782748,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 4156
+    },
+    {
+      "epoch": 6.640575079872205,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4157
+    },
+    {
+      "epoch": 6.642172523961661,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4158
+    },
+    {
+      "epoch": 6.643769968051118,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4159
+    },
+    {
+      "epoch": 6.645367412140575,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4160
+    },
+    {
+      "epoch": 6.646964856230032,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 4161
+    },
+    {
+      "epoch": 6.6485623003194885,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 4162
+    },
+    {
+      "epoch": 6.6501597444089455,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4163
+    },
+    {
+      "epoch": 6.651757188498403,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 4164
+    },
+    {
+      "epoch": 6.65335463258786,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 4165
+    },
+    {
+      "epoch": 6.654952076677317,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4166
+    },
+    {
+      "epoch": 6.656549520766773,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 4167
+    },
+    {
+      "epoch": 6.65814696485623,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 4168
+    },
+    {
+      "epoch": 6.659744408945687,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 4169
+    },
+    {
+      "epoch": 6.661341853035144,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 4170
+    },
+    {
+      "epoch": 6.6629392971246,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 4171
+    },
+    {
+      "epoch": 6.664536741214057,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4172
+    },
+    {
+      "epoch": 6.666134185303514,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 4173
+    },
+    {
+      "epoch": 6.667731629392971,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 4174
+    },
+    {
+      "epoch": 6.669329073482428,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 4175
+    },
+    {
+      "epoch": 6.6709265175718855,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 4176
+    },
+    {
+      "epoch": 6.672523961661342,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4177
+    },
+    {
+      "epoch": 6.674121405750799,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4178
+    },
+    {
+      "epoch": 6.675718849840256,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 4179
+    },
+    {
+      "epoch": 6.677316293929713,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 4180
+    },
+    {
+      "epoch": 6.678913738019169,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 4181
+    },
+    {
+      "epoch": 6.680511182108626,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 4182
+    },
+    {
+      "epoch": 6.682108626198083,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 4183
+    },
+    {
+      "epoch": 6.68370607028754,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4184
+    },
+    {
+      "epoch": 6.685303514376997,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4185
+    },
+    {
+      "epoch": 6.686900958466453,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 4186
+    },
+    {
+      "epoch": 6.68849840255591,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4187
+    },
+    {
+      "epoch": 6.6900958466453675,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 4188
+    },
+    {
+      "epoch": 6.6916932907348246,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4189
+    },
+    {
+      "epoch": 6.693290734824281,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 4190
+    },
+    {
+      "epoch": 6.694888178913738,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 4191
+    },
+    {
+      "epoch": 6.696485623003195,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4192
+    },
+    {
+      "epoch": 6.698083067092652,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 4193
+    },
+    {
+      "epoch": 6.699680511182109,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 4194
+    },
+    {
+      "epoch": 6.701277955271565,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4195
+    },
+    {
+      "epoch": 6.702875399361022,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1587,
+      "step": 4196
+    },
+    {
+      "epoch": 6.704472843450479,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 4197
+    },
+    {
+      "epoch": 6.706070287539936,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4198
+    },
+    {
+      "epoch": 6.707667731629393,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4199
+    },
+    {
+      "epoch": 6.7092651757188495,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 4200
+    },
+    {
+      "epoch": 6.710862619808307,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 4201
+    },
+    {
+      "epoch": 6.712460063897764,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 4202
+    },
+    {
+      "epoch": 6.714057507987221,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4203
+    },
+    {
+      "epoch": 6.715654952076678,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4204
+    },
+    {
+      "epoch": 6.717252396166134,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 4205
+    },
+    {
+      "epoch": 6.718849840255591,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 4206
+    },
+    {
+      "epoch": 6.720447284345048,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4207
+    },
+    {
+      "epoch": 6.722044728434505,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4208
+    },
+    {
+      "epoch": 6.723642172523961,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 4209
+    },
+    {
+      "epoch": 6.725239616613418,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4210
+    },
+    {
+      "epoch": 6.726837060702875,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 4211
+    },
+    {
+      "epoch": 6.728434504792332,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 4212
+    },
+    {
+      "epoch": 6.7300319488817895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 4213
+    },
+    {
+      "epoch": 6.731629392971246,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4214
+    },
+    {
+      "epoch": 6.733226837060703,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 4215
+    },
+    {
+      "epoch": 6.73482428115016,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 4216
+    },
+    {
+      "epoch": 6.736421725239617,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4217
+    },
+    {
+      "epoch": 6.738019169329074,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 4218
+    },
+    {
+      "epoch": 6.73961661341853,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 4219
+    },
+    {
+      "epoch": 6.741214057507987,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 4220
+    },
+    {
+      "epoch": 6.742811501597444,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 4221
+    },
+    {
+      "epoch": 6.744408945686901,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 4222
+    },
+    {
+      "epoch": 6.746006389776358,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 4223
+    },
+    {
+      "epoch": 6.747603833865814,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4224
+    },
+    {
+      "epoch": 6.7492012779552715,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4225
+    },
+    {
+      "epoch": 6.7507987220447285,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4226
+    },
+    {
+      "epoch": 6.752396166134186,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 4227
+    },
+    {
+      "epoch": 6.753993610223642,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 4228
+    },
+    {
+      "epoch": 6.755591054313099,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 4229
+    },
+    {
+      "epoch": 6.757188498402556,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4230
+    },
+    {
+      "epoch": 6.758785942492013,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4231
+    },
+    {
+      "epoch": 6.76038338658147,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 4232
+    },
+    {
+      "epoch": 6.761980830670926,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 4233
+    },
+    {
+      "epoch": 6.763578274760383,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4234
+    },
+    {
+      "epoch": 6.76517571884984,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4235
+    },
+    {
+      "epoch": 6.766773162939297,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 4236
+    },
+    {
+      "epoch": 6.768370607028754,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4237
+    },
+    {
+      "epoch": 6.7699680511182105,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 4238
+    },
+    {
+      "epoch": 6.771565495207668,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.1515,
+      "step": 4239
+    },
+    {
+      "epoch": 6.773162939297125,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4240
+    },
+    {
+      "epoch": 6.774760383386582,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 4241
+    },
+    {
+      "epoch": 6.776357827476039,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 4242
+    },
+    {
+      "epoch": 6.777955271565495,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 4243
+    },
+    {
+      "epoch": 6.779552715654952,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4244
+    },
+    {
+      "epoch": 6.781150159744409,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4245
+    },
+    {
+      "epoch": 6.782747603833866,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1556,
+      "step": 4246
+    },
+    {
+      "epoch": 6.784345047923322,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 4247
+    },
+    {
+      "epoch": 6.785942492012779,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 4248
+    },
+    {
+      "epoch": 6.787539936102236,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4249
+    },
+    {
+      "epoch": 6.789137380191693,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 4250
+    },
+    {
+      "epoch": 6.7907348242811505,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4251
+    },
+    {
+      "epoch": 6.792332268370607,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4252
+    },
+    {
+      "epoch": 6.793929712460064,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 4253
+    },
+    {
+      "epoch": 6.795527156549521,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4254
+    },
+    {
+      "epoch": 6.797124600638978,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 4255
+    },
+    {
+      "epoch": 6.798722044728435,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 4256
+    },
+    {
+      "epoch": 6.800319488817891,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1586,
+      "step": 4257
+    },
+    {
+      "epoch": 6.801916932907348,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 4258
+    },
+    {
+      "epoch": 6.803514376996805,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4259
+    },
+    {
+      "epoch": 6.805111821086262,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 4260
+    },
+    {
+      "epoch": 6.806709265175719,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4261
+    },
+    {
+      "epoch": 6.8083067092651754,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4262
+    },
+    {
+      "epoch": 6.8099041533546325,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 4263
+    },
+    {
+      "epoch": 6.81150159744409,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4264
+    },
+    {
+      "epoch": 6.813099041533547,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 4265
+    },
+    {
+      "epoch": 6.814696485623003,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4266
+    },
+    {
+      "epoch": 6.81629392971246,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4267
+    },
+    {
+      "epoch": 6.817891373801917,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4268
+    },
+    {
+      "epoch": 6.819488817891374,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 4269
+    },
+    {
+      "epoch": 6.821086261980831,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.155,
+      "step": 4270
+    },
+    {
+      "epoch": 6.822683706070287,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 4271
+    },
+    {
+      "epoch": 6.824281150159744,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4272
+    },
+    {
+      "epoch": 6.825878594249201,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 4273
+    },
+    {
+      "epoch": 6.827476038338658,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 4274
+    },
+    {
+      "epoch": 6.8290734824281145,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1572,
+      "step": 4275
+    },
+    {
+      "epoch": 6.830670926517572,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 4276
+    },
+    {
+      "epoch": 6.832268370607029,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 4277
+    },
+    {
+      "epoch": 6.833865814696486,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4278
+    },
+    {
+      "epoch": 6.835463258785943,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 4279
+    },
+    {
+      "epoch": 6.8370607028754,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4280
+    },
+    {
+      "epoch": 6.838658146964856,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 4281
+    },
+    {
+      "epoch": 6.840255591054313,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4282
+    },
+    {
+      "epoch": 6.84185303514377,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 4283
+    },
+    {
+      "epoch": 6.843450479233227,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 4284
+    },
+    {
+      "epoch": 6.845047923322683,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 4285
+    },
+    {
+      "epoch": 6.84664536741214,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4286
+    },
+    {
+      "epoch": 6.848242811501597,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4287
+    },
+    {
+      "epoch": 6.8498402555910545,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4288
+    },
+    {
+      "epoch": 6.8514376996805115,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 4289
+    },
+    {
+      "epoch": 6.853035143769968,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4290
+    },
+    {
+      "epoch": 6.854632587859425,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 4291
+    },
+    {
+      "epoch": 6.856230031948882,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 4292
+    },
+    {
+      "epoch": 6.857827476038339,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4293
+    },
+    {
+      "epoch": 6.859424920127795,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 4294
+    },
+    {
+      "epoch": 6.861022364217252,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 4295
+    },
+    {
+      "epoch": 6.862619808306709,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 4296
+    },
+    {
+      "epoch": 6.864217252396166,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 4297
+    },
+    {
+      "epoch": 6.865814696485623,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 4298
+    },
+    {
+      "epoch": 6.86741214057508,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4299
+    },
+    {
+      "epoch": 6.8690095846645365,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 4300
+    },
+    {
+      "epoch": 6.8706070287539935,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4301
+    },
+    {
+      "epoch": 6.872204472843451,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 4302
+    },
+    {
+      "epoch": 6.873801916932908,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4303
+    },
+    {
+      "epoch": 6.875399361022364,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 4304
+    },
+    {
+      "epoch": 6.876996805111821,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4305
+    },
+    {
+      "epoch": 6.878594249201278,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 4306
+    },
+    {
+      "epoch": 6.880191693290735,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1545,
+      "step": 4307
+    },
+    {
+      "epoch": 6.881789137380192,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4308
+    },
+    {
+      "epoch": 6.883386581469648,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4309
+    },
+    {
+      "epoch": 6.884984025559105,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4310
+    },
+    {
+      "epoch": 6.886581469648562,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 4311
+    },
+    {
+      "epoch": 6.888178913738019,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4312
+    },
+    {
+      "epoch": 6.8897763578274756,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 4313
+    },
+    {
+      "epoch": 6.891373801916933,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4314
+    },
+    {
+      "epoch": 6.89297124600639,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 4315
+    },
+    {
+      "epoch": 6.894568690095847,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 4316
+    },
+    {
+      "epoch": 6.896166134185304,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 4317
+    },
+    {
+      "epoch": 6.897763578274761,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4318
+    },
+    {
+      "epoch": 6.899361022364217,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 4319
+    },
+    {
+      "epoch": 6.900958466453674,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 4320
+    },
+    {
+      "epoch": 6.902555910543131,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 4321
+    },
+    {
+      "epoch": 6.904153354632588,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 4322
+    },
+    {
+      "epoch": 6.905750798722044,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4323
+    },
+    {
+      "epoch": 6.907348242811501,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 4324
+    },
+    {
+      "epoch": 6.9089456869009584,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 4325
+    },
+    {
+      "epoch": 6.9105431309904155,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1544,
+      "step": 4326
+    },
+    {
+      "epoch": 6.912140575079873,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 4327
+    },
+    {
+      "epoch": 6.913738019169329,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4328
+    },
+    {
+      "epoch": 6.915335463258786,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4329
+    },
+    {
+      "epoch": 6.916932907348243,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 4330
+    },
+    {
+      "epoch": 6.9185303514377,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4331
+    },
+    {
+      "epoch": 6.920127795527156,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4332
+    },
+    {
+      "epoch": 6.921725239616613,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4333
+    },
+    {
+      "epoch": 6.92332268370607,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 4334
+    },
+    {
+      "epoch": 6.924920127795527,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4335
+    },
+    {
+      "epoch": 6.926517571884984,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4336
+    },
+    {
+      "epoch": 6.928115015974441,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4337
+    },
+    {
+      "epoch": 6.9297124600638975,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 4338
+    },
+    {
+      "epoch": 6.931309904153355,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4339
+    },
+    {
+      "epoch": 6.932907348242812,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 4340
+    },
+    {
+      "epoch": 6.934504792332269,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4341
+    },
+    {
+      "epoch": 6.936102236421725,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1537,
+      "step": 4342
+    },
+    {
+      "epoch": 6.937699680511182,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4343
+    },
+    {
+      "epoch": 6.939297124600639,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 4344
+    },
+    {
+      "epoch": 6.940894568690096,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1632,
+      "step": 4345
+    },
+    {
+      "epoch": 6.942492012779553,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 4346
+    },
+    {
+      "epoch": 6.944089456869009,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4347
+    },
+    {
+      "epoch": 6.945686900958466,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.1608,
+      "step": 4348
+    },
+    {
+      "epoch": 6.947284345047923,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 4349
+    },
+    {
+      "epoch": 6.94888178913738,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 4350
+    },
+    {
+      "epoch": 6.950479233226837,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005,
+      "loss": 1.159,
+      "step": 4351
+    },
+    {
+      "epoch": 6.952076677316294,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 4352
+    },
+    {
+      "epoch": 6.953674121405751,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 4353
+    },
+    {
+      "epoch": 6.955271565495208,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4354
+    },
+    {
+      "epoch": 6.956869009584665,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4355
+    },
+    {
+      "epoch": 6.958466453674122,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 4356
+    },
+    {
+      "epoch": 6.960063897763578,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4357
+    },
+    {
+      "epoch": 6.961661341853035,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4358
+    },
+    {
+      "epoch": 6.963258785942492,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4359
+    },
+    {
+      "epoch": 6.964856230031949,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1491,
+      "step": 4360
+    },
+    {
+      "epoch": 6.966453674121405,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 4361
+    },
+    {
+      "epoch": 6.968051118210862,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4362
+    },
+    {
+      "epoch": 6.9696485623003195,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4363
+    },
+    {
+      "epoch": 6.9712460063897765,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4364
+    },
+    {
+      "epoch": 6.972843450479234,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 4365
+    },
+    {
+      "epoch": 6.97444089456869,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 4366
+    },
+    {
+      "epoch": 6.976038338658147,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 4367
+    },
+    {
+      "epoch": 6.977635782747604,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 4368
+    },
+    {
+      "epoch": 6.979233226837061,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4369
+    },
+    {
+      "epoch": 6.980830670926517,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4370
+    },
+    {
+      "epoch": 6.982428115015974,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 4371
+    },
+    {
+      "epoch": 6.984025559105431,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4372
+    },
+    {
+      "epoch": 6.985623003194888,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 4373
+    },
+    {
+      "epoch": 6.987220447284345,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 4374
+    },
+    {
+      "epoch": 6.988817891373802,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 4375
+    },
+    {
+      "epoch": 6.9904153354632586,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4376
+    },
+    {
+      "epoch": 6.992012779552716,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 4377
+    },
+    {
+      "epoch": 6.993610223642173,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 4378
+    },
+    {
+      "epoch": 6.99520766773163,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 4379
+    },
+    {
+      "epoch": 6.996805111821086,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 4380
+    },
+    {
+      "epoch": 6.998402555910543,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4381
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 4382
+    },
+    {
+      "epoch": 7.001597444089457,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1559,
+      "step": 4383
+    },
+    {
+      "epoch": 7.003194888178914,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4384
+    },
+    {
+      "epoch": 7.00479233226837,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 4385
+    },
+    {
+      "epoch": 7.006389776357827,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4386
+    },
+    {
+      "epoch": 7.007987220447284,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 4387
+    },
+    {
+      "epoch": 7.0095846645367414,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4388
+    },
+    {
+      "epoch": 7.0111821086261985,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4389
+    },
+    {
+      "epoch": 7.012779552715655,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 4390
+    },
+    {
+      "epoch": 7.014376996805112,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 4391
+    },
+    {
+      "epoch": 7.015974440894569,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 4392
+    },
+    {
+      "epoch": 7.017571884984026,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4393
+    },
+    {
+      "epoch": 7.019169329073482,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4394
+    },
+    {
+      "epoch": 7.020766773162939,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 4395
+    },
+    {
+      "epoch": 7.022364217252396,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 4396
+    },
+    {
+      "epoch": 7.023961661341853,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 4397
+    },
+    {
+      "epoch": 7.02555910543131,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 4398
+    },
+    {
+      "epoch": 7.027156549520766,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 4399
+    },
+    {
+      "epoch": 7.0287539936102235,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 4400
+    },
+    {
+      "epoch": 7.0303514376996805,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 4401
+    },
+    {
+      "epoch": 7.031948881789138,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 4402
+    },
+    {
+      "epoch": 7.033546325878595,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 4403
+    },
+    {
+      "epoch": 7.035143769968051,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 4404
+    },
+    {
+      "epoch": 7.036741214057508,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4405
+    },
+    {
+      "epoch": 7.038338658146965,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 4406
+    },
+    {
+      "epoch": 7.039936102236422,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 4407
+    },
+    {
+      "epoch": 7.041533546325879,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4408
+    },
+    {
+      "epoch": 7.043130990415335,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4409
+    },
+    {
+      "epoch": 7.044728434504792,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 4410
+    },
+    {
+      "epoch": 7.046325878594249,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 4411
+    },
+    {
+      "epoch": 7.047923322683706,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4412
+    },
+    {
+      "epoch": 7.0495207667731625,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 4413
+    },
+    {
+      "epoch": 7.05111821086262,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4414
+    },
+    {
+      "epoch": 7.052715654952077,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 4415
+    },
+    {
+      "epoch": 7.054313099041534,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4416
+    },
+    {
+      "epoch": 7.055910543130991,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4417
+    },
+    {
+      "epoch": 7.057507987220447,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4418
+    },
+    {
+      "epoch": 7.059105431309904,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 4419
+    },
+    {
+      "epoch": 7.060702875399361,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4420
+    },
+    {
+      "epoch": 7.062300319488818,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4421
+    },
+    {
+      "epoch": 7.063897763578275,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4422
+    },
+    {
+      "epoch": 7.065495207667731,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 4423
+    },
+    {
+      "epoch": 7.067092651757188,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4424
+    },
+    {
+      "epoch": 7.068690095846645,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 4425
+    },
+    {
+      "epoch": 7.0702875399361025,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 4426
+    },
+    {
+      "epoch": 7.0718849840255595,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4427
+    },
+    {
+      "epoch": 7.073482428115016,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 4428
+    },
+    {
+      "epoch": 7.075079872204473,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4429
+    },
+    {
+      "epoch": 7.07667731629393,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4430
+    },
+    {
+      "epoch": 7.078274760383387,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 4431
+    },
+    {
+      "epoch": 7.079872204472843,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 4432
+    },
+    {
+      "epoch": 7.0814696485623,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 4433
+    },
+    {
+      "epoch": 7.083067092651757,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 4434
+    },
+    {
+      "epoch": 7.084664536741214,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 4435
+    },
+    {
+      "epoch": 7.086261980830671,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4436
+    },
+    {
+      "epoch": 7.087859424920127,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1534,
+      "step": 4437
+    },
+    {
+      "epoch": 7.0894568690095845,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4438
+    },
+    {
+      "epoch": 7.0910543130990416,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 4439
+    },
+    {
+      "epoch": 7.092651757188499,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 4440
+    },
+    {
+      "epoch": 7.094249201277956,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4441
+    },
+    {
+      "epoch": 7.095846645367412,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 4442
+    },
+    {
+      "epoch": 7.097444089456869,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 4443
+    },
+    {
+      "epoch": 7.099041533546326,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4444
+    },
+    {
+      "epoch": 7.100638977635783,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4445
+    },
+    {
+      "epoch": 7.102236421725239,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 4446
+    },
+    {
+      "epoch": 7.103833865814696,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4447
+    },
+    {
+      "epoch": 7.105431309904153,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4448
+    },
+    {
+      "epoch": 7.10702875399361,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 4449
+    },
+    {
+      "epoch": 7.108626198083067,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 4450
+    },
+    {
+      "epoch": 7.110223642172524,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4451
+    },
+    {
+      "epoch": 7.111821086261981,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 4452
+    },
+    {
+      "epoch": 7.113418530351438,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 4453
+    },
+    {
+      "epoch": 7.115015974440895,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4454
+    },
+    {
+      "epoch": 7.116613418530352,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 4455
+    },
+    {
+      "epoch": 7.118210862619808,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 4456
+    },
+    {
+      "epoch": 7.119808306709265,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4457
+    },
+    {
+      "epoch": 7.121405750798722,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4458
+    },
+    {
+      "epoch": 7.123003194888179,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4459
+    },
+    {
+      "epoch": 7.124600638977636,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4460
+    },
+    {
+      "epoch": 7.126198083067092,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4461
+    },
+    {
+      "epoch": 7.127795527156549,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4462
+    },
+    {
+      "epoch": 7.1293929712460065,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4463
+    },
+    {
+      "epoch": 7.1309904153354635,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 4464
+    },
+    {
+      "epoch": 7.13258785942492,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 4465
+    },
+    {
+      "epoch": 7.134185303514377,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1565,
+      "step": 4466
+    },
+    {
+      "epoch": 7.135782747603834,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 4467
+    },
+    {
+      "epoch": 7.137380191693291,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 4468
+    },
+    {
+      "epoch": 7.138977635782748,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4469
+    },
+    {
+      "epoch": 7.140575079872204,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 4470
+    },
+    {
+      "epoch": 7.142172523961661,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4471
+    },
+    {
+      "epoch": 7.143769968051118,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4472
+    },
+    {
+      "epoch": 7.145367412140575,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4473
+    },
+    {
+      "epoch": 7.146964856230032,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 4474
+    },
+    {
+      "epoch": 7.1485623003194885,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 4475
+    },
+    {
+      "epoch": 7.1501597444089455,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 4476
+    },
+    {
+      "epoch": 7.151757188498403,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4477
+    },
+    {
+      "epoch": 7.15335463258786,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 4478
+    },
+    {
+      "epoch": 7.154952076677317,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 4479
+    },
+    {
+      "epoch": 7.156549520766773,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 4480
+    },
+    {
+      "epoch": 7.15814696485623,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 4481
+    },
+    {
+      "epoch": 7.159744408945687,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 4482
+    },
+    {
+      "epoch": 7.161341853035144,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4483
+    },
+    {
+      "epoch": 7.1629392971246,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 4484
+    },
+    {
+      "epoch": 7.164536741214057,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 4485
+    },
+    {
+      "epoch": 7.166134185303514,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 4486
+    },
+    {
+      "epoch": 7.167731629392971,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 4487
+    },
+    {
+      "epoch": 7.169329073482428,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 4488
+    },
+    {
+      "epoch": 7.170926517571885,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 4489
+    },
+    {
+      "epoch": 7.172523961661342,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4490
+    },
+    {
+      "epoch": 7.174121405750799,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 4491
+    },
+    {
+      "epoch": 7.175718849840256,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4492
+    },
+    {
+      "epoch": 7.177316293929713,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4493
+    },
+    {
+      "epoch": 7.178913738019169,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 4494
+    },
+    {
+      "epoch": 7.180511182108626,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 4495
+    },
+    {
+      "epoch": 7.182108626198083,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 4496
+    },
+    {
+      "epoch": 7.18370607028754,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4497
+    },
+    {
+      "epoch": 7.185303514376997,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 4498
+    },
+    {
+      "epoch": 7.186900958466453,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 4499
+    },
+    {
+      "epoch": 7.18849840255591,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 4500
+    },
+    {
+      "epoch": 7.1900958466453675,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 4501
+    },
+    {
+      "epoch": 7.1916932907348246,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 4502
+    },
+    {
+      "epoch": 7.193290734824281,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4503
+    },
+    {
+      "epoch": 7.194888178913738,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 4504
+    },
+    {
+      "epoch": 7.196485623003195,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4505
+    },
+    {
+      "epoch": 7.198083067092652,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4506
+    },
+    {
+      "epoch": 7.199680511182109,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4507
+    },
+    {
+      "epoch": 7.201277955271565,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4508
+    },
+    {
+      "epoch": 7.202875399361022,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4509
+    },
+    {
+      "epoch": 7.204472843450479,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 4510
+    },
+    {
+      "epoch": 7.206070287539936,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4511
+    },
+    {
+      "epoch": 7.207667731629393,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4512
+    },
+    {
+      "epoch": 7.2092651757188495,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4513
+    },
+    {
+      "epoch": 7.210862619808307,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4514
+    },
+    {
+      "epoch": 7.212460063897764,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 4515
+    },
+    {
+      "epoch": 7.214057507987221,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4516
+    },
+    {
+      "epoch": 7.215654952076678,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4517
+    },
+    {
+      "epoch": 7.217252396166134,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 4518
+    },
+    {
+      "epoch": 7.218849840255591,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4519
+    },
+    {
+      "epoch": 7.220447284345048,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 4520
+    },
+    {
+      "epoch": 7.222044728434505,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4521
+    },
+    {
+      "epoch": 7.223642172523961,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 4522
+    },
+    {
+      "epoch": 7.225239616613418,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 4523
+    },
+    {
+      "epoch": 7.226837060702875,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4524
+    },
+    {
+      "epoch": 7.228434504792332,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 4525
+    },
+    {
+      "epoch": 7.2300319488817895,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 4526
+    },
+    {
+      "epoch": 7.231629392971246,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 4527
+    },
+    {
+      "epoch": 7.233226837060703,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4528
+    },
+    {
+      "epoch": 7.23482428115016,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 4529
+    },
+    {
+      "epoch": 7.236421725239617,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4530
+    },
+    {
+      "epoch": 7.238019169329074,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4531
+    },
+    {
+      "epoch": 7.23961661341853,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 4532
+    },
+    {
+      "epoch": 7.241214057507987,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4533
+    },
+    {
+      "epoch": 7.242811501597444,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 4534
+    },
+    {
+      "epoch": 7.244408945686901,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4535
+    },
+    {
+      "epoch": 7.246006389776358,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 4536
+    },
+    {
+      "epoch": 7.247603833865814,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 4537
+    },
+    {
+      "epoch": 7.2492012779552715,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4538
+    },
+    {
+      "epoch": 7.2507987220447285,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4539
+    },
+    {
+      "epoch": 7.252396166134186,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 4540
+    },
+    {
+      "epoch": 7.253993610223642,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4541
+    },
+    {
+      "epoch": 7.255591054313099,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 4542
+    },
+    {
+      "epoch": 7.257188498402556,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4543
+    },
+    {
+      "epoch": 7.258785942492013,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4544
+    },
+    {
+      "epoch": 7.26038338658147,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 4545
+    },
+    {
+      "epoch": 7.261980830670926,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 4546
+    },
+    {
+      "epoch": 7.263578274760383,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4547
+    },
+    {
+      "epoch": 7.26517571884984,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4548
+    },
+    {
+      "epoch": 7.266773162939297,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4549
+    },
+    {
+      "epoch": 7.268370607028754,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 4550
+    },
+    {
+      "epoch": 7.2699680511182105,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4551
+    },
+    {
+      "epoch": 7.271565495207668,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 4552
+    },
+    {
+      "epoch": 7.273162939297125,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4553
+    },
+    {
+      "epoch": 7.274760383386582,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4554
+    },
+    {
+      "epoch": 7.276357827476039,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 4555
+    },
+    {
+      "epoch": 7.277955271565495,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4556
+    },
+    {
+      "epoch": 7.279552715654952,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 4557
+    },
+    {
+      "epoch": 7.281150159744409,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4558
+    },
+    {
+      "epoch": 7.282747603833866,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4559
+    },
+    {
+      "epoch": 7.284345047923322,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 4560
+    },
+    {
+      "epoch": 7.285942492012779,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 4561
+    },
+    {
+      "epoch": 7.287539936102236,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4562
+    },
+    {
+      "epoch": 7.289137380191693,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 4563
+    },
+    {
+      "epoch": 7.2907348242811505,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 4564
+    },
+    {
+      "epoch": 7.292332268370607,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4565
+    },
+    {
+      "epoch": 7.293929712460064,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 4566
+    },
+    {
+      "epoch": 7.295527156549521,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 4567
+    },
+    {
+      "epoch": 7.297124600638978,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 4568
+    },
+    {
+      "epoch": 7.298722044728435,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 4569
+    },
+    {
+      "epoch": 7.300319488817891,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 4570
+    },
+    {
+      "epoch": 7.301916932907348,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4571
+    },
+    {
+      "epoch": 7.303514376996805,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 4572
+    },
+    {
+      "epoch": 7.305111821086262,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4573
+    },
+    {
+      "epoch": 7.306709265175719,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 4574
+    },
+    {
+      "epoch": 7.3083067092651754,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4575
+    },
+    {
+      "epoch": 7.3099041533546325,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 4576
+    },
+    {
+      "epoch": 7.31150159744409,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4577
+    },
+    {
+      "epoch": 7.313099041533547,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 4578
+    },
+    {
+      "epoch": 7.314696485623003,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 4579
+    },
+    {
+      "epoch": 7.31629392971246,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 4580
+    },
+    {
+      "epoch": 7.317891373801917,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 4581
+    },
+    {
+      "epoch": 7.319488817891374,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 4582
+    },
+    {
+      "epoch": 7.321086261980831,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4583
+    },
+    {
+      "epoch": 7.322683706070287,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4584
+    },
+    {
+      "epoch": 7.324281150159744,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 4585
+    },
+    {
+      "epoch": 7.325878594249201,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4586
+    },
+    {
+      "epoch": 7.327476038338658,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 4587
+    },
+    {
+      "epoch": 7.329073482428115,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 4588
+    },
+    {
+      "epoch": 7.330670926517572,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4589
+    },
+    {
+      "epoch": 7.332268370607029,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 4590
+    },
+    {
+      "epoch": 7.333865814696486,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 4591
+    },
+    {
+      "epoch": 7.335463258785943,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 4592
+    },
+    {
+      "epoch": 7.3370607028754,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 4593
+    },
+    {
+      "epoch": 7.338658146964856,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4594
+    },
+    {
+      "epoch": 7.340255591054313,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 4595
+    },
+    {
+      "epoch": 7.34185303514377,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 4596
+    },
+    {
+      "epoch": 7.343450479233227,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 4597
+    },
+    {
+      "epoch": 7.345047923322683,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 4598
+    },
+    {
+      "epoch": 7.34664536741214,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 4599
+    },
+    {
+      "epoch": 7.348242811501597,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 4600
+    },
+    {
+      "epoch": 7.3498402555910545,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4601
+    },
+    {
+      "epoch": 7.3514376996805115,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 4602
+    },
+    {
+      "epoch": 7.353035143769968,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 4603
+    },
+    {
+      "epoch": 7.354632587859425,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 4604
+    },
+    {
+      "epoch": 7.356230031948882,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 4605
+    },
+    {
+      "epoch": 7.357827476038339,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4606
+    },
+    {
+      "epoch": 7.359424920127796,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 4607
+    },
+    {
+      "epoch": 7.361022364217252,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 4608
+    },
+    {
+      "epoch": 7.362619808306709,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4609
+    },
+    {
+      "epoch": 7.364217252396166,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 4610
+    },
+    {
+      "epoch": 7.365814696485623,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 4611
+    },
+    {
+      "epoch": 7.36741214057508,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 4612
+    },
+    {
+      "epoch": 7.3690095846645365,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 4613
+    },
+    {
+      "epoch": 7.3706070287539935,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 4614
+    },
+    {
+      "epoch": 7.372204472843451,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4615
+    },
+    {
+      "epoch": 7.373801916932908,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 4616
+    },
+    {
+      "epoch": 7.375399361022364,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4617
+    },
+    {
+      "epoch": 7.376996805111821,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 4618
+    },
+    {
+      "epoch": 7.378594249201278,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 4619
+    },
+    {
+      "epoch": 7.380191693290735,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4620
+    },
+    {
+      "epoch": 7.381789137380192,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 4621
+    },
+    {
+      "epoch": 7.383386581469648,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 4622
+    },
+    {
+      "epoch": 7.384984025559105,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4623
+    },
+    {
+      "epoch": 7.386581469648562,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4624
+    },
+    {
+      "epoch": 7.388178913738019,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4625
+    },
+    {
+      "epoch": 7.389776357827476,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 4626
+    },
+    {
+      "epoch": 7.391373801916933,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4627
+    },
+    {
+      "epoch": 7.39297124600639,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4628
+    },
+    {
+      "epoch": 7.394568690095847,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 4629
+    },
+    {
+      "epoch": 7.396166134185304,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 4630
+    },
+    {
+      "epoch": 7.397763578274761,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4631
+    },
+    {
+      "epoch": 7.399361022364217,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4632
+    },
+    {
+      "epoch": 7.400958466453674,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 4633
+    },
+    {
+      "epoch": 7.402555910543131,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4634
+    },
+    {
+      "epoch": 7.404153354632588,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4635
+    },
+    {
+      "epoch": 7.405750798722044,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 4636
+    },
+    {
+      "epoch": 7.407348242811501,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4637
+    },
+    {
+      "epoch": 7.4089456869009584,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 4638
+    },
+    {
+      "epoch": 7.4105431309904155,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4639
+    },
+    {
+      "epoch": 7.412140575079873,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 4640
+    },
+    {
+      "epoch": 7.413738019169329,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 4641
+    },
+    {
+      "epoch": 7.415335463258786,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 4642
+    },
+    {
+      "epoch": 7.416932907348243,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 4643
+    },
+    {
+      "epoch": 7.4185303514377,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 4644
+    },
+    {
+      "epoch": 7.420127795527157,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 4645
+    },
+    {
+      "epoch": 7.421725239616613,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4646
+    },
+    {
+      "epoch": 7.42332268370607,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4647
+    },
+    {
+      "epoch": 7.424920127795527,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 4648
+    },
+    {
+      "epoch": 7.426517571884984,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 4649
+    },
+    {
+      "epoch": 7.428115015974441,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1575,
+      "step": 4650
+    },
+    {
+      "epoch": 7.4297124600638975,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4651
+    },
+    {
+      "epoch": 7.431309904153355,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 4652
+    },
+    {
+      "epoch": 7.432907348242812,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 4653
+    },
+    {
+      "epoch": 7.434504792332269,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 4654
+    },
+    {
+      "epoch": 7.436102236421725,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 4655
+    },
+    {
+      "epoch": 7.437699680511182,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4656
+    },
+    {
+      "epoch": 7.439297124600639,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 4657
+    },
+    {
+      "epoch": 7.440894568690096,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4658
+    },
+    {
+      "epoch": 7.442492012779553,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 4659
+    },
+    {
+      "epoch": 7.444089456869009,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 4660
+    },
+    {
+      "epoch": 7.445686900958466,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4661
+    },
+    {
+      "epoch": 7.447284345047923,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4662
+    },
+    {
+      "epoch": 7.44888178913738,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4663
+    },
+    {
+      "epoch": 7.4504792332268375,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4664
+    },
+    {
+      "epoch": 7.452076677316294,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4665
+    },
+    {
+      "epoch": 7.453674121405751,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4666
+    },
+    {
+      "epoch": 7.455271565495208,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 4667
+    },
+    {
+      "epoch": 7.456869009584665,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 4668
+    },
+    {
+      "epoch": 7.458466453674122,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 4669
+    },
+    {
+      "epoch": 7.460063897763578,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 4670
+    },
+    {
+      "epoch": 7.461661341853035,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4671
+    },
+    {
+      "epoch": 7.463258785942492,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 4672
+    },
+    {
+      "epoch": 7.464856230031949,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 4673
+    },
+    {
+      "epoch": 7.466453674121405,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4674
+    },
+    {
+      "epoch": 7.468051118210862,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 4675
+    },
+    {
+      "epoch": 7.4696485623003195,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 4676
+    },
+    {
+      "epoch": 7.4712460063897765,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 4677
+    },
+    {
+      "epoch": 7.472843450479234,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 4678
+    },
+    {
+      "epoch": 7.47444089456869,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4679
+    },
+    {
+      "epoch": 7.476038338658147,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 4680
+    },
+    {
+      "epoch": 7.477635782747604,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 4681
+    },
+    {
+      "epoch": 7.479233226837061,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 4682
+    },
+    {
+      "epoch": 7.480830670926517,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 4683
+    },
+    {
+      "epoch": 7.482428115015974,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4684
+    },
+    {
+      "epoch": 7.484025559105431,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4685
+    },
+    {
+      "epoch": 7.485623003194888,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 4686
+    },
+    {
+      "epoch": 7.487220447284345,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4687
+    },
+    {
+      "epoch": 7.488817891373802,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 4688
+    },
+    {
+      "epoch": 7.4904153354632586,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4689
+    },
+    {
+      "epoch": 7.492012779552716,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4690
+    },
+    {
+      "epoch": 7.493610223642173,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 4691
+    },
+    {
+      "epoch": 7.49520766773163,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 4692
+    },
+    {
+      "epoch": 7.496805111821086,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 4693
+    },
+    {
+      "epoch": 7.498402555910543,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4694
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 4695
+    },
+    {
+      "epoch": 7.501597444089457,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 4696
+    },
+    {
+      "epoch": 7.503194888178914,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 4697
+    },
+    {
+      "epoch": 7.50479233226837,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4698
+    },
+    {
+      "epoch": 7.506389776357827,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4699
+    },
+    {
+      "epoch": 7.507987220447284,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 4700
+    },
+    {
+      "epoch": 7.5095846645367414,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4701
+    },
+    {
+      "epoch": 7.511182108626198,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4702
+    },
+    {
+      "epoch": 7.512779552715655,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 4703
+    },
+    {
+      "epoch": 7.514376996805112,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4704
+    },
+    {
+      "epoch": 7.515974440894569,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4705
+    },
+    {
+      "epoch": 7.517571884984026,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 4706
+    },
+    {
+      "epoch": 7.519169329073483,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 4707
+    },
+    {
+      "epoch": 7.520766773162939,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 4708
+    },
+    {
+      "epoch": 7.522364217252396,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 4709
+    },
+    {
+      "epoch": 7.523961661341853,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4710
+    },
+    {
+      "epoch": 7.52555910543131,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 4711
+    },
+    {
+      "epoch": 7.527156549520766,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 4712
+    },
+    {
+      "epoch": 7.5287539936102235,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4713
+    },
+    {
+      "epoch": 7.5303514376996805,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1561,
+      "step": 4714
+    },
+    {
+      "epoch": 7.531948881789138,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 4715
+    },
+    {
+      "epoch": 7.533546325878595,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 4716
+    },
+    {
+      "epoch": 7.535143769968051,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4717
+    },
+    {
+      "epoch": 7.536741214057508,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 4718
+    },
+    {
+      "epoch": 7.538338658146965,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 4719
+    },
+    {
+      "epoch": 7.539936102236422,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 4720
+    },
+    {
+      "epoch": 7.541533546325878,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4721
+    },
+    {
+      "epoch": 7.543130990415335,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 4722
+    },
+    {
+      "epoch": 7.544728434504792,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 4723
+    },
+    {
+      "epoch": 7.546325878594249,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 4724
+    },
+    {
+      "epoch": 7.547923322683706,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4725
+    },
+    {
+      "epoch": 7.549520766773163,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 4726
+    },
+    {
+      "epoch": 7.55111821086262,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 4727
+    },
+    {
+      "epoch": 7.552715654952077,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 4728
+    },
+    {
+      "epoch": 7.554313099041534,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 4729
+    },
+    {
+      "epoch": 7.555910543130991,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4730
+    },
+    {
+      "epoch": 7.557507987220447,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 4731
+    },
+    {
+      "epoch": 7.559105431309904,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 4732
+    },
+    {
+      "epoch": 7.560702875399361,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 4733
+    },
+    {
+      "epoch": 7.562300319488818,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4734
+    },
+    {
+      "epoch": 7.563897763578275,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 4735
+    },
+    {
+      "epoch": 7.565495207667731,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 4736
+    },
+    {
+      "epoch": 7.567092651757188,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1504,
+      "step": 4737
+    },
+    {
+      "epoch": 7.568690095846645,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4738
+    },
+    {
+      "epoch": 7.5702875399361025,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 4739
+    },
+    {
+      "epoch": 7.571884984025559,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4740
+    },
+    {
+      "epoch": 7.573482428115016,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 4741
+    },
+    {
+      "epoch": 7.575079872204473,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 4742
+    },
+    {
+      "epoch": 7.57667731629393,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 4743
+    },
+    {
+      "epoch": 7.578274760383387,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4744
+    },
+    {
+      "epoch": 7.579872204472844,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 4745
+    },
+    {
+      "epoch": 7.5814696485623,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4746
+    },
+    {
+      "epoch": 7.583067092651757,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 4747
+    },
+    {
+      "epoch": 7.584664536741214,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4748
+    },
+    {
+      "epoch": 7.586261980830671,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4749
+    },
+    {
+      "epoch": 7.587859424920127,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4750
+    },
+    {
+      "epoch": 7.5894568690095845,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4751
+    },
+    {
+      "epoch": 7.5910543130990416,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4752
+    },
+    {
+      "epoch": 7.592651757188499,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4753
+    },
+    {
+      "epoch": 7.594249201277956,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4754
+    },
+    {
+      "epoch": 7.595846645367412,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4755
+    },
+    {
+      "epoch": 7.597444089456869,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 4756
+    },
+    {
+      "epoch": 7.599041533546326,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4757
+    },
+    {
+      "epoch": 7.600638977635783,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 4758
+    },
+    {
+      "epoch": 7.602236421725239,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 4759
+    },
+    {
+      "epoch": 7.603833865814696,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 4760
+    },
+    {
+      "epoch": 7.605431309904153,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 4761
+    },
+    {
+      "epoch": 7.60702875399361,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 4762
+    },
+    {
+      "epoch": 7.608626198083067,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4763
+    },
+    {
+      "epoch": 7.6102236421725244,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4764
+    },
+    {
+      "epoch": 7.611821086261981,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 4765
+    },
+    {
+      "epoch": 7.613418530351438,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 4766
+    },
+    {
+      "epoch": 7.615015974440895,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4767
+    },
+    {
+      "epoch": 7.616613418530352,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 4768
+    },
+    {
+      "epoch": 7.618210862619808,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 4769
+    },
+    {
+      "epoch": 7.619808306709265,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 4770
+    },
+    {
+      "epoch": 7.621405750798722,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1541,
+      "step": 4771
+    },
+    {
+      "epoch": 7.623003194888179,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 4772
+    },
+    {
+      "epoch": 7.624600638977636,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4773
+    },
+    {
+      "epoch": 7.626198083067092,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 4774
+    },
+    {
+      "epoch": 7.627795527156549,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 4775
+    },
+    {
+      "epoch": 7.6293929712460065,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4776
+    },
+    {
+      "epoch": 7.6309904153354635,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 4777
+    },
+    {
+      "epoch": 7.63258785942492,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 4778
+    },
+    {
+      "epoch": 7.634185303514377,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 4779
+    },
+    {
+      "epoch": 7.635782747603834,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 4780
+    },
+    {
+      "epoch": 7.637380191693291,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 4781
+    },
+    {
+      "epoch": 7.638977635782748,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 4782
+    },
+    {
+      "epoch": 7.640575079872205,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4783
+    },
+    {
+      "epoch": 7.642172523961661,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 4784
+    },
+    {
+      "epoch": 7.643769968051118,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 4785
+    },
+    {
+      "epoch": 7.645367412140575,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 4786
+    },
+    {
+      "epoch": 7.646964856230032,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 4787
+    },
+    {
+      "epoch": 7.6485623003194885,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 4788
+    },
+    {
+      "epoch": 7.6501597444089455,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4789
+    },
+    {
+      "epoch": 7.651757188498403,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4790
+    },
+    {
+      "epoch": 7.65335463258786,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4791
+    },
+    {
+      "epoch": 7.654952076677317,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 4792
+    },
+    {
+      "epoch": 7.656549520766773,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 4793
+    },
+    {
+      "epoch": 7.65814696485623,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 4794
+    },
+    {
+      "epoch": 7.659744408945687,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 4795
+    },
+    {
+      "epoch": 7.661341853035144,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4796
+    },
+    {
+      "epoch": 7.6629392971246,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4797
+    },
+    {
+      "epoch": 7.664536741214057,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4798
+    },
+    {
+      "epoch": 7.666134185303514,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 4799
+    },
+    {
+      "epoch": 7.667731629392971,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 4800
+    },
+    {
+      "epoch": 7.669329073482428,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4801
+    },
+    {
+      "epoch": 7.6709265175718855,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 4802
+    },
+    {
+      "epoch": 7.672523961661342,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 4803
+    },
+    {
+      "epoch": 7.674121405750799,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 4804
+    },
+    {
+      "epoch": 7.675718849840256,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 4805
+    },
+    {
+      "epoch": 7.677316293929713,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4806
+    },
+    {
+      "epoch": 7.678913738019169,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 4807
+    },
+    {
+      "epoch": 7.680511182108626,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 4808
+    },
+    {
+      "epoch": 7.682108626198083,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4809
+    },
+    {
+      "epoch": 7.68370607028754,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 4810
+    },
+    {
+      "epoch": 7.685303514376997,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 4811
+    },
+    {
+      "epoch": 7.686900958466453,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 4812
+    },
+    {
+      "epoch": 7.68849840255591,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 4813
+    },
+    {
+      "epoch": 7.6900958466453675,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 4814
+    },
+    {
+      "epoch": 7.6916932907348246,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 4815
+    },
+    {
+      "epoch": 7.693290734824281,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 4816
+    },
+    {
+      "epoch": 7.694888178913738,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 4817
+    },
+    {
+      "epoch": 7.696485623003195,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 4818
+    },
+    {
+      "epoch": 7.698083067092652,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4819
+    },
+    {
+      "epoch": 7.699680511182109,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 4820
+    },
+    {
+      "epoch": 7.701277955271565,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 4821
+    },
+    {
+      "epoch": 7.702875399361022,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 4822
+    },
+    {
+      "epoch": 7.704472843450479,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 4823
+    },
+    {
+      "epoch": 7.706070287539936,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4824
+    },
+    {
+      "epoch": 7.707667731629393,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 4825
+    },
+    {
+      "epoch": 7.7092651757188495,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4826
+    },
+    {
+      "epoch": 7.710862619808307,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4827
+    },
+    {
+      "epoch": 7.712460063897764,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 4828
+    },
+    {
+      "epoch": 7.714057507987221,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 4829
+    },
+    {
+      "epoch": 7.715654952076678,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 4830
+    },
+    {
+      "epoch": 7.717252396166134,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 4831
+    },
+    {
+      "epoch": 7.718849840255591,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 4832
+    },
+    {
+      "epoch": 7.720447284345048,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.154,
+      "step": 4833
+    },
+    {
+      "epoch": 7.722044728434505,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4834
+    },
+    {
+      "epoch": 7.723642172523961,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 4835
+    },
+    {
+      "epoch": 7.725239616613418,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 4836
+    },
+    {
+      "epoch": 7.726837060702875,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 4837
+    },
+    {
+      "epoch": 7.728434504792332,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 4838
+    },
+    {
+      "epoch": 7.7300319488817895,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 4839
+    },
+    {
+      "epoch": 7.731629392971246,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 4840
+    },
+    {
+      "epoch": 7.733226837060703,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 4841
+    },
+    {
+      "epoch": 7.73482428115016,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 4842
+    },
+    {
+      "epoch": 7.736421725239617,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 4843
+    },
+    {
+      "epoch": 7.738019169329074,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4844
+    },
+    {
+      "epoch": 7.73961661341853,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 4845
+    },
+    {
+      "epoch": 7.741214057507987,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 4846
+    },
+    {
+      "epoch": 7.742811501597444,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 4847
+    },
+    {
+      "epoch": 7.744408945686901,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 4848
+    },
+    {
+      "epoch": 7.746006389776358,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 4849
+    },
+    {
+      "epoch": 7.747603833865814,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4850
+    },
+    {
+      "epoch": 7.7492012779552715,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1526,
+      "step": 4851
+    },
+    {
+      "epoch": 7.7507987220447285,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4852
+    },
+    {
+      "epoch": 7.752396166134186,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 4853
+    },
+    {
+      "epoch": 7.753993610223642,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 4854
+    },
+    {
+      "epoch": 7.755591054313099,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 4855
+    },
+    {
+      "epoch": 7.757188498402556,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 4856
+    },
+    {
+      "epoch": 7.758785942492013,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 4857
+    },
+    {
+      "epoch": 7.76038338658147,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 4858
+    },
+    {
+      "epoch": 7.761980830670926,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 4859
+    },
+    {
+      "epoch": 7.763578274760383,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 4860
+    },
+    {
+      "epoch": 7.76517571884984,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 4861
+    },
+    {
+      "epoch": 7.766773162939297,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 4862
+    },
+    {
+      "epoch": 7.768370607028754,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 4863
+    },
+    {
+      "epoch": 7.7699680511182105,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4864
+    },
+    {
+      "epoch": 7.771565495207668,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 4865
+    },
+    {
+      "epoch": 7.773162939297125,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 4866
+    },
+    {
+      "epoch": 7.774760383386582,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 4867
+    },
+    {
+      "epoch": 7.776357827476039,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4868
+    },
+    {
+      "epoch": 7.777955271565495,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 4869
+    },
+    {
+      "epoch": 7.779552715654952,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 4870
+    },
+    {
+      "epoch": 7.781150159744409,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 4871
+    },
+    {
+      "epoch": 7.782747603833866,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 4872
+    },
+    {
+      "epoch": 7.784345047923322,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 4873
+    },
+    {
+      "epoch": 7.785942492012779,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 4874
+    },
+    {
+      "epoch": 7.787539936102236,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 4875
+    },
+    {
+      "epoch": 7.789137380191693,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.152,
+      "step": 4876
+    },
+    {
+      "epoch": 7.7907348242811505,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 4877
+    },
+    {
+      "epoch": 7.792332268370607,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 4878
+    },
+    {
+      "epoch": 7.793929712460064,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4879
+    },
+    {
+      "epoch": 7.795527156549521,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4880
+    },
+    {
+      "epoch": 7.797124600638978,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 4881
+    },
+    {
+      "epoch": 7.798722044728435,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 4882
+    },
+    {
+      "epoch": 7.800319488817891,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 4883
+    },
+    {
+      "epoch": 7.801916932907348,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 4884
+    },
+    {
+      "epoch": 7.803514376996805,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 4885
+    },
+    {
+      "epoch": 7.805111821086262,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4886
+    },
+    {
+      "epoch": 7.806709265175719,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 4887
+    },
+    {
+      "epoch": 7.8083067092651754,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 4888
+    },
+    {
+      "epoch": 7.8099041533546325,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 4889
+    },
+    {
+      "epoch": 7.81150159744409,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 4890
+    },
+    {
+      "epoch": 7.813099041533547,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 4891
+    },
+    {
+      "epoch": 7.814696485623003,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 4892
+    },
+    {
+      "epoch": 7.81629392971246,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 4893
+    },
+    {
+      "epoch": 7.817891373801917,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 4894
+    },
+    {
+      "epoch": 7.819488817891374,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 4895
+    },
+    {
+      "epoch": 7.821086261980831,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 4896
+    },
+    {
+      "epoch": 7.822683706070287,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 4897
+    },
+    {
+      "epoch": 7.824281150159744,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 4898
+    },
+    {
+      "epoch": 7.825878594249201,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 4899
+    },
+    {
+      "epoch": 7.827476038338658,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 4900
+    },
+    {
+      "epoch": 7.8290734824281145,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 4901
+    },
+    {
+      "epoch": 7.830670926517572,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4902
+    },
+    {
+      "epoch": 7.832268370607029,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4903
+    },
+    {
+      "epoch": 7.833865814696486,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 4904
+    },
+    {
+      "epoch": 7.835463258785943,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 4905
+    },
+    {
+      "epoch": 7.8370607028754,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4906
+    },
+    {
+      "epoch": 7.838658146964856,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 4907
+    },
+    {
+      "epoch": 7.840255591054313,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 4908
+    },
+    {
+      "epoch": 7.84185303514377,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4909
+    },
+    {
+      "epoch": 7.843450479233227,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 4910
+    },
+    {
+      "epoch": 7.845047923322683,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 4911
+    },
+    {
+      "epoch": 7.84664536741214,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 4912
+    },
+    {
+      "epoch": 7.848242811501597,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 4913
+    },
+    {
+      "epoch": 7.8498402555910545,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 4914
+    },
+    {
+      "epoch": 7.8514376996805115,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 4915
+    },
+    {
+      "epoch": 7.853035143769968,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 4916
+    },
+    {
+      "epoch": 7.854632587859425,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 4917
+    },
+    {
+      "epoch": 7.856230031948882,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 4918
+    },
+    {
+      "epoch": 7.857827476038339,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 4919
+    },
+    {
+      "epoch": 7.859424920127795,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 4920
+    },
+    {
+      "epoch": 7.861022364217252,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 4921
+    },
+    {
+      "epoch": 7.862619808306709,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 4922
+    },
+    {
+      "epoch": 7.864217252396166,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 4923
+    },
+    {
+      "epoch": 7.865814696485623,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 4924
+    },
+    {
+      "epoch": 7.86741214057508,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 4925
+    },
+    {
+      "epoch": 7.8690095846645365,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 4926
+    },
+    {
+      "epoch": 7.8706070287539935,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4927
+    },
+    {
+      "epoch": 7.872204472843451,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4928
+    },
+    {
+      "epoch": 7.873801916932908,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 4929
+    },
+    {
+      "epoch": 7.875399361022364,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 4930
+    },
+    {
+      "epoch": 7.876996805111821,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 4931
+    },
+    {
+      "epoch": 7.878594249201278,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 4932
+    },
+    {
+      "epoch": 7.880191693290735,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 4933
+    },
+    {
+      "epoch": 7.881789137380192,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 4934
+    },
+    {
+      "epoch": 7.883386581469648,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 4935
+    },
+    {
+      "epoch": 7.884984025559105,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 4936
+    },
+    {
+      "epoch": 7.886581469648562,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 4937
+    },
+    {
+      "epoch": 7.888178913738019,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 4938
+    },
+    {
+      "epoch": 7.8897763578274756,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4939
+    },
+    {
+      "epoch": 7.891373801916933,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 4940
+    },
+    {
+      "epoch": 7.89297124600639,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 4941
+    },
+    {
+      "epoch": 7.894568690095847,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 4942
+    },
+    {
+      "epoch": 7.896166134185304,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 4943
+    },
+    {
+      "epoch": 7.897763578274761,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 4944
+    },
+    {
+      "epoch": 7.899361022364217,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 4945
+    },
+    {
+      "epoch": 7.900958466453674,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 4946
+    },
+    {
+      "epoch": 7.902555910543131,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 4947
+    },
+    {
+      "epoch": 7.904153354632588,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 4948
+    },
+    {
+      "epoch": 7.905750798722044,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 4949
+    },
+    {
+      "epoch": 7.907348242811501,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 4950
+    },
+    {
+      "epoch": 7.9089456869009584,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 4951
+    },
+    {
+      "epoch": 7.9105431309904155,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 4952
+    },
+    {
+      "epoch": 7.912140575079873,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 4953
+    },
+    {
+      "epoch": 7.913738019169329,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 4954
+    },
+    {
+      "epoch": 7.915335463258786,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 4955
+    },
+    {
+      "epoch": 7.916932907348243,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 4956
+    },
+    {
+      "epoch": 7.9185303514377,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 4957
+    },
+    {
+      "epoch": 7.920127795527156,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 4958
+    },
+    {
+      "epoch": 7.921725239616613,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 4959
+    },
+    {
+      "epoch": 7.92332268370607,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 4960
+    },
+    {
+      "epoch": 7.924920127795527,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 4961
+    },
+    {
+      "epoch": 7.926517571884984,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 4962
+    },
+    {
+      "epoch": 7.928115015974441,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 4963
+    },
+    {
+      "epoch": 7.9297124600638975,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 4964
+    },
+    {
+      "epoch": 7.931309904153355,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 4965
+    },
+    {
+      "epoch": 7.932907348242812,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 4966
+    },
+    {
+      "epoch": 7.934504792332269,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 4967
+    },
+    {
+      "epoch": 7.936102236421725,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 4968
+    },
+    {
+      "epoch": 7.937699680511182,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 4969
+    },
+    {
+      "epoch": 7.939297124600639,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 4970
+    },
+    {
+      "epoch": 7.940894568690096,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4971
+    },
+    {
+      "epoch": 7.942492012779553,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 4972
+    },
+    {
+      "epoch": 7.944089456869009,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 4973
+    },
+    {
+      "epoch": 7.945686900958466,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 4974
+    },
+    {
+      "epoch": 7.947284345047923,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 4975
+    },
+    {
+      "epoch": 7.94888178913738,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 4976
+    },
+    {
+      "epoch": 7.950479233226837,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 4977
+    },
+    {
+      "epoch": 7.952076677316294,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 4978
+    },
+    {
+      "epoch": 7.953674121405751,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 4979
+    },
+    {
+      "epoch": 7.955271565495208,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 4980
+    },
+    {
+      "epoch": 7.956869009584665,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 4981
+    },
+    {
+      "epoch": 7.958466453674122,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 4982
+    },
+    {
+      "epoch": 7.960063897763578,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 4983
+    },
+    {
+      "epoch": 7.961661341853035,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 4984
+    },
+    {
+      "epoch": 7.963258785942492,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 4985
+    },
+    {
+      "epoch": 7.964856230031949,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 4986
+    },
+    {
+      "epoch": 7.966453674121405,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 4987
+    },
+    {
+      "epoch": 7.968051118210862,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 4988
+    },
+    {
+      "epoch": 7.9696485623003195,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 4989
+    },
+    {
+      "epoch": 7.9712460063897765,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 4990
+    },
+    {
+      "epoch": 7.972843450479234,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 4991
+    },
+    {
+      "epoch": 7.97444089456869,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1529,
+      "step": 4992
+    },
+    {
+      "epoch": 7.976038338658147,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 4993
+    },
+    {
+      "epoch": 7.977635782747604,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 4994
+    },
+    {
+      "epoch": 7.979233226837061,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 4995
+    },
+    {
+      "epoch": 7.980830670926517,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 4996
+    },
+    {
+      "epoch": 7.982428115015974,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 4997
+    },
+    {
+      "epoch": 7.984025559105431,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 4998
+    },
+    {
+      "epoch": 7.985623003194888,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 4999
+    },
+    {
+      "epoch": 7.987220447284345,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 5000
+    },
+    {
+      "epoch": 7.988817891373802,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 5001
+    },
+    {
+      "epoch": 7.9904153354632586,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5002
+    },
+    {
+      "epoch": 7.992012779552716,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 5003
+    },
+    {
+      "epoch": 7.993610223642173,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 5004
+    },
+    {
+      "epoch": 7.99520766773163,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 5005
+    },
+    {
+      "epoch": 7.996805111821086,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5006
+    },
+    {
+      "epoch": 7.998402555910543,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 5007
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5008
+    },
+    {
+      "epoch": 8.001597444089457,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 5009
+    },
+    {
+      "epoch": 8.003194888178914,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 5010
+    },
+    {
+      "epoch": 8.004792332268371,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5011
+    },
+    {
+      "epoch": 8.006389776357828,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 5012
+    },
+    {
+      "epoch": 8.007987220447284,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5013
+    },
+    {
+      "epoch": 8.00958466453674,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 5014
+    },
+    {
+      "epoch": 8.011182108626198,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 5015
+    },
+    {
+      "epoch": 8.012779552715655,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5016
+    },
+    {
+      "epoch": 8.014376996805112,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 5017
+    },
+    {
+      "epoch": 8.015974440894569,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5018
+    },
+    {
+      "epoch": 8.017571884984026,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 5019
+    },
+    {
+      "epoch": 8.019169329073483,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 5020
+    },
+    {
+      "epoch": 8.02076677316294,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 5021
+    },
+    {
+      "epoch": 8.022364217252397,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5022
+    },
+    {
+      "epoch": 8.023961661341852,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 5023
+    },
+    {
+      "epoch": 8.02555910543131,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 5024
+    },
+    {
+      "epoch": 8.027156549520766,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 5025
+    },
+    {
+      "epoch": 8.028753993610223,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 5026
+    },
+    {
+      "epoch": 8.03035143769968,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 5027
+    },
+    {
+      "epoch": 8.031948881789138,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5028
+    },
+    {
+      "epoch": 8.033546325878595,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5029
+    },
+    {
+      "epoch": 8.035143769968052,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 5030
+    },
+    {
+      "epoch": 8.036741214057509,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 5031
+    },
+    {
+      "epoch": 8.038338658146964,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 5032
+    },
+    {
+      "epoch": 8.039936102236421,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 5033
+    },
+    {
+      "epoch": 8.041533546325878,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5034
+    },
+    {
+      "epoch": 8.043130990415335,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5035
+    },
+    {
+      "epoch": 8.044728434504792,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 5036
+    },
+    {
+      "epoch": 8.04632587859425,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5037
+    },
+    {
+      "epoch": 8.047923322683706,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 5038
+    },
+    {
+      "epoch": 8.049520766773163,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5039
+    },
+    {
+      "epoch": 8.05111821086262,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5040
+    },
+    {
+      "epoch": 8.052715654952078,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 5041
+    },
+    {
+      "epoch": 8.054313099041533,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5042
+    },
+    {
+      "epoch": 8.05591054313099,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5043
+    },
+    {
+      "epoch": 8.057507987220447,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1519,
+      "step": 5044
+    },
+    {
+      "epoch": 8.059105431309904,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 5045
+    },
+    {
+      "epoch": 8.060702875399361,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 5046
+    },
+    {
+      "epoch": 8.062300319488818,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 5047
+    },
+    {
+      "epoch": 8.063897763578275,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 5048
+    },
+    {
+      "epoch": 8.065495207667732,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5049
+    },
+    {
+      "epoch": 8.06709265175719,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 5050
+    },
+    {
+      "epoch": 8.068690095846645,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 5051
+    },
+    {
+      "epoch": 8.070287539936102,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5052
+    },
+    {
+      "epoch": 8.071884984025559,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5053
+    },
+    {
+      "epoch": 8.073482428115016,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5054
+    },
+    {
+      "epoch": 8.075079872204473,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 5055
+    },
+    {
+      "epoch": 8.07667731629393,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5056
+    },
+    {
+      "epoch": 8.078274760383387,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 5057
+    },
+    {
+      "epoch": 8.079872204472844,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5058
+    },
+    {
+      "epoch": 8.081469648562301,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5059
+    },
+    {
+      "epoch": 8.083067092651758,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1472,
+      "step": 5060
+    },
+    {
+      "epoch": 8.084664536741213,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5061
+    },
+    {
+      "epoch": 8.08626198083067,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 5062
+    },
+    {
+      "epoch": 8.087859424920127,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 5063
+    },
+    {
+      "epoch": 8.089456869009584,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 5064
+    },
+    {
+      "epoch": 8.091054313099042,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5065
+    },
+    {
+      "epoch": 8.092651757188499,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 5066
+    },
+    {
+      "epoch": 8.094249201277956,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5067
+    },
+    {
+      "epoch": 8.095846645367413,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 5068
+    },
+    {
+      "epoch": 8.09744408945687,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5069
+    },
+    {
+      "epoch": 8.099041533546325,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 5070
+    },
+    {
+      "epoch": 8.100638977635782,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 5071
+    },
+    {
+      "epoch": 8.10223642172524,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5072
+    },
+    {
+      "epoch": 8.103833865814696,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5073
+    },
+    {
+      "epoch": 8.105431309904153,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 5074
+    },
+    {
+      "epoch": 8.10702875399361,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5075
+    },
+    {
+      "epoch": 8.108626198083067,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5076
+    },
+    {
+      "epoch": 8.110223642172524,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 5077
+    },
+    {
+      "epoch": 8.111821086261982,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5078
+    },
+    {
+      "epoch": 8.113418530351439,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5079
+    },
+    {
+      "epoch": 8.115015974440894,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5080
+    },
+    {
+      "epoch": 8.116613418530351,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5081
+    },
+    {
+      "epoch": 8.118210862619808,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 5082
+    },
+    {
+      "epoch": 8.119808306709265,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 5083
+    },
+    {
+      "epoch": 8.121405750798722,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 5084
+    },
+    {
+      "epoch": 8.12300319488818,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5085
+    },
+    {
+      "epoch": 8.124600638977636,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5086
+    },
+    {
+      "epoch": 8.126198083067093,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 5087
+    },
+    {
+      "epoch": 8.12779552715655,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 5088
+    },
+    {
+      "epoch": 8.129392971246006,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 5089
+    },
+    {
+      "epoch": 8.130990415335463,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5090
+    },
+    {
+      "epoch": 8.13258785942492,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 5091
+    },
+    {
+      "epoch": 8.134185303514377,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 5092
+    },
+    {
+      "epoch": 8.135782747603834,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 5093
+    },
+    {
+      "epoch": 8.13738019169329,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 5094
+    },
+    {
+      "epoch": 8.138977635782748,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5095
+    },
+    {
+      "epoch": 8.140575079872205,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 5096
+    },
+    {
+      "epoch": 8.142172523961662,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 5097
+    },
+    {
+      "epoch": 8.143769968051119,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 5098
+    },
+    {
+      "epoch": 8.145367412140574,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5099
+    },
+    {
+      "epoch": 8.146964856230031,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 5100
+    },
+    {
+      "epoch": 8.148562300319488,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 5101
+    },
+    {
+      "epoch": 8.150159744408946,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 5102
+    },
+    {
+      "epoch": 8.151757188498403,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 5103
+    },
+    {
+      "epoch": 8.15335463258786,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5104
+    },
+    {
+      "epoch": 8.154952076677317,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5105
+    },
+    {
+      "epoch": 8.156549520766774,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 5106
+    },
+    {
+      "epoch": 8.15814696485623,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 5107
+    },
+    {
+      "epoch": 8.159744408945686,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 5108
+    },
+    {
+      "epoch": 8.161341853035143,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 5109
+    },
+    {
+      "epoch": 8.1629392971246,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5110
+    },
+    {
+      "epoch": 8.164536741214057,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 5111
+    },
+    {
+      "epoch": 8.166134185303514,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5112
+    },
+    {
+      "epoch": 8.167731629392971,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 5113
+    },
+    {
+      "epoch": 8.169329073482428,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 5114
+    },
+    {
+      "epoch": 8.170926517571885,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5115
+    },
+    {
+      "epoch": 8.172523961661343,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5116
+    },
+    {
+      "epoch": 8.1741214057508,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 5117
+    },
+    {
+      "epoch": 8.175718849840255,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 5118
+    },
+    {
+      "epoch": 8.177316293929712,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5119
+    },
+    {
+      "epoch": 8.178913738019169,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 5120
+    },
+    {
+      "epoch": 8.180511182108626,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 5121
+    },
+    {
+      "epoch": 8.182108626198083,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 5122
+    },
+    {
+      "epoch": 8.18370607028754,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5123
+    },
+    {
+      "epoch": 8.185303514376997,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 5124
+    },
+    {
+      "epoch": 8.186900958466454,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 5125
+    },
+    {
+      "epoch": 8.188498402555911,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 5126
+    },
+    {
+      "epoch": 8.190095846645367,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 5127
+    },
+    {
+      "epoch": 8.191693290734824,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 5128
+    },
+    {
+      "epoch": 8.19329073482428,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5129
+    },
+    {
+      "epoch": 8.194888178913738,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 5130
+    },
+    {
+      "epoch": 8.196485623003195,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 5131
+    },
+    {
+      "epoch": 8.198083067092652,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 5132
+    },
+    {
+      "epoch": 8.199680511182109,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5133
+    },
+    {
+      "epoch": 8.201277955271566,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5134
+    },
+    {
+      "epoch": 8.202875399361023,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 5135
+    },
+    {
+      "epoch": 8.204472843450478,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5136
+    },
+    {
+      "epoch": 8.206070287539935,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5137
+    },
+    {
+      "epoch": 8.207667731629392,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5138
+    },
+    {
+      "epoch": 8.20926517571885,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5139
+    },
+    {
+      "epoch": 8.210862619808307,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 5140
+    },
+    {
+      "epoch": 8.212460063897764,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5141
+    },
+    {
+      "epoch": 8.21405750798722,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 5142
+    },
+    {
+      "epoch": 8.215654952076678,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5143
+    },
+    {
+      "epoch": 8.217252396166135,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 5144
+    },
+    {
+      "epoch": 8.218849840255592,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 5145
+    },
+    {
+      "epoch": 8.220447284345047,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 5146
+    },
+    {
+      "epoch": 8.222044728434504,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5147
+    },
+    {
+      "epoch": 8.223642172523961,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5148
+    },
+    {
+      "epoch": 8.225239616613418,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 5149
+    },
+    {
+      "epoch": 8.226837060702875,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 5150
+    },
+    {
+      "epoch": 8.228434504792332,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 5151
+    },
+    {
+      "epoch": 8.23003194888179,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5152
+    },
+    {
+      "epoch": 8.231629392971247,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5153
+    },
+    {
+      "epoch": 8.233226837060704,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 5154
+    },
+    {
+      "epoch": 8.23482428115016,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 5155
+    },
+    {
+      "epoch": 8.236421725239616,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 5156
+    },
+    {
+      "epoch": 8.238019169329073,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5157
+    },
+    {
+      "epoch": 8.23961661341853,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 5158
+    },
+    {
+      "epoch": 8.241214057507987,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 5159
+    },
+    {
+      "epoch": 8.242811501597444,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 5160
+    },
+    {
+      "epoch": 8.244408945686901,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5161
+    },
+    {
+      "epoch": 8.246006389776358,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 5162
+    },
+    {
+      "epoch": 8.247603833865815,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5163
+    },
+    {
+      "epoch": 8.249201277955272,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 5164
+    },
+    {
+      "epoch": 8.250798722044728,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 5165
+    },
+    {
+      "epoch": 8.252396166134185,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 5166
+    },
+    {
+      "epoch": 8.253993610223642,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5167
+    },
+    {
+      "epoch": 8.255591054313099,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 5168
+    },
+    {
+      "epoch": 8.257188498402556,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 5169
+    },
+    {
+      "epoch": 8.258785942492013,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5170
+    },
+    {
+      "epoch": 8.26038338658147,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 5171
+    },
+    {
+      "epoch": 8.261980830670927,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5172
+    },
+    {
+      "epoch": 8.263578274760384,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 5173
+    },
+    {
+      "epoch": 8.26517571884984,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 5174
+    },
+    {
+      "epoch": 8.266773162939296,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 5175
+    },
+    {
+      "epoch": 8.268370607028753,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5176
+    },
+    {
+      "epoch": 8.26996805111821,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 5177
+    },
+    {
+      "epoch": 8.271565495207668,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5178
+    },
+    {
+      "epoch": 8.273162939297125,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 5179
+    },
+    {
+      "epoch": 8.274760383386582,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5180
+    },
+    {
+      "epoch": 8.276357827476039,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 5181
+    },
+    {
+      "epoch": 8.277955271565496,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1549,
+      "step": 5182
+    },
+    {
+      "epoch": 8.279552715654953,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5183
+    },
+    {
+      "epoch": 8.281150159744408,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 5184
+    },
+    {
+      "epoch": 8.282747603833865,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 5185
+    },
+    {
+      "epoch": 8.284345047923322,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 5186
+    },
+    {
+      "epoch": 8.28594249201278,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 5187
+    },
+    {
+      "epoch": 8.287539936102236,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 5188
+    },
+    {
+      "epoch": 8.289137380191693,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 5189
+    },
+    {
+      "epoch": 8.29073482428115,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 5190
+    },
+    {
+      "epoch": 8.292332268370608,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5191
+    },
+    {
+      "epoch": 8.293929712460065,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5192
+    },
+    {
+      "epoch": 8.295527156549522,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5193
+    },
+    {
+      "epoch": 8.297124600638977,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5194
+    },
+    {
+      "epoch": 8.298722044728434,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 5195
+    },
+    {
+      "epoch": 8.300319488817891,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5196
+    },
+    {
+      "epoch": 8.301916932907348,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 5197
+    },
+    {
+      "epoch": 8.303514376996805,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5198
+    },
+    {
+      "epoch": 8.305111821086262,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 5199
+    },
+    {
+      "epoch": 8.30670926517572,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 5200
+    },
+    {
+      "epoch": 8.308306709265176,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5201
+    },
+    {
+      "epoch": 8.309904153354633,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 5202
+    },
+    {
+      "epoch": 8.311501597444089,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5203
+    },
+    {
+      "epoch": 8.313099041533546,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 5204
+    },
+    {
+      "epoch": 8.314696485623003,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 5205
+    },
+    {
+      "epoch": 8.31629392971246,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 5206
+    },
+    {
+      "epoch": 8.317891373801917,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 5207
+    },
+    {
+      "epoch": 8.319488817891374,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5208
+    },
+    {
+      "epoch": 8.321086261980831,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5209
+    },
+    {
+      "epoch": 8.322683706070288,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 5210
+    },
+    {
+      "epoch": 8.324281150159745,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5211
+    },
+    {
+      "epoch": 8.3258785942492,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 5212
+    },
+    {
+      "epoch": 8.327476038338657,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5213
+    },
+    {
+      "epoch": 8.329073482428115,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 5214
+    },
+    {
+      "epoch": 8.330670926517572,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5215
+    },
+    {
+      "epoch": 8.332268370607029,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 5216
+    },
+    {
+      "epoch": 8.333865814696486,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 5217
+    },
+    {
+      "epoch": 8.335463258785943,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 5218
+    },
+    {
+      "epoch": 8.3370607028754,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 5219
+    },
+    {
+      "epoch": 8.338658146964857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 5220
+    },
+    {
+      "epoch": 8.340255591054314,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5221
+    },
+    {
+      "epoch": 8.34185303514377,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 5222
+    },
+    {
+      "epoch": 8.343450479233226,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 5223
+    },
+    {
+      "epoch": 8.345047923322683,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 5224
+    },
+    {
+      "epoch": 8.34664536741214,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5225
+    },
+    {
+      "epoch": 8.348242811501597,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 5226
+    },
+    {
+      "epoch": 8.349840255591054,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 5227
+    },
+    {
+      "epoch": 8.351437699680512,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 5228
+    },
+    {
+      "epoch": 8.353035143769969,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 5229
+    },
+    {
+      "epoch": 8.354632587859426,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 5230
+    },
+    {
+      "epoch": 8.356230031948883,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5231
+    },
+    {
+      "epoch": 8.357827476038338,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 5232
+    },
+    {
+      "epoch": 8.359424920127795,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5233
+    },
+    {
+      "epoch": 8.361022364217252,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5234
+    },
+    {
+      "epoch": 8.36261980830671,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 5235
+    },
+    {
+      "epoch": 8.364217252396166,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 5236
+    },
+    {
+      "epoch": 8.365814696485623,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 5237
+    },
+    {
+      "epoch": 8.36741214057508,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5238
+    },
+    {
+      "epoch": 8.369009584664537,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5239
+    },
+    {
+      "epoch": 8.370607028753994,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 5240
+    },
+    {
+      "epoch": 8.37220447284345,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5241
+    },
+    {
+      "epoch": 8.373801916932907,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 5242
+    },
+    {
+      "epoch": 8.375399361022364,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 5243
+    },
+    {
+      "epoch": 8.37699680511182,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5244
+    },
+    {
+      "epoch": 8.378594249201278,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 5245
+    },
+    {
+      "epoch": 8.380191693290735,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5246
+    },
+    {
+      "epoch": 8.381789137380192,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5247
+    },
+    {
+      "epoch": 8.383386581469649,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 5248
+    },
+    {
+      "epoch": 8.384984025559106,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 5249
+    },
+    {
+      "epoch": 8.386581469648561,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 5250
+    },
+    {
+      "epoch": 8.388178913738018,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5251
+    },
+    {
+      "epoch": 8.389776357827476,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 5252
+    },
+    {
+      "epoch": 8.391373801916933,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5253
+    },
+    {
+      "epoch": 8.39297124600639,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 5254
+    },
+    {
+      "epoch": 8.394568690095847,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 5255
+    },
+    {
+      "epoch": 8.396166134185304,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5256
+    },
+    {
+      "epoch": 8.39776357827476,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 5257
+    },
+    {
+      "epoch": 8.399361022364218,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 5258
+    },
+    {
+      "epoch": 8.400958466453675,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 5259
+    },
+    {
+      "epoch": 8.40255591054313,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 5260
+    },
+    {
+      "epoch": 8.404153354632587,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5261
+    },
+    {
+      "epoch": 8.405750798722044,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5262
+    },
+    {
+      "epoch": 8.407348242811501,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 5263
+    },
+    {
+      "epoch": 8.408945686900958,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 5264
+    },
+    {
+      "epoch": 8.410543130990416,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 5265
+    },
+    {
+      "epoch": 8.412140575079873,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 5266
+    },
+    {
+      "epoch": 8.41373801916933,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 5267
+    },
+    {
+      "epoch": 8.415335463258787,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5268
+    },
+    {
+      "epoch": 8.416932907348242,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 5269
+    },
+    {
+      "epoch": 8.418530351437699,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5270
+    },
+    {
+      "epoch": 8.420127795527156,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 5271
+    },
+    {
+      "epoch": 8.421725239616613,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 5272
+    },
+    {
+      "epoch": 8.42332268370607,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5273
+    },
+    {
+      "epoch": 8.424920127795527,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1479,
+      "step": 5274
+    },
+    {
+      "epoch": 8.426517571884984,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 5275
+    },
+    {
+      "epoch": 8.428115015974441,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5276
+    },
+    {
+      "epoch": 8.429712460063898,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 5277
+    },
+    {
+      "epoch": 8.431309904153355,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5278
+    },
+    {
+      "epoch": 8.43290734824281,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 5279
+    },
+    {
+      "epoch": 8.434504792332268,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 5280
+    },
+    {
+      "epoch": 8.436102236421725,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 5281
+    },
+    {
+      "epoch": 8.437699680511182,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 5282
+    },
+    {
+      "epoch": 8.439297124600639,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5283
+    },
+    {
+      "epoch": 8.440894568690096,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 5284
+    },
+    {
+      "epoch": 8.442492012779553,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5285
+    },
+    {
+      "epoch": 8.44408945686901,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5286
+    },
+    {
+      "epoch": 8.445686900958467,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1553,
+      "step": 5287
+    },
+    {
+      "epoch": 8.447284345047922,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 5288
+    },
+    {
+      "epoch": 8.44888178913738,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 5289
+    },
+    {
+      "epoch": 8.450479233226837,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 5290
+    },
+    {
+      "epoch": 8.452076677316294,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 5291
+    },
+    {
+      "epoch": 8.45367412140575,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5292
+    },
+    {
+      "epoch": 8.455271565495208,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 5293
+    },
+    {
+      "epoch": 8.456869009584665,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 5294
+    },
+    {
+      "epoch": 8.458466453674122,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 5295
+    },
+    {
+      "epoch": 8.460063897763579,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5296
+    },
+    {
+      "epoch": 8.461661341853036,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 5297
+    },
+    {
+      "epoch": 8.463258785942491,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5298
+    },
+    {
+      "epoch": 8.464856230031948,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5299
+    },
+    {
+      "epoch": 8.466453674121405,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 5300
+    },
+    {
+      "epoch": 8.468051118210862,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 5301
+    },
+    {
+      "epoch": 8.46964856230032,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 5302
+    },
+    {
+      "epoch": 8.471246006389777,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5303
+    },
+    {
+      "epoch": 8.472843450479234,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 5304
+    },
+    {
+      "epoch": 8.47444089456869,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 5305
+    },
+    {
+      "epoch": 8.476038338658148,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 5306
+    },
+    {
+      "epoch": 8.477635782747603,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 5307
+    },
+    {
+      "epoch": 8.47923322683706,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 5308
+    },
+    {
+      "epoch": 8.480830670926517,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 5309
+    },
+    {
+      "epoch": 8.482428115015974,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 5310
+    },
+    {
+      "epoch": 8.484025559105431,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5311
+    },
+    {
+      "epoch": 8.485623003194888,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1539,
+      "step": 5312
+    },
+    {
+      "epoch": 8.487220447284345,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5313
+    },
+    {
+      "epoch": 8.488817891373802,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5314
+    },
+    {
+      "epoch": 8.49041533546326,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5315
+    },
+    {
+      "epoch": 8.492012779552716,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5316
+    },
+    {
+      "epoch": 8.493610223642172,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 5317
+    },
+    {
+      "epoch": 8.495207667731629,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5318
+    },
+    {
+      "epoch": 8.496805111821086,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 5319
+    },
+    {
+      "epoch": 8.498402555910543,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 5320
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 5321
+    },
+    {
+      "epoch": 8.501597444089457,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 5322
+    },
+    {
+      "epoch": 8.503194888178914,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5323
+    },
+    {
+      "epoch": 8.504792332268371,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5324
+    },
+    {
+      "epoch": 8.506389776357828,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 5325
+    },
+    {
+      "epoch": 8.507987220447284,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5326
+    },
+    {
+      "epoch": 8.50958466453674,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 5327
+    },
+    {
+      "epoch": 8.511182108626198,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5328
+    },
+    {
+      "epoch": 8.512779552715655,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 5329
+    },
+    {
+      "epoch": 8.514376996805112,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5330
+    },
+    {
+      "epoch": 8.515974440894569,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5331
+    },
+    {
+      "epoch": 8.517571884984026,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5332
+    },
+    {
+      "epoch": 8.519169329073483,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 5333
+    },
+    {
+      "epoch": 8.52076677316294,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 5334
+    },
+    {
+      "epoch": 8.522364217252395,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 5335
+    },
+    {
+      "epoch": 8.523961661341852,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5336
+    },
+    {
+      "epoch": 8.52555910543131,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 5337
+    },
+    {
+      "epoch": 8.527156549520766,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 5338
+    },
+    {
+      "epoch": 8.528753993610223,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 5339
+    },
+    {
+      "epoch": 8.53035143769968,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 5340
+    },
+    {
+      "epoch": 8.531948881789138,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 5341
+    },
+    {
+      "epoch": 8.533546325878595,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 5342
+    },
+    {
+      "epoch": 8.535143769968052,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5343
+    },
+    {
+      "epoch": 8.536741214057509,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 5344
+    },
+    {
+      "epoch": 8.538338658146966,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5345
+    },
+    {
+      "epoch": 8.539936102236421,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 5346
+    },
+    {
+      "epoch": 8.541533546325878,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 5347
+    },
+    {
+      "epoch": 8.543130990415335,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5348
+    },
+    {
+      "epoch": 8.544728434504792,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5349
+    },
+    {
+      "epoch": 8.54632587859425,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 5350
+    },
+    {
+      "epoch": 8.547923322683706,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5351
+    },
+    {
+      "epoch": 8.549520766773163,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 5352
+    },
+    {
+      "epoch": 8.55111821086262,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5353
+    },
+    {
+      "epoch": 8.552715654952078,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 5354
+    },
+    {
+      "epoch": 8.554313099041533,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5355
+    },
+    {
+      "epoch": 8.55591054313099,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 5356
+    },
+    {
+      "epoch": 8.557507987220447,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5357
+    },
+    {
+      "epoch": 8.559105431309904,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5358
+    },
+    {
+      "epoch": 8.560702875399361,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5359
+    },
+    {
+      "epoch": 8.562300319488818,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 5360
+    },
+    {
+      "epoch": 8.563897763578275,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 5361
+    },
+    {
+      "epoch": 8.565495207667732,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5362
+    },
+    {
+      "epoch": 8.56709265175719,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 5363
+    },
+    {
+      "epoch": 8.568690095846645,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 5364
+    },
+    {
+      "epoch": 8.570287539936102,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5365
+    },
+    {
+      "epoch": 8.571884984025559,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5366
+    },
+    {
+      "epoch": 8.573482428115016,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 5367
+    },
+    {
+      "epoch": 8.575079872204473,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 5368
+    },
+    {
+      "epoch": 8.57667731629393,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5369
+    },
+    {
+      "epoch": 8.578274760383387,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 5370
+    },
+    {
+      "epoch": 8.579872204472844,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5371
+    },
+    {
+      "epoch": 8.581469648562301,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5372
+    },
+    {
+      "epoch": 8.583067092651756,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 5373
+    },
+    {
+      "epoch": 8.584664536741213,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5374
+    },
+    {
+      "epoch": 8.58626198083067,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5375
+    },
+    {
+      "epoch": 8.587859424920127,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5376
+    },
+    {
+      "epoch": 8.589456869009584,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 5377
+    },
+    {
+      "epoch": 8.591054313099042,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5378
+    },
+    {
+      "epoch": 8.592651757188499,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 5379
+    },
+    {
+      "epoch": 8.594249201277956,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 5380
+    },
+    {
+      "epoch": 8.595846645367413,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5381
+    },
+    {
+      "epoch": 8.59744408945687,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 5382
+    },
+    {
+      "epoch": 8.599041533546325,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5383
+    },
+    {
+      "epoch": 8.600638977635782,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 5384
+    },
+    {
+      "epoch": 8.60223642172524,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5385
+    },
+    {
+      "epoch": 8.603833865814696,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5386
+    },
+    {
+      "epoch": 8.605431309904153,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5387
+    },
+    {
+      "epoch": 8.60702875399361,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5388
+    },
+    {
+      "epoch": 8.608626198083067,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 5389
+    },
+    {
+      "epoch": 8.610223642172524,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 5390
+    },
+    {
+      "epoch": 8.611821086261982,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5391
+    },
+    {
+      "epoch": 8.613418530351439,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5392
+    },
+    {
+      "epoch": 8.615015974440894,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5393
+    },
+    {
+      "epoch": 8.616613418530351,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 5394
+    },
+    {
+      "epoch": 8.618210862619808,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5395
+    },
+    {
+      "epoch": 8.619808306709265,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 5396
+    },
+    {
+      "epoch": 8.621405750798722,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5397
+    },
+    {
+      "epoch": 8.62300319488818,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5398
+    },
+    {
+      "epoch": 8.624600638977636,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5399
+    },
+    {
+      "epoch": 8.626198083067093,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 5400
+    },
+    {
+      "epoch": 8.62779552715655,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 5401
+    },
+    {
+      "epoch": 8.629392971246006,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 5402
+    },
+    {
+      "epoch": 8.630990415335463,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5403
+    },
+    {
+      "epoch": 8.63258785942492,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5404
+    },
+    {
+      "epoch": 8.634185303514377,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1522,
+      "step": 5405
+    },
+    {
+      "epoch": 8.635782747603834,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 5406
+    },
+    {
+      "epoch": 8.63738019169329,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 5407
+    },
+    {
+      "epoch": 8.638977635782748,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 5408
+    },
+    {
+      "epoch": 8.640575079872205,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 5409
+    },
+    {
+      "epoch": 8.642172523961662,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 5410
+    },
+    {
+      "epoch": 8.643769968051117,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 5411
+    },
+    {
+      "epoch": 8.645367412140574,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 5412
+    },
+    {
+      "epoch": 8.646964856230031,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 5413
+    },
+    {
+      "epoch": 8.648562300319488,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5414
+    },
+    {
+      "epoch": 8.650159744408946,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 5415
+    },
+    {
+      "epoch": 8.651757188498403,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 5416
+    },
+    {
+      "epoch": 8.65335463258786,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 5417
+    },
+    {
+      "epoch": 8.654952076677317,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 5418
+    },
+    {
+      "epoch": 8.656549520766774,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 5419
+    },
+    {
+      "epoch": 8.65814696485623,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5420
+    },
+    {
+      "epoch": 8.659744408945686,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5421
+    },
+    {
+      "epoch": 8.661341853035143,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5422
+    },
+    {
+      "epoch": 8.6629392971246,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 5423
+    },
+    {
+      "epoch": 8.664536741214057,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5424
+    },
+    {
+      "epoch": 8.666134185303514,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 5425
+    },
+    {
+      "epoch": 8.667731629392971,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5426
+    },
+    {
+      "epoch": 8.669329073482428,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5427
+    },
+    {
+      "epoch": 8.670926517571885,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5428
+    },
+    {
+      "epoch": 8.672523961661343,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 5429
+    },
+    {
+      "epoch": 8.6741214057508,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5430
+    },
+    {
+      "epoch": 8.675718849840255,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 5431
+    },
+    {
+      "epoch": 8.677316293929712,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 5432
+    },
+    {
+      "epoch": 8.678913738019169,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5433
+    },
+    {
+      "epoch": 8.680511182108626,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 5434
+    },
+    {
+      "epoch": 8.682108626198083,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 5435
+    },
+    {
+      "epoch": 8.68370607028754,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 5436
+    },
+    {
+      "epoch": 8.685303514376997,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5437
+    },
+    {
+      "epoch": 8.686900958466454,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 5438
+    },
+    {
+      "epoch": 8.688498402555911,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 5439
+    },
+    {
+      "epoch": 8.690095846645367,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 5440
+    },
+    {
+      "epoch": 8.691693290734824,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 5441
+    },
+    {
+      "epoch": 8.69329073482428,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 5442
+    },
+    {
+      "epoch": 8.694888178913738,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5443
+    },
+    {
+      "epoch": 8.696485623003195,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5444
+    },
+    {
+      "epoch": 8.698083067092652,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5445
+    },
+    {
+      "epoch": 8.699680511182109,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 5446
+    },
+    {
+      "epoch": 8.701277955271566,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5447
+    },
+    {
+      "epoch": 8.702875399361023,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 5448
+    },
+    {
+      "epoch": 8.704472843450478,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 5449
+    },
+    {
+      "epoch": 8.706070287539935,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 5450
+    },
+    {
+      "epoch": 8.707667731629392,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 5451
+    },
+    {
+      "epoch": 8.70926517571885,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 5452
+    },
+    {
+      "epoch": 8.710862619808307,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 5453
+    },
+    {
+      "epoch": 8.712460063897764,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 5454
+    },
+    {
+      "epoch": 8.71405750798722,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 5455
+    },
+    {
+      "epoch": 8.715654952076678,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 5456
+    },
+    {
+      "epoch": 8.717252396166135,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 5457
+    },
+    {
+      "epoch": 8.718849840255592,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 5458
+    },
+    {
+      "epoch": 8.720447284345047,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5459
+    },
+    {
+      "epoch": 8.722044728434504,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 5460
+    },
+    {
+      "epoch": 8.723642172523961,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5461
+    },
+    {
+      "epoch": 8.725239616613418,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5462
+    },
+    {
+      "epoch": 8.726837060702875,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 5463
+    },
+    {
+      "epoch": 8.728434504792332,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5464
+    },
+    {
+      "epoch": 8.73003194888179,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5465
+    },
+    {
+      "epoch": 8.731629392971247,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5466
+    },
+    {
+      "epoch": 8.733226837060704,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 5467
+    },
+    {
+      "epoch": 8.73482428115016,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5468
+    },
+    {
+      "epoch": 8.736421725239616,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5469
+    },
+    {
+      "epoch": 8.738019169329073,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 5470
+    },
+    {
+      "epoch": 8.73961661341853,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 5471
+    },
+    {
+      "epoch": 8.741214057507987,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 5472
+    },
+    {
+      "epoch": 8.742811501597444,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 5473
+    },
+    {
+      "epoch": 8.744408945686901,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5474
+    },
+    {
+      "epoch": 8.746006389776358,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5475
+    },
+    {
+      "epoch": 8.747603833865815,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5476
+    },
+    {
+      "epoch": 8.749201277955272,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 5477
+    },
+    {
+      "epoch": 8.750798722044728,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5478
+    },
+    {
+      "epoch": 8.752396166134185,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5479
+    },
+    {
+      "epoch": 8.753993610223642,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5480
+    },
+    {
+      "epoch": 8.755591054313099,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 5481
+    },
+    {
+      "epoch": 8.757188498402556,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5482
+    },
+    {
+      "epoch": 8.758785942492013,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 5483
+    },
+    {
+      "epoch": 8.76038338658147,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 5484
+    },
+    {
+      "epoch": 8.761980830670927,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 5485
+    },
+    {
+      "epoch": 8.763578274760384,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 5486
+    },
+    {
+      "epoch": 8.76517571884984,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5487
+    },
+    {
+      "epoch": 8.766773162939296,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 5488
+    },
+    {
+      "epoch": 8.768370607028753,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5489
+    },
+    {
+      "epoch": 8.76996805111821,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 5490
+    },
+    {
+      "epoch": 8.771565495207668,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5491
+    },
+    {
+      "epoch": 8.773162939297125,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5492
+    },
+    {
+      "epoch": 8.774760383386582,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 5493
+    },
+    {
+      "epoch": 8.776357827476039,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 5494
+    },
+    {
+      "epoch": 8.777955271565496,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5495
+    },
+    {
+      "epoch": 8.779552715654953,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 5496
+    },
+    {
+      "epoch": 8.781150159744408,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 5497
+    },
+    {
+      "epoch": 8.782747603833865,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 5498
+    },
+    {
+      "epoch": 8.784345047923322,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 5499
+    },
+    {
+      "epoch": 8.78594249201278,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 5500
+    },
+    {
+      "epoch": 8.787539936102236,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1179,
+      "step": 5501
+    },
+    {
+      "epoch": 8.789137380191693,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 5502
+    },
+    {
+      "epoch": 8.79073482428115,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 5503
+    },
+    {
+      "epoch": 8.792332268370608,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 5504
+    },
+    {
+      "epoch": 8.793929712460065,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 5505
+    },
+    {
+      "epoch": 8.795527156549522,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5506
+    },
+    {
+      "epoch": 8.797124600638977,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5507
+    },
+    {
+      "epoch": 8.798722044728434,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 5508
+    },
+    {
+      "epoch": 8.800319488817891,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 5509
+    },
+    {
+      "epoch": 8.801916932907348,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 5510
+    },
+    {
+      "epoch": 8.803514376996805,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 5511
+    },
+    {
+      "epoch": 8.805111821086262,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5512
+    },
+    {
+      "epoch": 8.80670926517572,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5513
+    },
+    {
+      "epoch": 8.808306709265176,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 5514
+    },
+    {
+      "epoch": 8.809904153354633,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 5515
+    },
+    {
+      "epoch": 8.811501597444089,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 5516
+    },
+    {
+      "epoch": 8.813099041533546,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5517
+    },
+    {
+      "epoch": 8.814696485623003,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 5518
+    },
+    {
+      "epoch": 8.81629392971246,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5519
+    },
+    {
+      "epoch": 8.817891373801917,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 5520
+    },
+    {
+      "epoch": 8.819488817891374,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5521
+    },
+    {
+      "epoch": 8.821086261980831,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1521,
+      "step": 5522
+    },
+    {
+      "epoch": 8.822683706070288,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 5523
+    },
+    {
+      "epoch": 8.824281150159745,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1206,
+      "step": 5524
+    },
+    {
+      "epoch": 8.8258785942492,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 5525
+    },
+    {
+      "epoch": 8.827476038338657,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 5526
+    },
+    {
+      "epoch": 8.829073482428115,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5527
+    },
+    {
+      "epoch": 8.830670926517572,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 5528
+    },
+    {
+      "epoch": 8.832268370607029,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 5529
+    },
+    {
+      "epoch": 8.833865814696486,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5530
+    },
+    {
+      "epoch": 8.835463258785943,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5531
+    },
+    {
+      "epoch": 8.8370607028754,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5532
+    },
+    {
+      "epoch": 8.838658146964857,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 5533
+    },
+    {
+      "epoch": 8.840255591054314,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5534
+    },
+    {
+      "epoch": 8.84185303514377,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 5535
+    },
+    {
+      "epoch": 8.843450479233226,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5536
+    },
+    {
+      "epoch": 8.845047923322683,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 5537
+    },
+    {
+      "epoch": 8.84664536741214,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1484,
+      "step": 5538
+    },
+    {
+      "epoch": 8.848242811501597,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 5539
+    },
+    {
+      "epoch": 8.849840255591054,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 5540
+    },
+    {
+      "epoch": 8.851437699680512,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5541
+    },
+    {
+      "epoch": 8.853035143769969,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5542
+    },
+    {
+      "epoch": 8.854632587859426,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 5543
+    },
+    {
+      "epoch": 8.856230031948883,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 5544
+    },
+    {
+      "epoch": 8.857827476038338,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5545
+    },
+    {
+      "epoch": 8.859424920127795,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 5546
+    },
+    {
+      "epoch": 8.861022364217252,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 5547
+    },
+    {
+      "epoch": 8.86261980830671,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 5548
+    },
+    {
+      "epoch": 8.864217252396166,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 5549
+    },
+    {
+      "epoch": 8.865814696485623,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 5550
+    },
+    {
+      "epoch": 8.86741214057508,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5551
+    },
+    {
+      "epoch": 8.869009584664537,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5552
+    },
+    {
+      "epoch": 8.870607028753994,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5553
+    },
+    {
+      "epoch": 8.87220447284345,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5554
+    },
+    {
+      "epoch": 8.873801916932907,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 5555
+    },
+    {
+      "epoch": 8.875399361022364,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 5556
+    },
+    {
+      "epoch": 8.87699680511182,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 5557
+    },
+    {
+      "epoch": 8.878594249201278,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5558
+    },
+    {
+      "epoch": 8.880191693290735,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 5559
+    },
+    {
+      "epoch": 8.881789137380192,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 5560
+    },
+    {
+      "epoch": 8.883386581469649,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5561
+    },
+    {
+      "epoch": 8.884984025559106,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5562
+    },
+    {
+      "epoch": 8.886581469648561,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 5563
+    },
+    {
+      "epoch": 8.888178913738018,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 5564
+    },
+    {
+      "epoch": 8.889776357827476,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 5565
+    },
+    {
+      "epoch": 8.891373801916933,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 5566
+    },
+    {
+      "epoch": 8.89297124600639,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 5567
+    },
+    {
+      "epoch": 8.894568690095847,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 5568
+    },
+    {
+      "epoch": 8.896166134185304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5569
+    },
+    {
+      "epoch": 8.89776357827476,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5570
+    },
+    {
+      "epoch": 8.899361022364218,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 5571
+    },
+    {
+      "epoch": 8.900958466453675,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5572
+    },
+    {
+      "epoch": 8.90255591054313,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 5573
+    },
+    {
+      "epoch": 8.904153354632587,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 5574
+    },
+    {
+      "epoch": 8.905750798722044,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5575
+    },
+    {
+      "epoch": 8.907348242811501,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5576
+    },
+    {
+      "epoch": 8.908945686900958,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5577
+    },
+    {
+      "epoch": 8.910543130990416,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 5578
+    },
+    {
+      "epoch": 8.912140575079873,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 5579
+    },
+    {
+      "epoch": 8.91373801916933,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 5580
+    },
+    {
+      "epoch": 8.915335463258787,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 5581
+    },
+    {
+      "epoch": 8.916932907348244,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5582
+    },
+    {
+      "epoch": 8.918530351437699,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 5583
+    },
+    {
+      "epoch": 8.920127795527156,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5584
+    },
+    {
+      "epoch": 8.921725239616613,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5585
+    },
+    {
+      "epoch": 8.92332268370607,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5586
+    },
+    {
+      "epoch": 8.924920127795527,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1557,
+      "step": 5587
+    },
+    {
+      "epoch": 8.926517571884984,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 5588
+    },
+    {
+      "epoch": 8.928115015974441,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 5589
+    },
+    {
+      "epoch": 8.929712460063898,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 5590
+    },
+    {
+      "epoch": 8.931309904153355,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 5591
+    },
+    {
+      "epoch": 8.93290734824281,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5592
+    },
+    {
+      "epoch": 8.934504792332268,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 5593
+    },
+    {
+      "epoch": 8.936102236421725,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5594
+    },
+    {
+      "epoch": 8.937699680511182,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5595
+    },
+    {
+      "epoch": 8.939297124600639,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 5596
+    },
+    {
+      "epoch": 8.940894568690096,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 5597
+    },
+    {
+      "epoch": 8.942492012779553,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 5598
+    },
+    {
+      "epoch": 8.94408945686901,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5599
+    },
+    {
+      "epoch": 8.945686900958467,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5600
+    },
+    {
+      "epoch": 8.947284345047922,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5601
+    },
+    {
+      "epoch": 8.94888178913738,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 5602
+    },
+    {
+      "epoch": 8.950479233226837,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5603
+    },
+    {
+      "epoch": 8.952076677316294,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 5604
+    },
+    {
+      "epoch": 8.95367412140575,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 5605
+    },
+    {
+      "epoch": 8.955271565495208,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5606
+    },
+    {
+      "epoch": 8.956869009584665,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1535,
+      "step": 5607
+    },
+    {
+      "epoch": 8.958466453674122,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 5608
+    },
+    {
+      "epoch": 8.960063897763579,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 5609
+    },
+    {
+      "epoch": 8.961661341853034,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 5610
+    },
+    {
+      "epoch": 8.963258785942491,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 5611
+    },
+    {
+      "epoch": 8.964856230031948,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5612
+    },
+    {
+      "epoch": 8.966453674121405,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 5613
+    },
+    {
+      "epoch": 8.968051118210862,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5614
+    },
+    {
+      "epoch": 8.96964856230032,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 5615
+    },
+    {
+      "epoch": 8.971246006389777,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 5616
+    },
+    {
+      "epoch": 8.972843450479234,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 5617
+    },
+    {
+      "epoch": 8.97444089456869,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 5618
+    },
+    {
+      "epoch": 8.976038338658148,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5619
+    },
+    {
+      "epoch": 8.977635782747605,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 5620
+    },
+    {
+      "epoch": 8.97923322683706,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 5621
+    },
+    {
+      "epoch": 8.980830670926517,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5622
+    },
+    {
+      "epoch": 8.982428115015974,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 5623
+    },
+    {
+      "epoch": 8.984025559105431,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 5624
+    },
+    {
+      "epoch": 8.985623003194888,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 5625
+    },
+    {
+      "epoch": 8.987220447284345,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 5626
+    },
+    {
+      "epoch": 8.988817891373802,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 5627
+    },
+    {
+      "epoch": 8.99041533546326,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5628
+    },
+    {
+      "epoch": 8.992012779552716,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5629
+    },
+    {
+      "epoch": 8.993610223642172,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5630
+    },
+    {
+      "epoch": 8.995207667731629,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 5631
+    },
+    {
+      "epoch": 8.996805111821086,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 5632
+    },
+    {
+      "epoch": 8.998402555910543,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5633
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5634
+    },
+    {
+      "epoch": 9.001597444089457,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5635
+    },
+    {
+      "epoch": 9.003194888178914,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 5636
+    },
+    {
+      "epoch": 9.004792332268371,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 5637
+    },
+    {
+      "epoch": 9.006389776357828,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5638
+    },
+    {
+      "epoch": 9.007987220447284,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 5639
+    },
+    {
+      "epoch": 9.00958466453674,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5640
+    },
+    {
+      "epoch": 9.011182108626198,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 5641
+    },
+    {
+      "epoch": 9.012779552715655,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5642
+    },
+    {
+      "epoch": 9.014376996805112,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5643
+    },
+    {
+      "epoch": 9.015974440894569,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 5644
+    },
+    {
+      "epoch": 9.017571884984026,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 5645
+    },
+    {
+      "epoch": 9.019169329073483,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 5646
+    },
+    {
+      "epoch": 9.02076677316294,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 5647
+    },
+    {
+      "epoch": 9.022364217252397,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5648
+    },
+    {
+      "epoch": 9.023961661341852,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1481,
+      "step": 5649
+    },
+    {
+      "epoch": 9.02555910543131,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5650
+    },
+    {
+      "epoch": 9.027156549520766,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5651
+    },
+    {
+      "epoch": 9.028753993610223,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5652
+    },
+    {
+      "epoch": 9.03035143769968,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 5653
+    },
+    {
+      "epoch": 9.031948881789138,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 5654
+    },
+    {
+      "epoch": 9.033546325878595,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 5655
+    },
+    {
+      "epoch": 9.035143769968052,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 5656
+    },
+    {
+      "epoch": 9.036741214057509,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5657
+    },
+    {
+      "epoch": 9.038338658146964,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 5658
+    },
+    {
+      "epoch": 9.039936102236421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5659
+    },
+    {
+      "epoch": 9.041533546325878,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 5660
+    },
+    {
+      "epoch": 9.043130990415335,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 5661
+    },
+    {
+      "epoch": 9.044728434504792,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 5662
+    },
+    {
+      "epoch": 9.04632587859425,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5663
+    },
+    {
+      "epoch": 9.047923322683706,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5664
+    },
+    {
+      "epoch": 9.049520766773163,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5665
+    },
+    {
+      "epoch": 9.05111821086262,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 5666
+    },
+    {
+      "epoch": 9.052715654952078,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5667
+    },
+    {
+      "epoch": 9.054313099041533,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 5668
+    },
+    {
+      "epoch": 9.05591054313099,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 5669
+    },
+    {
+      "epoch": 9.057507987220447,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 5670
+    },
+    {
+      "epoch": 9.059105431309904,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5671
+    },
+    {
+      "epoch": 9.060702875399361,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 5672
+    },
+    {
+      "epoch": 9.062300319488818,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 5673
+    },
+    {
+      "epoch": 9.063897763578275,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5674
+    },
+    {
+      "epoch": 9.065495207667732,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 5675
+    },
+    {
+      "epoch": 9.06709265175719,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 5676
+    },
+    {
+      "epoch": 9.068690095846645,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5677
+    },
+    {
+      "epoch": 9.070287539936102,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5678
+    },
+    {
+      "epoch": 9.071884984025559,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5679
+    },
+    {
+      "epoch": 9.073482428115016,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5680
+    },
+    {
+      "epoch": 9.075079872204473,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5681
+    },
+    {
+      "epoch": 9.07667731629393,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 5682
+    },
+    {
+      "epoch": 9.078274760383387,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5683
+    },
+    {
+      "epoch": 9.079872204472844,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 5684
+    },
+    {
+      "epoch": 9.081469648562301,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 5685
+    },
+    {
+      "epoch": 9.083067092651758,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 5686
+    },
+    {
+      "epoch": 9.084664536741213,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5687
+    },
+    {
+      "epoch": 9.08626198083067,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 5688
+    },
+    {
+      "epoch": 9.087859424920127,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 5689
+    },
+    {
+      "epoch": 9.089456869009584,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 5690
+    },
+    {
+      "epoch": 9.091054313099042,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5691
+    },
+    {
+      "epoch": 9.092651757188499,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 5692
+    },
+    {
+      "epoch": 9.094249201277956,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 5693
+    },
+    {
+      "epoch": 9.095846645367413,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5694
+    },
+    {
+      "epoch": 9.09744408945687,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5695
+    },
+    {
+      "epoch": 9.099041533546325,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 5696
+    },
+    {
+      "epoch": 9.100638977635782,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 5697
+    },
+    {
+      "epoch": 9.10223642172524,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 5698
+    },
+    {
+      "epoch": 9.103833865814696,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 5699
+    },
+    {
+      "epoch": 9.105431309904153,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 5700
+    },
+    {
+      "epoch": 9.10702875399361,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 5701
+    },
+    {
+      "epoch": 9.108626198083067,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5702
+    },
+    {
+      "epoch": 9.110223642172524,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5703
+    },
+    {
+      "epoch": 9.111821086261982,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5704
+    },
+    {
+      "epoch": 9.113418530351439,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 5705
+    },
+    {
+      "epoch": 9.115015974440894,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5706
+    },
+    {
+      "epoch": 9.116613418530351,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 5707
+    },
+    {
+      "epoch": 9.118210862619808,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5708
+    },
+    {
+      "epoch": 9.119808306709265,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5709
+    },
+    {
+      "epoch": 9.121405750798722,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 5710
+    },
+    {
+      "epoch": 9.12300319488818,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 5711
+    },
+    {
+      "epoch": 9.124600638977636,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 5712
+    },
+    {
+      "epoch": 9.126198083067093,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5713
+    },
+    {
+      "epoch": 9.12779552715655,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 5714
+    },
+    {
+      "epoch": 9.129392971246006,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5715
+    },
+    {
+      "epoch": 9.130990415335463,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 5716
+    },
+    {
+      "epoch": 9.13258785942492,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 5717
+    },
+    {
+      "epoch": 9.134185303514377,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 5718
+    },
+    {
+      "epoch": 9.135782747603834,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 5719
+    },
+    {
+      "epoch": 9.13738019169329,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 5720
+    },
+    {
+      "epoch": 9.138977635782748,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 5721
+    },
+    {
+      "epoch": 9.140575079872205,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 5722
+    },
+    {
+      "epoch": 9.142172523961662,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5723
+    },
+    {
+      "epoch": 9.143769968051119,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5724
+    },
+    {
+      "epoch": 9.145367412140574,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 5725
+    },
+    {
+      "epoch": 9.146964856230031,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 5726
+    },
+    {
+      "epoch": 9.148562300319488,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5727
+    },
+    {
+      "epoch": 9.150159744408946,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 5728
+    },
+    {
+      "epoch": 9.151757188498403,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5729
+    },
+    {
+      "epoch": 9.15335463258786,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 5730
+    },
+    {
+      "epoch": 9.154952076677317,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5731
+    },
+    {
+      "epoch": 9.156549520766774,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5732
+    },
+    {
+      "epoch": 9.15814696485623,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 5733
+    },
+    {
+      "epoch": 9.159744408945686,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 5734
+    },
+    {
+      "epoch": 9.161341853035143,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 5735
+    },
+    {
+      "epoch": 9.1629392971246,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 5736
+    },
+    {
+      "epoch": 9.164536741214057,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 5737
+    },
+    {
+      "epoch": 9.166134185303514,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 5738
+    },
+    {
+      "epoch": 9.167731629392971,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 5739
+    },
+    {
+      "epoch": 9.169329073482428,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 5740
+    },
+    {
+      "epoch": 9.170926517571885,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 5741
+    },
+    {
+      "epoch": 9.172523961661343,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5742
+    },
+    {
+      "epoch": 9.1741214057508,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 5743
+    },
+    {
+      "epoch": 9.175718849840255,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 5744
+    },
+    {
+      "epoch": 9.177316293929712,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5745
+    },
+    {
+      "epoch": 9.178913738019169,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 5746
+    },
+    {
+      "epoch": 9.180511182108626,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 5747
+    },
+    {
+      "epoch": 9.182108626198083,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 5748
+    },
+    {
+      "epoch": 9.18370607028754,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5749
+    },
+    {
+      "epoch": 9.185303514376997,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 5750
+    },
+    {
+      "epoch": 9.186900958466454,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 5751
+    },
+    {
+      "epoch": 9.188498402555911,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 5752
+    },
+    {
+      "epoch": 9.190095846645367,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 5753
+    },
+    {
+      "epoch": 9.191693290734824,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 5754
+    },
+    {
+      "epoch": 9.19329073482428,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5755
+    },
+    {
+      "epoch": 9.194888178913738,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5756
+    },
+    {
+      "epoch": 9.196485623003195,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5757
+    },
+    {
+      "epoch": 9.198083067092652,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 5758
+    },
+    {
+      "epoch": 9.199680511182109,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 5759
+    },
+    {
+      "epoch": 9.201277955271566,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5760
+    },
+    {
+      "epoch": 9.202875399361023,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 5761
+    },
+    {
+      "epoch": 9.204472843450478,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5762
+    },
+    {
+      "epoch": 9.206070287539935,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 5763
+    },
+    {
+      "epoch": 9.207667731629392,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5764
+    },
+    {
+      "epoch": 9.20926517571885,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 5765
+    },
+    {
+      "epoch": 9.210862619808307,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 5766
+    },
+    {
+      "epoch": 9.212460063897764,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5767
+    },
+    {
+      "epoch": 9.21405750798722,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 5768
+    },
+    {
+      "epoch": 9.215654952076678,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 5769
+    },
+    {
+      "epoch": 9.217252396166135,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5770
+    },
+    {
+      "epoch": 9.218849840255592,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 5771
+    },
+    {
+      "epoch": 9.220447284345047,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5772
+    },
+    {
+      "epoch": 9.222044728434504,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 5773
+    },
+    {
+      "epoch": 9.223642172523961,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 5774
+    },
+    {
+      "epoch": 9.225239616613418,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5775
+    },
+    {
+      "epoch": 9.226837060702875,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 5776
+    },
+    {
+      "epoch": 9.228434504792332,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 5777
+    },
+    {
+      "epoch": 9.23003194888179,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5778
+    },
+    {
+      "epoch": 9.231629392971247,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 5779
+    },
+    {
+      "epoch": 9.233226837060704,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 5780
+    },
+    {
+      "epoch": 9.23482428115016,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5781
+    },
+    {
+      "epoch": 9.236421725239616,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5782
+    },
+    {
+      "epoch": 9.238019169329073,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 5783
+    },
+    {
+      "epoch": 9.23961661341853,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5784
+    },
+    {
+      "epoch": 9.241214057507987,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 5785
+    },
+    {
+      "epoch": 9.242811501597444,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 5786
+    },
+    {
+      "epoch": 9.244408945686901,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 5787
+    },
+    {
+      "epoch": 9.246006389776358,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5788
+    },
+    {
+      "epoch": 9.247603833865815,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5789
+    },
+    {
+      "epoch": 9.249201277955272,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5790
+    },
+    {
+      "epoch": 9.250798722044728,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 5791
+    },
+    {
+      "epoch": 9.252396166134185,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 5792
+    },
+    {
+      "epoch": 9.253993610223642,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 5793
+    },
+    {
+      "epoch": 9.255591054313099,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 5794
+    },
+    {
+      "epoch": 9.257188498402556,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 5795
+    },
+    {
+      "epoch": 9.258785942492013,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 5796
+    },
+    {
+      "epoch": 9.26038338658147,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 5797
+    },
+    {
+      "epoch": 9.261980830670927,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1192,
+      "step": 5798
+    },
+    {
+      "epoch": 9.263578274760384,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5799
+    },
+    {
+      "epoch": 9.26517571884984,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5800
+    },
+    {
+      "epoch": 9.266773162939296,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5801
+    },
+    {
+      "epoch": 9.268370607028753,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 5802
+    },
+    {
+      "epoch": 9.26996805111821,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 5803
+    },
+    {
+      "epoch": 9.271565495207668,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 5804
+    },
+    {
+      "epoch": 9.273162939297125,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 5805
+    },
+    {
+      "epoch": 9.274760383386582,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5806
+    },
+    {
+      "epoch": 9.276357827476039,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5807
+    },
+    {
+      "epoch": 9.277955271565496,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5808
+    },
+    {
+      "epoch": 9.279552715654953,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 5809
+    },
+    {
+      "epoch": 9.281150159744408,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 5810
+    },
+    {
+      "epoch": 9.282747603833865,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 5811
+    },
+    {
+      "epoch": 9.284345047923322,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 5812
+    },
+    {
+      "epoch": 9.28594249201278,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 5813
+    },
+    {
+      "epoch": 9.287539936102236,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5814
+    },
+    {
+      "epoch": 9.289137380191693,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 5815
+    },
+    {
+      "epoch": 9.29073482428115,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5816
+    },
+    {
+      "epoch": 9.292332268370608,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 5817
+    },
+    {
+      "epoch": 9.293929712460065,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 5818
+    },
+    {
+      "epoch": 9.295527156549522,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 5819
+    },
+    {
+      "epoch": 9.297124600638977,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 5820
+    },
+    {
+      "epoch": 9.298722044728434,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5821
+    },
+    {
+      "epoch": 9.300319488817891,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 5822
+    },
+    {
+      "epoch": 9.301916932907348,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5823
+    },
+    {
+      "epoch": 9.303514376996805,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 5824
+    },
+    {
+      "epoch": 9.305111821086262,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 5825
+    },
+    {
+      "epoch": 9.30670926517572,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 5826
+    },
+    {
+      "epoch": 9.308306709265176,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 5827
+    },
+    {
+      "epoch": 9.309904153354633,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 5828
+    },
+    {
+      "epoch": 9.311501597444089,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 5829
+    },
+    {
+      "epoch": 9.313099041533546,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 5830
+    },
+    {
+      "epoch": 9.314696485623003,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5831
+    },
+    {
+      "epoch": 9.31629392971246,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 5832
+    },
+    {
+      "epoch": 9.317891373801917,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5833
+    },
+    {
+      "epoch": 9.319488817891374,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1536,
+      "step": 5834
+    },
+    {
+      "epoch": 9.321086261980831,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5835
+    },
+    {
+      "epoch": 9.322683706070288,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5836
+    },
+    {
+      "epoch": 9.324281150159745,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 5837
+    },
+    {
+      "epoch": 9.3258785942492,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1459,
+      "step": 5838
+    },
+    {
+      "epoch": 9.327476038338657,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 5839
+    },
+    {
+      "epoch": 9.329073482428115,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 5840
+    },
+    {
+      "epoch": 9.330670926517572,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5841
+    },
+    {
+      "epoch": 9.332268370607029,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 5842
+    },
+    {
+      "epoch": 9.333865814696486,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5843
+    },
+    {
+      "epoch": 9.335463258785943,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 5844
+    },
+    {
+      "epoch": 9.3370607028754,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 5845
+    },
+    {
+      "epoch": 9.338658146964857,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 5846
+    },
+    {
+      "epoch": 9.340255591054314,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5847
+    },
+    {
+      "epoch": 9.34185303514377,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 5848
+    },
+    {
+      "epoch": 9.343450479233226,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5849
+    },
+    {
+      "epoch": 9.345047923322683,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 5850
+    },
+    {
+      "epoch": 9.34664536741214,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 5851
+    },
+    {
+      "epoch": 9.348242811501597,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 5852
+    },
+    {
+      "epoch": 9.349840255591054,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5853
+    },
+    {
+      "epoch": 9.351437699680512,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5854
+    },
+    {
+      "epoch": 9.353035143769969,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 5855
+    },
+    {
+      "epoch": 9.354632587859426,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 5856
+    },
+    {
+      "epoch": 9.356230031948883,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 5857
+    },
+    {
+      "epoch": 9.357827476038338,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 5858
+    },
+    {
+      "epoch": 9.359424920127795,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5859
+    },
+    {
+      "epoch": 9.361022364217252,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 5860
+    },
+    {
+      "epoch": 9.36261980830671,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 5861
+    },
+    {
+      "epoch": 9.364217252396166,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 5862
+    },
+    {
+      "epoch": 9.365814696485623,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5863
+    },
+    {
+      "epoch": 9.36741214057508,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 5864
+    },
+    {
+      "epoch": 9.369009584664537,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 5865
+    },
+    {
+      "epoch": 9.370607028753994,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 5866
+    },
+    {
+      "epoch": 9.37220447284345,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5867
+    },
+    {
+      "epoch": 9.373801916932907,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 5868
+    },
+    {
+      "epoch": 9.375399361022364,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 5869
+    },
+    {
+      "epoch": 9.37699680511182,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5870
+    },
+    {
+      "epoch": 9.378594249201278,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 5871
+    },
+    {
+      "epoch": 9.380191693290735,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 5872
+    },
+    {
+      "epoch": 9.381789137380192,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 5873
+    },
+    {
+      "epoch": 9.383386581469649,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 5874
+    },
+    {
+      "epoch": 9.384984025559106,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 5875
+    },
+    {
+      "epoch": 9.386581469648561,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 5876
+    },
+    {
+      "epoch": 9.388178913738018,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 5877
+    },
+    {
+      "epoch": 9.389776357827476,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 5878
+    },
+    {
+      "epoch": 9.391373801916933,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 5879
+    },
+    {
+      "epoch": 9.39297124600639,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 5880
+    },
+    {
+      "epoch": 9.394568690095847,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 5881
+    },
+    {
+      "epoch": 9.396166134185304,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5882
+    },
+    {
+      "epoch": 9.39776357827476,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 5883
+    },
+    {
+      "epoch": 9.399361022364218,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 5884
+    },
+    {
+      "epoch": 9.400958466453675,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 5885
+    },
+    {
+      "epoch": 9.40255591054313,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 5886
+    },
+    {
+      "epoch": 9.404153354632587,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.147,
+      "step": 5887
+    },
+    {
+      "epoch": 9.405750798722044,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 5888
+    },
+    {
+      "epoch": 9.407348242811501,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 5889
+    },
+    {
+      "epoch": 9.408945686900958,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5890
+    },
+    {
+      "epoch": 9.410543130990416,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 5891
+    },
+    {
+      "epoch": 9.412140575079873,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5892
+    },
+    {
+      "epoch": 9.41373801916933,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 5893
+    },
+    {
+      "epoch": 9.415335463258787,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 5894
+    },
+    {
+      "epoch": 9.416932907348242,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1493,
+      "step": 5895
+    },
+    {
+      "epoch": 9.418530351437699,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 5896
+    },
+    {
+      "epoch": 9.420127795527156,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 5897
+    },
+    {
+      "epoch": 9.421725239616613,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 5898
+    },
+    {
+      "epoch": 9.42332268370607,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 5899
+    },
+    {
+      "epoch": 9.424920127795527,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 5900
+    },
+    {
+      "epoch": 9.426517571884984,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 5901
+    },
+    {
+      "epoch": 9.428115015974441,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 5902
+    },
+    {
+      "epoch": 9.429712460063898,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 5903
+    },
+    {
+      "epoch": 9.431309904153355,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0005,
+      "loss": 1.1735,
+      "step": 5904
+    },
+    {
+      "epoch": 9.43290734824281,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0005,
+      "loss": 1.1734,
+      "step": 5905
+    },
+    {
+      "epoch": 9.434504792332268,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 5906
+    },
+    {
+      "epoch": 9.436102236421725,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0005,
+      "loss": 1.1745,
+      "step": 5907
+    },
+    {
+      "epoch": 9.437699680511182,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1776,
+      "step": 5908
+    },
+    {
+      "epoch": 9.439297124600639,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 5909
+    },
+    {
+      "epoch": 9.440894568690096,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1527,
+      "step": 5910
+    },
+    {
+      "epoch": 9.442492012779553,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 5911
+    },
+    {
+      "epoch": 9.44408945686901,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1546,
+      "step": 5912
+    },
+    {
+      "epoch": 9.445686900958467,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 5913
+    },
+    {
+      "epoch": 9.447284345047922,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 5914
+    },
+    {
+      "epoch": 9.44888178913738,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5915
+    },
+    {
+      "epoch": 9.450479233226837,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 5916
+    },
+    {
+      "epoch": 9.452076677316294,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5917
+    },
+    {
+      "epoch": 9.45367412140575,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 5918
+    },
+    {
+      "epoch": 9.455271565495208,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 5919
+    },
+    {
+      "epoch": 9.456869009584665,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 5920
+    },
+    {
+      "epoch": 9.458466453674122,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 5921
+    },
+    {
+      "epoch": 9.460063897763579,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 5922
+    },
+    {
+      "epoch": 9.461661341853036,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5923
+    },
+    {
+      "epoch": 9.463258785942491,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 5924
+    },
+    {
+      "epoch": 9.464856230031948,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 5925
+    },
+    {
+      "epoch": 9.466453674121405,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 5926
+    },
+    {
+      "epoch": 9.468051118210862,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 5927
+    },
+    {
+      "epoch": 9.46964856230032,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 5928
+    },
+    {
+      "epoch": 9.471246006389777,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1463,
+      "step": 5929
+    },
+    {
+      "epoch": 9.472843450479234,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5930
+    },
+    {
+      "epoch": 9.47444089456869,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 5931
+    },
+    {
+      "epoch": 9.476038338658148,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5932
+    },
+    {
+      "epoch": 9.477635782747603,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 5933
+    },
+    {
+      "epoch": 9.47923322683706,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 5934
+    },
+    {
+      "epoch": 9.480830670926517,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 5935
+    },
+    {
+      "epoch": 9.482428115015974,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 5936
+    },
+    {
+      "epoch": 9.484025559105431,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5937
+    },
+    {
+      "epoch": 9.485623003194888,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 5938
+    },
+    {
+      "epoch": 9.487220447284345,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5939
+    },
+    {
+      "epoch": 9.488817891373802,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 5940
+    },
+    {
+      "epoch": 9.49041533546326,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 5941
+    },
+    {
+      "epoch": 9.492012779552716,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 5942
+    },
+    {
+      "epoch": 9.493610223642172,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 5943
+    },
+    {
+      "epoch": 9.495207667731629,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 5944
+    },
+    {
+      "epoch": 9.496805111821086,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 5945
+    },
+    {
+      "epoch": 9.498402555910543,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5946
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 5947
+    },
+    {
+      "epoch": 9.501597444089457,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 5948
+    },
+    {
+      "epoch": 9.503194888178914,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 5949
+    },
+    {
+      "epoch": 9.504792332268371,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 5950
+    },
+    {
+      "epoch": 9.506389776357828,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 5951
+    },
+    {
+      "epoch": 9.507987220447284,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 5952
+    },
+    {
+      "epoch": 9.50958466453674,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1476,
+      "step": 5953
+    },
+    {
+      "epoch": 9.511182108626198,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 5954
+    },
+    {
+      "epoch": 9.512779552715655,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 5955
+    },
+    {
+      "epoch": 9.514376996805112,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 5956
+    },
+    {
+      "epoch": 9.515974440894569,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5957
+    },
+    {
+      "epoch": 9.517571884984026,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 5958
+    },
+    {
+      "epoch": 9.519169329073483,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 5959
+    },
+    {
+      "epoch": 9.52076677316294,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 5960
+    },
+    {
+      "epoch": 9.522364217252395,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 5961
+    },
+    {
+      "epoch": 9.523961661341852,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 5962
+    },
+    {
+      "epoch": 9.52555910543131,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 5963
+    },
+    {
+      "epoch": 9.527156549520766,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 5964
+    },
+    {
+      "epoch": 9.528753993610223,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 5965
+    },
+    {
+      "epoch": 9.53035143769968,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 5966
+    },
+    {
+      "epoch": 9.531948881789138,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 5967
+    },
+    {
+      "epoch": 9.533546325878595,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 5968
+    },
+    {
+      "epoch": 9.535143769968052,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 5969
+    },
+    {
+      "epoch": 9.536741214057509,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 5970
+    },
+    {
+      "epoch": 9.538338658146966,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 5971
+    },
+    {
+      "epoch": 9.539936102236421,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 5972
+    },
+    {
+      "epoch": 9.541533546325878,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 5973
+    },
+    {
+      "epoch": 9.543130990415335,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 5974
+    },
+    {
+      "epoch": 9.544728434504792,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 5975
+    },
+    {
+      "epoch": 9.54632587859425,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 5976
+    },
+    {
+      "epoch": 9.547923322683706,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 5977
+    },
+    {
+      "epoch": 9.549520766773163,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 5978
+    },
+    {
+      "epoch": 9.55111821086262,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 5979
+    },
+    {
+      "epoch": 9.552715654952078,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 5980
+    },
+    {
+      "epoch": 9.554313099041533,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 5981
+    },
+    {
+      "epoch": 9.55591054313099,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 5982
+    },
+    {
+      "epoch": 9.557507987220447,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 5983
+    },
+    {
+      "epoch": 9.559105431309904,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 5984
+    },
+    {
+      "epoch": 9.560702875399361,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 5985
+    },
+    {
+      "epoch": 9.562300319488818,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 5986
+    },
+    {
+      "epoch": 9.563897763578275,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 5987
+    },
+    {
+      "epoch": 9.565495207667732,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 5988
+    },
+    {
+      "epoch": 9.56709265175719,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 5989
+    },
+    {
+      "epoch": 9.568690095846645,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 5990
+    },
+    {
+      "epoch": 9.570287539936102,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 5991
+    },
+    {
+      "epoch": 9.571884984025559,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 5992
+    },
+    {
+      "epoch": 9.573482428115016,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 5993
+    },
+    {
+      "epoch": 9.575079872204473,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1507,
+      "step": 5994
+    },
+    {
+      "epoch": 9.57667731629393,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 5995
+    },
+    {
+      "epoch": 9.578274760383387,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 5996
+    },
+    {
+      "epoch": 9.579872204472844,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 5997
+    },
+    {
+      "epoch": 9.581469648562301,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 5998
+    },
+    {
+      "epoch": 9.583067092651756,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 5999
+    },
+    {
+      "epoch": 9.584664536741213,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 6000
+    },
+    {
+      "epoch": 9.58626198083067,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 6001
+    },
+    {
+      "epoch": 9.587859424920127,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 6002
+    },
+    {
+      "epoch": 9.589456869009584,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 6003
+    },
+    {
+      "epoch": 9.591054313099042,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 6004
+    },
+    {
+      "epoch": 9.592651757188499,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6005
+    },
+    {
+      "epoch": 9.594249201277956,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6006
+    },
+    {
+      "epoch": 9.595846645367413,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 6007
+    },
+    {
+      "epoch": 9.59744408945687,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 6008
+    },
+    {
+      "epoch": 9.599041533546325,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6009
+    },
+    {
+      "epoch": 9.600638977635782,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6010
+    },
+    {
+      "epoch": 9.60223642172524,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 6011
+    },
+    {
+      "epoch": 9.603833865814696,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 6012
+    },
+    {
+      "epoch": 9.605431309904153,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6013
+    },
+    {
+      "epoch": 9.60702875399361,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6014
+    },
+    {
+      "epoch": 9.608626198083067,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 6015
+    },
+    {
+      "epoch": 9.610223642172524,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6016
+    },
+    {
+      "epoch": 9.611821086261982,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 6017
+    },
+    {
+      "epoch": 9.613418530351439,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6018
+    },
+    {
+      "epoch": 9.615015974440894,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 6019
+    },
+    {
+      "epoch": 9.616613418530351,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 6020
+    },
+    {
+      "epoch": 9.618210862619808,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6021
+    },
+    {
+      "epoch": 9.619808306709265,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 6022
+    },
+    {
+      "epoch": 9.621405750798722,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6023
+    },
+    {
+      "epoch": 9.62300319488818,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6024
+    },
+    {
+      "epoch": 9.624600638977636,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6025
+    },
+    {
+      "epoch": 9.626198083067093,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6026
+    },
+    {
+      "epoch": 9.62779552715655,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6027
+    },
+    {
+      "epoch": 9.629392971246006,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 6028
+    },
+    {
+      "epoch": 9.630990415335463,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6029
+    },
+    {
+      "epoch": 9.63258785942492,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6030
+    },
+    {
+      "epoch": 9.634185303514377,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 6031
+    },
+    {
+      "epoch": 9.635782747603834,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 6032
+    },
+    {
+      "epoch": 9.63738019169329,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6033
+    },
+    {
+      "epoch": 9.638977635782748,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 6034
+    },
+    {
+      "epoch": 9.640575079872205,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6035
+    },
+    {
+      "epoch": 9.642172523961662,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6036
+    },
+    {
+      "epoch": 9.643769968051117,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 6037
+    },
+    {
+      "epoch": 9.645367412140574,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6038
+    },
+    {
+      "epoch": 9.646964856230031,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 6039
+    },
+    {
+      "epoch": 9.648562300319488,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 6040
+    },
+    {
+      "epoch": 9.650159744408946,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 6041
+    },
+    {
+      "epoch": 9.651757188498403,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6042
+    },
+    {
+      "epoch": 9.65335463258786,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 6043
+    },
+    {
+      "epoch": 9.654952076677317,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6044
+    },
+    {
+      "epoch": 9.656549520766774,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 6045
+    },
+    {
+      "epoch": 9.65814696485623,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 6046
+    },
+    {
+      "epoch": 9.659744408945686,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1532,
+      "step": 6047
+    },
+    {
+      "epoch": 9.661341853035143,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6048
+    },
+    {
+      "epoch": 9.6629392971246,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 6049
+    },
+    {
+      "epoch": 9.664536741214057,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6050
+    },
+    {
+      "epoch": 9.666134185303514,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 6051
+    },
+    {
+      "epoch": 9.667731629392971,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 6052
+    },
+    {
+      "epoch": 9.669329073482428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6053
+    },
+    {
+      "epoch": 9.670926517571885,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6054
+    },
+    {
+      "epoch": 9.672523961661343,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1483,
+      "step": 6055
+    },
+    {
+      "epoch": 9.6741214057508,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6056
+    },
+    {
+      "epoch": 9.675718849840255,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6057
+    },
+    {
+      "epoch": 9.677316293929712,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 6058
+    },
+    {
+      "epoch": 9.678913738019169,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 6059
+    },
+    {
+      "epoch": 9.680511182108626,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6060
+    },
+    {
+      "epoch": 9.682108626198083,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 6061
+    },
+    {
+      "epoch": 9.68370607028754,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6062
+    },
+    {
+      "epoch": 9.685303514376997,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 6063
+    },
+    {
+      "epoch": 9.686900958466454,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6064
+    },
+    {
+      "epoch": 9.688498402555911,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6065
+    },
+    {
+      "epoch": 9.690095846645367,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 6066
+    },
+    {
+      "epoch": 9.691693290734824,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6067
+    },
+    {
+      "epoch": 9.69329073482428,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 6068
+    },
+    {
+      "epoch": 9.694888178913738,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6069
+    },
+    {
+      "epoch": 9.696485623003195,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6070
+    },
+    {
+      "epoch": 9.698083067092652,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 6071
+    },
+    {
+      "epoch": 9.699680511182109,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 6072
+    },
+    {
+      "epoch": 9.701277955271566,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6073
+    },
+    {
+      "epoch": 9.702875399361023,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 6074
+    },
+    {
+      "epoch": 9.704472843450478,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 6075
+    },
+    {
+      "epoch": 9.706070287539935,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 6076
+    },
+    {
+      "epoch": 9.707667731629392,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6077
+    },
+    {
+      "epoch": 9.70926517571885,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 6078
+    },
+    {
+      "epoch": 9.710862619808307,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6079
+    },
+    {
+      "epoch": 9.712460063897764,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 6080
+    },
+    {
+      "epoch": 9.71405750798722,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6081
+    },
+    {
+      "epoch": 9.715654952076678,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 6082
+    },
+    {
+      "epoch": 9.717252396166135,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 6083
+    },
+    {
+      "epoch": 9.718849840255592,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6084
+    },
+    {
+      "epoch": 9.720447284345047,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6085
+    },
+    {
+      "epoch": 9.722044728434504,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 6086
+    },
+    {
+      "epoch": 9.723642172523961,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 6087
+    },
+    {
+      "epoch": 9.725239616613418,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 6088
+    },
+    {
+      "epoch": 9.726837060702875,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6089
+    },
+    {
+      "epoch": 9.728434504792332,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 6090
+    },
+    {
+      "epoch": 9.73003194888179,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6091
+    },
+    {
+      "epoch": 9.731629392971247,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 6092
+    },
+    {
+      "epoch": 9.733226837060704,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 6093
+    },
+    {
+      "epoch": 9.73482428115016,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 6094
+    },
+    {
+      "epoch": 9.736421725239616,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6095
+    },
+    {
+      "epoch": 9.738019169329073,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6096
+    },
+    {
+      "epoch": 9.73961661341853,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 6097
+    },
+    {
+      "epoch": 9.741214057507987,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 6098
+    },
+    {
+      "epoch": 9.742811501597444,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 6099
+    },
+    {
+      "epoch": 9.744408945686901,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 6100
+    },
+    {
+      "epoch": 9.746006389776358,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 6101
+    },
+    {
+      "epoch": 9.747603833865815,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6102
+    },
+    {
+      "epoch": 9.749201277955272,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 6103
+    },
+    {
+      "epoch": 9.750798722044728,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6104
+    },
+    {
+      "epoch": 9.752396166134185,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6105
+    },
+    {
+      "epoch": 9.753993610223642,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6106
+    },
+    {
+      "epoch": 9.755591054313099,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 6107
+    },
+    {
+      "epoch": 9.757188498402556,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6108
+    },
+    {
+      "epoch": 9.758785942492013,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 6109
+    },
+    {
+      "epoch": 9.76038338658147,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 6110
+    },
+    {
+      "epoch": 9.761980830670927,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 6111
+    },
+    {
+      "epoch": 9.763578274760384,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 6112
+    },
+    {
+      "epoch": 9.76517571884984,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 6113
+    },
+    {
+      "epoch": 9.766773162939296,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 6114
+    },
+    {
+      "epoch": 9.768370607028753,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6115
+    },
+    {
+      "epoch": 9.76996805111821,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6116
+    },
+    {
+      "epoch": 9.771565495207668,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 6117
+    },
+    {
+      "epoch": 9.773162939297125,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6118
+    },
+    {
+      "epoch": 9.774760383386582,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 6119
+    },
+    {
+      "epoch": 9.776357827476039,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 6120
+    },
+    {
+      "epoch": 9.777955271565496,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6121
+    },
+    {
+      "epoch": 9.779552715654953,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 6122
+    },
+    {
+      "epoch": 9.781150159744408,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1505,
+      "step": 6123
+    },
+    {
+      "epoch": 9.782747603833865,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6124
+    },
+    {
+      "epoch": 9.784345047923322,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 6125
+    },
+    {
+      "epoch": 9.78594249201278,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6126
+    },
+    {
+      "epoch": 9.787539936102236,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 6127
+    },
+    {
+      "epoch": 9.789137380191693,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 6128
+    },
+    {
+      "epoch": 9.79073482428115,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 6129
+    },
+    {
+      "epoch": 9.792332268370608,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 6130
+    },
+    {
+      "epoch": 9.793929712460065,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 6131
+    },
+    {
+      "epoch": 9.795527156549522,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 6132
+    },
+    {
+      "epoch": 9.797124600638977,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 6133
+    },
+    {
+      "epoch": 9.798722044728434,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 6134
+    },
+    {
+      "epoch": 9.800319488817891,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6135
+    },
+    {
+      "epoch": 9.801916932907348,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 6136
+    },
+    {
+      "epoch": 9.803514376996805,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6137
+    },
+    {
+      "epoch": 9.805111821086262,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 6138
+    },
+    {
+      "epoch": 9.80670926517572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6139
+    },
+    {
+      "epoch": 9.808306709265176,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6140
+    },
+    {
+      "epoch": 9.809904153354633,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6141
+    },
+    {
+      "epoch": 9.811501597444089,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 6142
+    },
+    {
+      "epoch": 9.813099041533546,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6143
+    },
+    {
+      "epoch": 9.814696485623003,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6144
+    },
+    {
+      "epoch": 9.81629392971246,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 6145
+    },
+    {
+      "epoch": 9.817891373801917,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6146
+    },
+    {
+      "epoch": 9.819488817891374,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6147
+    },
+    {
+      "epoch": 9.821086261980831,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 6148
+    },
+    {
+      "epoch": 9.822683706070288,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6149
+    },
+    {
+      "epoch": 9.824281150159745,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6150
+    },
+    {
+      "epoch": 9.8258785942492,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 6151
+    },
+    {
+      "epoch": 9.827476038338657,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 6152
+    },
+    {
+      "epoch": 9.829073482428115,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6153
+    },
+    {
+      "epoch": 9.830670926517572,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 6154
+    },
+    {
+      "epoch": 9.832268370607029,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 6155
+    },
+    {
+      "epoch": 9.833865814696486,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6156
+    },
+    {
+      "epoch": 9.835463258785943,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6157
+    },
+    {
+      "epoch": 9.8370607028754,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6158
+    },
+    {
+      "epoch": 9.838658146964857,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6159
+    },
+    {
+      "epoch": 9.840255591054314,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 6160
+    },
+    {
+      "epoch": 9.84185303514377,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6161
+    },
+    {
+      "epoch": 9.843450479233226,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 6162
+    },
+    {
+      "epoch": 9.845047923322683,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 6163
+    },
+    {
+      "epoch": 9.84664536741214,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 6164
+    },
+    {
+      "epoch": 9.848242811501597,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 6165
+    },
+    {
+      "epoch": 9.849840255591054,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6166
+    },
+    {
+      "epoch": 9.851437699680512,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 6167
+    },
+    {
+      "epoch": 9.853035143769969,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6168
+    },
+    {
+      "epoch": 9.854632587859426,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 6169
+    },
+    {
+      "epoch": 9.856230031948883,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 6170
+    },
+    {
+      "epoch": 9.857827476038338,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6171
+    },
+    {
+      "epoch": 9.859424920127795,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 6172
+    },
+    {
+      "epoch": 9.861022364217252,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 6173
+    },
+    {
+      "epoch": 9.86261980830671,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 6174
+    },
+    {
+      "epoch": 9.864217252396166,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6175
+    },
+    {
+      "epoch": 9.865814696485623,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6176
+    },
+    {
+      "epoch": 9.86741214057508,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6177
+    },
+    {
+      "epoch": 9.869009584664537,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 6178
+    },
+    {
+      "epoch": 9.870607028753994,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 6179
+    },
+    {
+      "epoch": 9.87220447284345,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6180
+    },
+    {
+      "epoch": 9.873801916932907,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 6181
+    },
+    {
+      "epoch": 9.875399361022364,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6182
+    },
+    {
+      "epoch": 9.87699680511182,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 6183
+    },
+    {
+      "epoch": 9.878594249201278,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 6184
+    },
+    {
+      "epoch": 9.880191693290735,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6185
+    },
+    {
+      "epoch": 9.881789137380192,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6186
+    },
+    {
+      "epoch": 9.883386581469649,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6187
+    },
+    {
+      "epoch": 9.884984025559106,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6188
+    },
+    {
+      "epoch": 9.886581469648561,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 6189
+    },
+    {
+      "epoch": 9.888178913738018,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6190
+    },
+    {
+      "epoch": 9.889776357827476,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 6191
+    },
+    {
+      "epoch": 9.891373801916933,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6192
+    },
+    {
+      "epoch": 9.89297124600639,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6193
+    },
+    {
+      "epoch": 9.894568690095847,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6194
+    },
+    {
+      "epoch": 9.896166134185304,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6195
+    },
+    {
+      "epoch": 9.89776357827476,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6196
+    },
+    {
+      "epoch": 9.899361022364218,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6197
+    },
+    {
+      "epoch": 9.900958466453675,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6198
+    },
+    {
+      "epoch": 9.90255591054313,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 6199
+    },
+    {
+      "epoch": 9.904153354632587,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6200
+    },
+    {
+      "epoch": 9.905750798722044,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6201
+    },
+    {
+      "epoch": 9.907348242811501,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6202
+    },
+    {
+      "epoch": 9.908945686900958,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6203
+    },
+    {
+      "epoch": 9.910543130990416,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 6204
+    },
+    {
+      "epoch": 9.912140575079873,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 6205
+    },
+    {
+      "epoch": 9.91373801916933,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 6206
+    },
+    {
+      "epoch": 9.915335463258787,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 6207
+    },
+    {
+      "epoch": 9.916932907348244,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 6208
+    },
+    {
+      "epoch": 9.918530351437699,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 6209
+    },
+    {
+      "epoch": 9.920127795527156,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 6210
+    },
+    {
+      "epoch": 9.921725239616613,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 6211
+    },
+    {
+      "epoch": 9.92332268370607,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 6212
+    },
+    {
+      "epoch": 9.924920127795527,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6213
+    },
+    {
+      "epoch": 9.926517571884984,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6214
+    },
+    {
+      "epoch": 9.928115015974441,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6215
+    },
+    {
+      "epoch": 9.929712460063898,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 6216
+    },
+    {
+      "epoch": 9.931309904153355,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 6217
+    },
+    {
+      "epoch": 9.93290734824281,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6218
+    },
+    {
+      "epoch": 9.934504792332268,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6219
+    },
+    {
+      "epoch": 9.936102236421725,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 6220
+    },
+    {
+      "epoch": 9.937699680511182,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6221
+    },
+    {
+      "epoch": 9.939297124600639,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 6222
+    },
+    {
+      "epoch": 9.940894568690096,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6223
+    },
+    {
+      "epoch": 9.942492012779553,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 6224
+    },
+    {
+      "epoch": 9.94408945686901,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 6225
+    },
+    {
+      "epoch": 9.945686900958467,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 6226
+    },
+    {
+      "epoch": 9.947284345047922,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6227
+    },
+    {
+      "epoch": 9.94888178913738,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6228
+    },
+    {
+      "epoch": 9.950479233226837,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6229
+    },
+    {
+      "epoch": 9.952076677316294,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1502,
+      "step": 6230
+    },
+    {
+      "epoch": 9.95367412140575,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6231
+    },
+    {
+      "epoch": 9.955271565495208,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 6232
+    },
+    {
+      "epoch": 9.956869009584665,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 6233
+    },
+    {
+      "epoch": 9.958466453674122,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 6234
+    },
+    {
+      "epoch": 9.960063897763579,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 6235
+    },
+    {
+      "epoch": 9.961661341853034,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 6236
+    },
+    {
+      "epoch": 9.963258785942491,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6237
+    },
+    {
+      "epoch": 9.964856230031948,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6238
+    },
+    {
+      "epoch": 9.966453674121405,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 6239
+    },
+    {
+      "epoch": 9.968051118210862,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 6240
+    },
+    {
+      "epoch": 9.96964856230032,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6241
+    },
+    {
+      "epoch": 9.971246006389777,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 6242
+    },
+    {
+      "epoch": 9.972843450479234,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6243
+    },
+    {
+      "epoch": 9.97444089456869,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6244
+    },
+    {
+      "epoch": 9.976038338658148,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 6245
+    },
+    {
+      "epoch": 9.977635782747605,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6246
+    },
+    {
+      "epoch": 9.97923322683706,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6247
+    },
+    {
+      "epoch": 9.980830670926517,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6248
+    },
+    {
+      "epoch": 9.982428115015974,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 6249
+    },
+    {
+      "epoch": 9.984025559105431,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6250
+    },
+    {
+      "epoch": 9.985623003194888,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 6251
+    },
+    {
+      "epoch": 9.987220447284345,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 6252
+    },
+    {
+      "epoch": 9.988817891373802,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6253
+    },
+    {
+      "epoch": 9.99041533546326,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 6254
+    },
+    {
+      "epoch": 9.992012779552716,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 6255
+    },
+    {
+      "epoch": 9.993610223642172,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 6256
+    },
+    {
+      "epoch": 9.995207667731629,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 6257
+    },
+    {
+      "epoch": 9.996805111821086,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6258
+    },
+    {
+      "epoch": 9.998402555910543,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 6259
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 6260
+    },
+    {
+      "epoch": 10.001597444089457,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 6261
+    },
+    {
+      "epoch": 10.003194888178914,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6262
+    },
+    {
+      "epoch": 10.004792332268371,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 6263
+    },
+    {
+      "epoch": 10.006389776357828,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6264
+    },
+    {
+      "epoch": 10.007987220447284,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 6265
+    },
+    {
+      "epoch": 10.00958466453674,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6266
+    },
+    {
+      "epoch": 10.011182108626198,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6267
+    },
+    {
+      "epoch": 10.012779552715655,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6268
+    },
+    {
+      "epoch": 10.014376996805112,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 6269
+    },
+    {
+      "epoch": 10.015974440894569,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6270
+    },
+    {
+      "epoch": 10.017571884984026,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 6271
+    },
+    {
+      "epoch": 10.019169329073483,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6272
+    },
+    {
+      "epoch": 10.02076677316294,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 6273
+    },
+    {
+      "epoch": 10.022364217252397,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 6274
+    },
+    {
+      "epoch": 10.023961661341852,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 6275
+    },
+    {
+      "epoch": 10.02555910543131,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6276
+    },
+    {
+      "epoch": 10.027156549520766,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 6277
+    },
+    {
+      "epoch": 10.028753993610223,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 6278
+    },
+    {
+      "epoch": 10.03035143769968,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 6279
+    },
+    {
+      "epoch": 10.031948881789138,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 6280
+    },
+    {
+      "epoch": 10.033546325878595,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6281
+    },
+    {
+      "epoch": 10.035143769968052,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1466,
+      "step": 6282
+    },
+    {
+      "epoch": 10.036741214057509,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 6283
+    },
+    {
+      "epoch": 10.038338658146964,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 6284
+    },
+    {
+      "epoch": 10.039936102236421,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 6285
+    },
+    {
+      "epoch": 10.041533546325878,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 6286
+    },
+    {
+      "epoch": 10.043130990415335,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 6287
+    },
+    {
+      "epoch": 10.044728434504792,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 6288
+    },
+    {
+      "epoch": 10.04632587859425,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 6289
+    },
+    {
+      "epoch": 10.047923322683706,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 6290
+    },
+    {
+      "epoch": 10.049520766773163,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 6291
+    },
+    {
+      "epoch": 10.05111821086262,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 6292
+    },
+    {
+      "epoch": 10.052715654952078,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 6293
+    },
+    {
+      "epoch": 10.054313099041533,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 6294
+    },
+    {
+      "epoch": 10.05591054313099,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6295
+    },
+    {
+      "epoch": 10.057507987220447,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 6296
+    },
+    {
+      "epoch": 10.059105431309904,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6297
+    },
+    {
+      "epoch": 10.060702875399361,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 6298
+    },
+    {
+      "epoch": 10.062300319488818,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 6299
+    },
+    {
+      "epoch": 10.063897763578275,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6300
+    },
+    {
+      "epoch": 10.065495207667732,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 6301
+    },
+    {
+      "epoch": 10.06709265175719,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6302
+    },
+    {
+      "epoch": 10.068690095846645,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6303
+    },
+    {
+      "epoch": 10.070287539936102,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 6304
+    },
+    {
+      "epoch": 10.071884984025559,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 6305
+    },
+    {
+      "epoch": 10.073482428115016,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 6306
+    },
+    {
+      "epoch": 10.075079872204473,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 6307
+    },
+    {
+      "epoch": 10.07667731629393,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6308
+    },
+    {
+      "epoch": 10.078274760383387,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6309
+    },
+    {
+      "epoch": 10.079872204472844,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 6310
+    },
+    {
+      "epoch": 10.081469648562301,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6311
+    },
+    {
+      "epoch": 10.083067092651758,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 6312
+    },
+    {
+      "epoch": 10.084664536741213,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 6313
+    },
+    {
+      "epoch": 10.08626198083067,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 6314
+    },
+    {
+      "epoch": 10.087859424920127,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 6315
+    },
+    {
+      "epoch": 10.089456869009584,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 6316
+    },
+    {
+      "epoch": 10.091054313099042,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6317
+    },
+    {
+      "epoch": 10.092651757188499,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 6318
+    },
+    {
+      "epoch": 10.094249201277956,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6319
+    },
+    {
+      "epoch": 10.095846645367413,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 6320
+    },
+    {
+      "epoch": 10.09744408945687,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6321
+    },
+    {
+      "epoch": 10.099041533546325,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 6322
+    },
+    {
+      "epoch": 10.100638977635782,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 6323
+    },
+    {
+      "epoch": 10.10223642172524,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 6324
+    },
+    {
+      "epoch": 10.103833865814696,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 6325
+    },
+    {
+      "epoch": 10.105431309904153,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6326
+    },
+    {
+      "epoch": 10.10702875399361,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 6327
+    },
+    {
+      "epoch": 10.108626198083067,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6328
+    },
+    {
+      "epoch": 10.110223642172524,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 6329
+    },
+    {
+      "epoch": 10.111821086261982,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6330
+    },
+    {
+      "epoch": 10.113418530351439,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6331
+    },
+    {
+      "epoch": 10.115015974440894,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 6332
+    },
+    {
+      "epoch": 10.116613418530351,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 6333
+    },
+    {
+      "epoch": 10.118210862619808,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6334
+    },
+    {
+      "epoch": 10.119808306709265,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6335
+    },
+    {
+      "epoch": 10.121405750798722,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6336
+    },
+    {
+      "epoch": 10.12300319488818,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6337
+    },
+    {
+      "epoch": 10.124600638977636,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6338
+    },
+    {
+      "epoch": 10.126198083067093,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1475,
+      "step": 6339
+    },
+    {
+      "epoch": 10.12779552715655,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 6340
+    },
+    {
+      "epoch": 10.129392971246006,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 6341
+    },
+    {
+      "epoch": 10.130990415335463,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6342
+    },
+    {
+      "epoch": 10.13258785942492,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6343
+    },
+    {
+      "epoch": 10.134185303514377,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 6344
+    },
+    {
+      "epoch": 10.135782747603834,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 6345
+    },
+    {
+      "epoch": 10.13738019169329,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 6346
+    },
+    {
+      "epoch": 10.138977635782748,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 6347
+    },
+    {
+      "epoch": 10.140575079872205,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 6348
+    },
+    {
+      "epoch": 10.142172523961662,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 6349
+    },
+    {
+      "epoch": 10.143769968051119,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6350
+    },
+    {
+      "epoch": 10.145367412140574,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 6351
+    },
+    {
+      "epoch": 10.146964856230031,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 6352
+    },
+    {
+      "epoch": 10.148562300319488,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 6353
+    },
+    {
+      "epoch": 10.150159744408946,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6354
+    },
+    {
+      "epoch": 10.151757188498403,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6355
+    },
+    {
+      "epoch": 10.15335463258786,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6356
+    },
+    {
+      "epoch": 10.154952076677317,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6357
+    },
+    {
+      "epoch": 10.156549520766774,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 6358
+    },
+    {
+      "epoch": 10.15814696485623,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 6359
+    },
+    {
+      "epoch": 10.159744408945686,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6360
+    },
+    {
+      "epoch": 10.161341853035143,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 6361
+    },
+    {
+      "epoch": 10.1629392971246,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 6362
+    },
+    {
+      "epoch": 10.164536741214057,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6363
+    },
+    {
+      "epoch": 10.166134185303514,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6364
+    },
+    {
+      "epoch": 10.167731629392971,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1201,
+      "step": 6365
+    },
+    {
+      "epoch": 10.169329073482428,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 6366
+    },
+    {
+      "epoch": 10.170926517571885,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6367
+    },
+    {
+      "epoch": 10.172523961661343,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 6368
+    },
+    {
+      "epoch": 10.1741214057508,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1457,
+      "step": 6369
+    },
+    {
+      "epoch": 10.175718849840255,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6370
+    },
+    {
+      "epoch": 10.177316293929712,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6371
+    },
+    {
+      "epoch": 10.178913738019169,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6372
+    },
+    {
+      "epoch": 10.180511182108626,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 6373
+    },
+    {
+      "epoch": 10.182108626198083,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 6374
+    },
+    {
+      "epoch": 10.18370607028754,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6375
+    },
+    {
+      "epoch": 10.185303514376997,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 6376
+    },
+    {
+      "epoch": 10.186900958466454,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 6377
+    },
+    {
+      "epoch": 10.188498402555911,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6378
+    },
+    {
+      "epoch": 10.190095846645367,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 6379
+    },
+    {
+      "epoch": 10.191693290734824,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 6380
+    },
+    {
+      "epoch": 10.19329073482428,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 6381
+    },
+    {
+      "epoch": 10.194888178913738,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 6382
+    },
+    {
+      "epoch": 10.196485623003195,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6383
+    },
+    {
+      "epoch": 10.198083067092652,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6384
+    },
+    {
+      "epoch": 10.199680511182109,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 6385
+    },
+    {
+      "epoch": 10.201277955271566,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 6386
+    },
+    {
+      "epoch": 10.202875399361023,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6387
+    },
+    {
+      "epoch": 10.204472843450478,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6388
+    },
+    {
+      "epoch": 10.206070287539935,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 6389
+    },
+    {
+      "epoch": 10.207667731629392,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6390
+    },
+    {
+      "epoch": 10.20926517571885,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6391
+    },
+    {
+      "epoch": 10.210862619808307,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 6392
+    },
+    {
+      "epoch": 10.212460063897764,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 6393
+    },
+    {
+      "epoch": 10.21405750798722,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6394
+    },
+    {
+      "epoch": 10.215654952076678,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6395
+    },
+    {
+      "epoch": 10.217252396166135,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 6396
+    },
+    {
+      "epoch": 10.218849840255592,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6397
+    },
+    {
+      "epoch": 10.220447284345047,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6398
+    },
+    {
+      "epoch": 10.222044728434504,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 6399
+    },
+    {
+      "epoch": 10.223642172523961,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 6400
+    },
+    {
+      "epoch": 10.225239616613418,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 6401
+    },
+    {
+      "epoch": 10.226837060702875,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6402
+    },
+    {
+      "epoch": 10.228434504792332,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6403
+    },
+    {
+      "epoch": 10.23003194888179,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6404
+    },
+    {
+      "epoch": 10.231629392971247,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 6405
+    },
+    {
+      "epoch": 10.233226837060704,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6406
+    },
+    {
+      "epoch": 10.23482428115016,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6407
+    },
+    {
+      "epoch": 10.236421725239616,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 6408
+    },
+    {
+      "epoch": 10.238019169329073,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6409
+    },
+    {
+      "epoch": 10.23961661341853,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6410
+    },
+    {
+      "epoch": 10.241214057507987,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 6411
+    },
+    {
+      "epoch": 10.242811501597444,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6412
+    },
+    {
+      "epoch": 10.244408945686901,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 6413
+    },
+    {
+      "epoch": 10.246006389776358,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 6414
+    },
+    {
+      "epoch": 10.247603833865815,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6415
+    },
+    {
+      "epoch": 10.249201277955272,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6416
+    },
+    {
+      "epoch": 10.250798722044728,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6417
+    },
+    {
+      "epoch": 10.252396166134185,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 6418
+    },
+    {
+      "epoch": 10.253993610223642,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6419
+    },
+    {
+      "epoch": 10.255591054313099,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 6420
+    },
+    {
+      "epoch": 10.257188498402556,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 6421
+    },
+    {
+      "epoch": 10.258785942492013,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 6422
+    },
+    {
+      "epoch": 10.26038338658147,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 6423
+    },
+    {
+      "epoch": 10.261980830670927,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6424
+    },
+    {
+      "epoch": 10.263578274760384,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 6425
+    },
+    {
+      "epoch": 10.26517571884984,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6426
+    },
+    {
+      "epoch": 10.266773162939296,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6427
+    },
+    {
+      "epoch": 10.268370607028753,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 6428
+    },
+    {
+      "epoch": 10.26996805111821,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 6429
+    },
+    {
+      "epoch": 10.271565495207668,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6430
+    },
+    {
+      "epoch": 10.273162939297125,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 6431
+    },
+    {
+      "epoch": 10.274760383386582,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 6432
+    },
+    {
+      "epoch": 10.276357827476039,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1552,
+      "step": 6433
+    },
+    {
+      "epoch": 10.277955271565496,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 6434
+    },
+    {
+      "epoch": 10.279552715654953,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 6435
+    },
+    {
+      "epoch": 10.281150159744408,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 6436
+    },
+    {
+      "epoch": 10.282747603833865,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 6437
+    },
+    {
+      "epoch": 10.284345047923322,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1461,
+      "step": 6438
+    },
+    {
+      "epoch": 10.28594249201278,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 6439
+    },
+    {
+      "epoch": 10.287539936102236,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 6440
+    },
+    {
+      "epoch": 10.289137380191693,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.123,
+      "step": 6441
+    },
+    {
+      "epoch": 10.29073482428115,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 6442
+    },
+    {
+      "epoch": 10.292332268370608,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 6443
+    },
+    {
+      "epoch": 10.293929712460065,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 6444
+    },
+    {
+      "epoch": 10.295527156549522,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 6445
+    },
+    {
+      "epoch": 10.297124600638977,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 6446
+    },
+    {
+      "epoch": 10.298722044728434,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6447
+    },
+    {
+      "epoch": 10.300319488817891,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 6448
+    },
+    {
+      "epoch": 10.301916932907348,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6449
+    },
+    {
+      "epoch": 10.303514376996805,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1462,
+      "step": 6450
+    },
+    {
+      "epoch": 10.305111821086262,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6451
+    },
+    {
+      "epoch": 10.30670926517572,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 6452
+    },
+    {
+      "epoch": 10.308306709265176,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6453
+    },
+    {
+      "epoch": 10.309904153354633,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 6454
+    },
+    {
+      "epoch": 10.311501597444089,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6455
+    },
+    {
+      "epoch": 10.313099041533546,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 6456
+    },
+    {
+      "epoch": 10.314696485623003,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 6457
+    },
+    {
+      "epoch": 10.31629392971246,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6458
+    },
+    {
+      "epoch": 10.317891373801917,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6459
+    },
+    {
+      "epoch": 10.319488817891374,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 6460
+    },
+    {
+      "epoch": 10.321086261980831,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 6461
+    },
+    {
+      "epoch": 10.322683706070288,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 6462
+    },
+    {
+      "epoch": 10.324281150159745,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6463
+    },
+    {
+      "epoch": 10.3258785942492,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6464
+    },
+    {
+      "epoch": 10.327476038338657,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6465
+    },
+    {
+      "epoch": 10.329073482428115,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6466
+    },
+    {
+      "epoch": 10.330670926517572,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6467
+    },
+    {
+      "epoch": 10.332268370607029,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 6468
+    },
+    {
+      "epoch": 10.333865814696486,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6469
+    },
+    {
+      "epoch": 10.335463258785943,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6470
+    },
+    {
+      "epoch": 10.3370607028754,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 6471
+    },
+    {
+      "epoch": 10.338658146964857,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6472
+    },
+    {
+      "epoch": 10.340255591054314,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 6473
+    },
+    {
+      "epoch": 10.34185303514377,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 6474
+    },
+    {
+      "epoch": 10.343450479233226,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6475
+    },
+    {
+      "epoch": 10.345047923322683,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6476
+    },
+    {
+      "epoch": 10.34664536741214,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6477
+    },
+    {
+      "epoch": 10.348242811501597,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6478
+    },
+    {
+      "epoch": 10.349840255591054,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 6479
+    },
+    {
+      "epoch": 10.351437699680512,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6480
+    },
+    {
+      "epoch": 10.353035143769969,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6481
+    },
+    {
+      "epoch": 10.354632587859426,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 6482
+    },
+    {
+      "epoch": 10.356230031948883,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 6483
+    },
+    {
+      "epoch": 10.357827476038338,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6484
+    },
+    {
+      "epoch": 10.359424920127795,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 6485
+    },
+    {
+      "epoch": 10.361022364217252,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6486
+    },
+    {
+      "epoch": 10.36261980830671,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 6487
+    },
+    {
+      "epoch": 10.364217252396166,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6488
+    },
+    {
+      "epoch": 10.365814696485623,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 6489
+    },
+    {
+      "epoch": 10.36741214057508,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 6490
+    },
+    {
+      "epoch": 10.369009584664537,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1214,
+      "step": 6491
+    },
+    {
+      "epoch": 10.370607028753994,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 6492
+    },
+    {
+      "epoch": 10.37220447284345,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6493
+    },
+    {
+      "epoch": 10.373801916932907,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 6494
+    },
+    {
+      "epoch": 10.375399361022364,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 6495
+    },
+    {
+      "epoch": 10.37699680511182,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6496
+    },
+    {
+      "epoch": 10.378594249201278,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 6497
+    },
+    {
+      "epoch": 10.380191693290735,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6498
+    },
+    {
+      "epoch": 10.381789137380192,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6499
+    },
+    {
+      "epoch": 10.383386581469649,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6500
+    },
+    {
+      "epoch": 10.384984025559106,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 6501
+    },
+    {
+      "epoch": 10.386581469648561,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 6502
+    },
+    {
+      "epoch": 10.388178913738018,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 6503
+    },
+    {
+      "epoch": 10.389776357827476,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6504
+    },
+    {
+      "epoch": 10.391373801916933,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 6505
+    },
+    {
+      "epoch": 10.39297124600639,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 6506
+    },
+    {
+      "epoch": 10.394568690095847,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 6507
+    },
+    {
+      "epoch": 10.396166134185304,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 6508
+    },
+    {
+      "epoch": 10.39776357827476,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6509
+    },
+    {
+      "epoch": 10.399361022364218,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 6510
+    },
+    {
+      "epoch": 10.400958466453675,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 6511
+    },
+    {
+      "epoch": 10.40255591054313,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 6512
+    },
+    {
+      "epoch": 10.404153354632587,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 6513
+    },
+    {
+      "epoch": 10.405750798722044,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 6514
+    },
+    {
+      "epoch": 10.407348242811501,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6515
+    },
+    {
+      "epoch": 10.408945686900958,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 6516
+    },
+    {
+      "epoch": 10.410543130990416,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 6517
+    },
+    {
+      "epoch": 10.412140575079873,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6518
+    },
+    {
+      "epoch": 10.41373801916933,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 6519
+    },
+    {
+      "epoch": 10.415335463258787,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6520
+    },
+    {
+      "epoch": 10.416932907348242,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6521
+    },
+    {
+      "epoch": 10.418530351437699,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6522
+    },
+    {
+      "epoch": 10.420127795527156,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 6523
+    },
+    {
+      "epoch": 10.421725239616613,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 6524
+    },
+    {
+      "epoch": 10.42332268370607,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6525
+    },
+    {
+      "epoch": 10.424920127795527,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 6526
+    },
+    {
+      "epoch": 10.426517571884984,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 6527
+    },
+    {
+      "epoch": 10.428115015974441,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6528
+    },
+    {
+      "epoch": 10.429712460063898,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6529
+    },
+    {
+      "epoch": 10.431309904153355,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 6530
+    },
+    {
+      "epoch": 10.43290734824281,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6531
+    },
+    {
+      "epoch": 10.434504792332268,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 6532
+    },
+    {
+      "epoch": 10.436102236421725,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 6533
+    },
+    {
+      "epoch": 10.437699680511182,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 6534
+    },
+    {
+      "epoch": 10.439297124600639,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 6535
+    },
+    {
+      "epoch": 10.440894568690096,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6536
+    },
+    {
+      "epoch": 10.442492012779553,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 6537
+    },
+    {
+      "epoch": 10.44408945686901,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6538
+    },
+    {
+      "epoch": 10.445686900958467,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6539
+    },
+    {
+      "epoch": 10.447284345047922,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6540
+    },
+    {
+      "epoch": 10.44888178913738,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 6541
+    },
+    {
+      "epoch": 10.450479233226837,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1224,
+      "step": 6542
+    },
+    {
+      "epoch": 10.452076677316294,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6543
+    },
+    {
+      "epoch": 10.45367412140575,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 6544
+    },
+    {
+      "epoch": 10.455271565495208,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 6545
+    },
+    {
+      "epoch": 10.456869009584665,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6546
+    },
+    {
+      "epoch": 10.458466453674122,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 6547
+    },
+    {
+      "epoch": 10.460063897763579,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 6548
+    },
+    {
+      "epoch": 10.461661341853036,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 6549
+    },
+    {
+      "epoch": 10.463258785942491,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6550
+    },
+    {
+      "epoch": 10.464856230031948,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6551
+    },
+    {
+      "epoch": 10.466453674121405,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6552
+    },
+    {
+      "epoch": 10.468051118210862,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6553
+    },
+    {
+      "epoch": 10.46964856230032,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 6554
+    },
+    {
+      "epoch": 10.471246006389777,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 6555
+    },
+    {
+      "epoch": 10.472843450479234,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 6556
+    },
+    {
+      "epoch": 10.47444089456869,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6557
+    },
+    {
+      "epoch": 10.476038338658148,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6558
+    },
+    {
+      "epoch": 10.477635782747603,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 6559
+    },
+    {
+      "epoch": 10.47923322683706,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 6560
+    },
+    {
+      "epoch": 10.480830670926517,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6561
+    },
+    {
+      "epoch": 10.482428115015974,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 6562
+    },
+    {
+      "epoch": 10.484025559105431,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 6563
+    },
+    {
+      "epoch": 10.485623003194888,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6564
+    },
+    {
+      "epoch": 10.487220447284345,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 6565
+    },
+    {
+      "epoch": 10.488817891373802,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1467,
+      "step": 6566
+    },
+    {
+      "epoch": 10.49041533546326,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6567
+    },
+    {
+      "epoch": 10.492012779552716,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 6568
+    },
+    {
+      "epoch": 10.493610223642172,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 6569
+    },
+    {
+      "epoch": 10.495207667731629,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6570
+    },
+    {
+      "epoch": 10.496805111821086,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 6571
+    },
+    {
+      "epoch": 10.498402555910543,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6572
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 6573
+    },
+    {
+      "epoch": 10.501597444089457,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1433,
+      "step": 6574
+    },
+    {
+      "epoch": 10.503194888178914,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6575
+    },
+    {
+      "epoch": 10.504792332268371,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6576
+    },
+    {
+      "epoch": 10.506389776357828,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6577
+    },
+    {
+      "epoch": 10.507987220447284,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6578
+    },
+    {
+      "epoch": 10.50958466453674,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6579
+    },
+    {
+      "epoch": 10.511182108626198,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 6580
+    },
+    {
+      "epoch": 10.512779552715655,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6581
+    },
+    {
+      "epoch": 10.514376996805112,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 6582
+    },
+    {
+      "epoch": 10.515974440894569,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6583
+    },
+    {
+      "epoch": 10.517571884984026,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 6584
+    },
+    {
+      "epoch": 10.519169329073483,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6585
+    },
+    {
+      "epoch": 10.52076677316294,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 6586
+    },
+    {
+      "epoch": 10.522364217252395,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 6587
+    },
+    {
+      "epoch": 10.523961661341852,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 6588
+    },
+    {
+      "epoch": 10.52555910543131,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 6589
+    },
+    {
+      "epoch": 10.527156549520766,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 6590
+    },
+    {
+      "epoch": 10.528753993610223,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 6591
+    },
+    {
+      "epoch": 10.53035143769968,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 6592
+    },
+    {
+      "epoch": 10.531948881789138,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 6593
+    },
+    {
+      "epoch": 10.533546325878595,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 6594
+    },
+    {
+      "epoch": 10.535143769968052,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 6595
+    },
+    {
+      "epoch": 10.536741214057509,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 6596
+    },
+    {
+      "epoch": 10.538338658146966,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6597
+    },
+    {
+      "epoch": 10.539936102236421,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 6598
+    },
+    {
+      "epoch": 10.541533546325878,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 6599
+    },
+    {
+      "epoch": 10.543130990415335,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 6600
+    },
+    {
+      "epoch": 10.544728434504792,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 6601
+    },
+    {
+      "epoch": 10.54632587859425,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6602
+    },
+    {
+      "epoch": 10.547923322683706,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6603
+    },
+    {
+      "epoch": 10.549520766773163,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 6604
+    },
+    {
+      "epoch": 10.55111821086262,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6605
+    },
+    {
+      "epoch": 10.552715654952078,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 6606
+    },
+    {
+      "epoch": 10.554313099041533,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6607
+    },
+    {
+      "epoch": 10.55591054313099,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 6608
+    },
+    {
+      "epoch": 10.557507987220447,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 6609
+    },
+    {
+      "epoch": 10.559105431309904,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6610
+    },
+    {
+      "epoch": 10.560702875399361,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6611
+    },
+    {
+      "epoch": 10.562300319488818,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6612
+    },
+    {
+      "epoch": 10.563897763578275,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6613
+    },
+    {
+      "epoch": 10.565495207667732,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 6614
+    },
+    {
+      "epoch": 10.56709265175719,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6615
+    },
+    {
+      "epoch": 10.568690095846645,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6616
+    },
+    {
+      "epoch": 10.570287539936102,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6617
+    },
+    {
+      "epoch": 10.571884984025559,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 6618
+    },
+    {
+      "epoch": 10.573482428115016,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6619
+    },
+    {
+      "epoch": 10.575079872204473,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6620
+    },
+    {
+      "epoch": 10.57667731629393,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6621
+    },
+    {
+      "epoch": 10.578274760383387,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6622
+    },
+    {
+      "epoch": 10.579872204472844,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6623
+    },
+    {
+      "epoch": 10.581469648562301,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 6624
+    },
+    {
+      "epoch": 10.583067092651756,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 6625
+    },
+    {
+      "epoch": 10.584664536741213,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 6626
+    },
+    {
+      "epoch": 10.58626198083067,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6627
+    },
+    {
+      "epoch": 10.587859424920127,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1471,
+      "step": 6628
+    },
+    {
+      "epoch": 10.589456869009584,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 6629
+    },
+    {
+      "epoch": 10.591054313099042,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 6630
+    },
+    {
+      "epoch": 10.592651757188499,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6631
+    },
+    {
+      "epoch": 10.594249201277956,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 6632
+    },
+    {
+      "epoch": 10.595846645367413,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6633
+    },
+    {
+      "epoch": 10.59744408945687,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6634
+    },
+    {
+      "epoch": 10.599041533546325,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6635
+    },
+    {
+      "epoch": 10.600638977635782,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 6636
+    },
+    {
+      "epoch": 10.60223642172524,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6637
+    },
+    {
+      "epoch": 10.603833865814696,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6638
+    },
+    {
+      "epoch": 10.605431309904153,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 6639
+    },
+    {
+      "epoch": 10.60702875399361,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 6640
+    },
+    {
+      "epoch": 10.608626198083067,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6641
+    },
+    {
+      "epoch": 10.610223642172524,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6642
+    },
+    {
+      "epoch": 10.611821086261982,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6643
+    },
+    {
+      "epoch": 10.613418530351439,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 6644
+    },
+    {
+      "epoch": 10.615015974440894,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6645
+    },
+    {
+      "epoch": 10.616613418530351,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6646
+    },
+    {
+      "epoch": 10.618210862619808,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6647
+    },
+    {
+      "epoch": 10.619808306709265,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 6648
+    },
+    {
+      "epoch": 10.621405750798722,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6649
+    },
+    {
+      "epoch": 10.62300319488818,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6650
+    },
+    {
+      "epoch": 10.624600638977636,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 6651
+    },
+    {
+      "epoch": 10.626198083067093,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6652
+    },
+    {
+      "epoch": 10.62779552715655,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6653
+    },
+    {
+      "epoch": 10.629392971246006,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6654
+    },
+    {
+      "epoch": 10.630990415335463,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 6655
+    },
+    {
+      "epoch": 10.63258785942492,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6656
+    },
+    {
+      "epoch": 10.634185303514377,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6657
+    },
+    {
+      "epoch": 10.635782747603834,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6658
+    },
+    {
+      "epoch": 10.63738019169329,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 6659
+    },
+    {
+      "epoch": 10.638977635782748,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 6660
+    },
+    {
+      "epoch": 10.640575079872205,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6661
+    },
+    {
+      "epoch": 10.642172523961662,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6662
+    },
+    {
+      "epoch": 10.643769968051117,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 6663
+    },
+    {
+      "epoch": 10.645367412140574,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6664
+    },
+    {
+      "epoch": 10.646964856230031,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6665
+    },
+    {
+      "epoch": 10.648562300319488,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 6666
+    },
+    {
+      "epoch": 10.650159744408946,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 6667
+    },
+    {
+      "epoch": 10.651757188498403,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 6668
+    },
+    {
+      "epoch": 10.65335463258786,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 6669
+    },
+    {
+      "epoch": 10.654952076677317,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6670
+    },
+    {
+      "epoch": 10.656549520766774,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6671
+    },
+    {
+      "epoch": 10.65814696485623,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6672
+    },
+    {
+      "epoch": 10.659744408945686,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 6673
+    },
+    {
+      "epoch": 10.661341853035143,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 6674
+    },
+    {
+      "epoch": 10.6629392971246,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 6675
+    },
+    {
+      "epoch": 10.664536741214057,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6676
+    },
+    {
+      "epoch": 10.666134185303514,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6677
+    },
+    {
+      "epoch": 10.667731629392971,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6678
+    },
+    {
+      "epoch": 10.669329073482428,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 6679
+    },
+    {
+      "epoch": 10.670926517571885,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 6680
+    },
+    {
+      "epoch": 10.672523961661343,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 6681
+    },
+    {
+      "epoch": 10.6741214057508,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6682
+    },
+    {
+      "epoch": 10.675718849840255,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 6683
+    },
+    {
+      "epoch": 10.677316293929712,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 6684
+    },
+    {
+      "epoch": 10.678913738019169,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6685
+    },
+    {
+      "epoch": 10.680511182108626,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6686
+    },
+    {
+      "epoch": 10.682108626198083,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6687
+    },
+    {
+      "epoch": 10.68370607028754,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 6688
+    },
+    {
+      "epoch": 10.685303514376997,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6689
+    },
+    {
+      "epoch": 10.686900958466454,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 6690
+    },
+    {
+      "epoch": 10.688498402555911,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6691
+    },
+    {
+      "epoch": 10.690095846645367,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6692
+    },
+    {
+      "epoch": 10.691693290734824,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 6693
+    },
+    {
+      "epoch": 10.69329073482428,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 6694
+    },
+    {
+      "epoch": 10.694888178913738,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6695
+    },
+    {
+      "epoch": 10.696485623003195,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 6696
+    },
+    {
+      "epoch": 10.698083067092652,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 6697
+    },
+    {
+      "epoch": 10.699680511182109,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 6698
+    },
+    {
+      "epoch": 10.701277955271566,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 6699
+    },
+    {
+      "epoch": 10.702875399361023,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6700
+    },
+    {
+      "epoch": 10.704472843450478,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 6701
+    },
+    {
+      "epoch": 10.706070287539935,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 6702
+    },
+    {
+      "epoch": 10.707667731629392,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6703
+    },
+    {
+      "epoch": 10.70926517571885,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 6704
+    },
+    {
+      "epoch": 10.710862619808307,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 6705
+    },
+    {
+      "epoch": 10.712460063897764,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6706
+    },
+    {
+      "epoch": 10.71405750798722,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 6707
+    },
+    {
+      "epoch": 10.715654952076678,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 6708
+    },
+    {
+      "epoch": 10.717252396166135,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6709
+    },
+    {
+      "epoch": 10.718849840255592,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 6710
+    },
+    {
+      "epoch": 10.720447284345047,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 6711
+    },
+    {
+      "epoch": 10.722044728434504,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 6712
+    },
+    {
+      "epoch": 10.723642172523961,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 6713
+    },
+    {
+      "epoch": 10.725239616613418,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6714
+    },
+    {
+      "epoch": 10.726837060702875,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6715
+    },
+    {
+      "epoch": 10.728434504792332,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6716
+    },
+    {
+      "epoch": 10.73003194888179,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 6717
+    },
+    {
+      "epoch": 10.731629392971247,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 6718
+    },
+    {
+      "epoch": 10.733226837060704,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 6719
+    },
+    {
+      "epoch": 10.73482428115016,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 6720
+    },
+    {
+      "epoch": 10.736421725239616,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 6721
+    },
+    {
+      "epoch": 10.738019169329073,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 6722
+    },
+    {
+      "epoch": 10.73961661341853,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 6723
+    },
+    {
+      "epoch": 10.741214057507987,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 6724
+    },
+    {
+      "epoch": 10.742811501597444,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6725
+    },
+    {
+      "epoch": 10.744408945686901,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 6726
+    },
+    {
+      "epoch": 10.746006389776358,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 6727
+    },
+    {
+      "epoch": 10.747603833865815,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 6728
+    },
+    {
+      "epoch": 10.749201277955272,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 6729
+    },
+    {
+      "epoch": 10.750798722044728,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6730
+    },
+    {
+      "epoch": 10.752396166134185,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6731
+    },
+    {
+      "epoch": 10.753993610223642,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6732
+    },
+    {
+      "epoch": 10.755591054313099,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6733
+    },
+    {
+      "epoch": 10.757188498402556,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 6734
+    },
+    {
+      "epoch": 10.758785942492013,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 6735
+    },
+    {
+      "epoch": 10.76038338658147,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6736
+    },
+    {
+      "epoch": 10.761980830670927,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6737
+    },
+    {
+      "epoch": 10.763578274760384,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6738
+    },
+    {
+      "epoch": 10.76517571884984,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 6739
+    },
+    {
+      "epoch": 10.766773162939296,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 6740
+    },
+    {
+      "epoch": 10.768370607028753,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6741
+    },
+    {
+      "epoch": 10.76996805111821,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 6742
+    },
+    {
+      "epoch": 10.771565495207668,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 6743
+    },
+    {
+      "epoch": 10.773162939297125,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 6744
+    },
+    {
+      "epoch": 10.774760383386582,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 6745
+    },
+    {
+      "epoch": 10.776357827476039,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 6746
+    },
+    {
+      "epoch": 10.777955271565496,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 6747
+    },
+    {
+      "epoch": 10.779552715654953,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 6748
+    },
+    {
+      "epoch": 10.781150159744408,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6749
+    },
+    {
+      "epoch": 10.782747603833865,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6750
+    },
+    {
+      "epoch": 10.784345047923322,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1454,
+      "step": 6751
+    },
+    {
+      "epoch": 10.78594249201278,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 6752
+    },
+    {
+      "epoch": 10.787539936102236,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6753
+    },
+    {
+      "epoch": 10.789137380191693,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 6754
+    },
+    {
+      "epoch": 10.79073482428115,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6755
+    },
+    {
+      "epoch": 10.792332268370608,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 6756
+    },
+    {
+      "epoch": 10.793929712460065,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 6757
+    },
+    {
+      "epoch": 10.795527156549522,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6758
+    },
+    {
+      "epoch": 10.797124600638977,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 6759
+    },
+    {
+      "epoch": 10.798722044728434,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 6760
+    },
+    {
+      "epoch": 10.800319488817891,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6761
+    },
+    {
+      "epoch": 10.801916932907348,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 6762
+    },
+    {
+      "epoch": 10.803514376996805,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6763
+    },
+    {
+      "epoch": 10.805111821086262,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 6764
+    },
+    {
+      "epoch": 10.80670926517572,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 6765
+    },
+    {
+      "epoch": 10.808306709265176,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6766
+    },
+    {
+      "epoch": 10.809904153354633,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 6767
+    },
+    {
+      "epoch": 10.811501597444089,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6768
+    },
+    {
+      "epoch": 10.813099041533546,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 6769
+    },
+    {
+      "epoch": 10.814696485623003,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6770
+    },
+    {
+      "epoch": 10.81629392971246,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6771
+    },
+    {
+      "epoch": 10.817891373801917,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6772
+    },
+    {
+      "epoch": 10.819488817891374,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6773
+    },
+    {
+      "epoch": 10.821086261980831,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6774
+    },
+    {
+      "epoch": 10.822683706070288,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 6775
+    },
+    {
+      "epoch": 10.824281150159745,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 6776
+    },
+    {
+      "epoch": 10.8258785942492,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6777
+    },
+    {
+      "epoch": 10.827476038338657,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 6778
+    },
+    {
+      "epoch": 10.829073482428115,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 6779
+    },
+    {
+      "epoch": 10.830670926517572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 6780
+    },
+    {
+      "epoch": 10.832268370607029,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 6781
+    },
+    {
+      "epoch": 10.833865814696486,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6782
+    },
+    {
+      "epoch": 10.835463258785943,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 6783
+    },
+    {
+      "epoch": 10.8370607028754,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6784
+    },
+    {
+      "epoch": 10.838658146964857,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 6785
+    },
+    {
+      "epoch": 10.840255591054314,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6786
+    },
+    {
+      "epoch": 10.84185303514377,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6787
+    },
+    {
+      "epoch": 10.843450479233226,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 6788
+    },
+    {
+      "epoch": 10.845047923322683,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 6789
+    },
+    {
+      "epoch": 10.84664536741214,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6790
+    },
+    {
+      "epoch": 10.848242811501597,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 6791
+    },
+    {
+      "epoch": 10.849840255591054,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 6792
+    },
+    {
+      "epoch": 10.851437699680512,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 6793
+    },
+    {
+      "epoch": 10.853035143769969,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 6794
+    },
+    {
+      "epoch": 10.854632587859426,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 6795
+    },
+    {
+      "epoch": 10.856230031948883,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 6796
+    },
+    {
+      "epoch": 10.857827476038338,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6797
+    },
+    {
+      "epoch": 10.859424920127795,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 6798
+    },
+    {
+      "epoch": 10.861022364217252,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 6799
+    },
+    {
+      "epoch": 10.86261980830671,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 6800
+    },
+    {
+      "epoch": 10.864217252396166,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 6801
+    },
+    {
+      "epoch": 10.865814696485623,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 6802
+    },
+    {
+      "epoch": 10.86741214057508,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 6803
+    },
+    {
+      "epoch": 10.869009584664537,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 6804
+    },
+    {
+      "epoch": 10.870607028753994,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 6805
+    },
+    {
+      "epoch": 10.87220447284345,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 6806
+    },
+    {
+      "epoch": 10.873801916932907,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 6807
+    },
+    {
+      "epoch": 10.875399361022364,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6808
+    },
+    {
+      "epoch": 10.87699680511182,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6809
+    },
+    {
+      "epoch": 10.878594249201278,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 6810
+    },
+    {
+      "epoch": 10.880191693290735,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6811
+    },
+    {
+      "epoch": 10.881789137380192,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 6812
+    },
+    {
+      "epoch": 10.883386581469649,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 6813
+    },
+    {
+      "epoch": 10.884984025559106,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1194,
+      "step": 6814
+    },
+    {
+      "epoch": 10.886581469648561,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6815
+    },
+    {
+      "epoch": 10.888178913738018,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 6816
+    },
+    {
+      "epoch": 10.889776357827476,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 6817
+    },
+    {
+      "epoch": 10.891373801916933,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 6818
+    },
+    {
+      "epoch": 10.89297124600639,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 6819
+    },
+    {
+      "epoch": 10.894568690095847,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6820
+    },
+    {
+      "epoch": 10.896166134185304,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 6821
+    },
+    {
+      "epoch": 10.89776357827476,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 6822
+    },
+    {
+      "epoch": 10.899361022364218,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 6823
+    },
+    {
+      "epoch": 10.900958466453675,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 6824
+    },
+    {
+      "epoch": 10.90255591054313,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6825
+    },
+    {
+      "epoch": 10.904153354632587,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1217,
+      "step": 6826
+    },
+    {
+      "epoch": 10.905750798722044,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 6827
+    },
+    {
+      "epoch": 10.907348242811501,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 6828
+    },
+    {
+      "epoch": 10.908945686900958,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 6829
+    },
+    {
+      "epoch": 10.910543130990416,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 6830
+    },
+    {
+      "epoch": 10.912140575079873,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 6831
+    },
+    {
+      "epoch": 10.91373801916933,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 6832
+    },
+    {
+      "epoch": 10.915335463258787,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.143,
+      "step": 6833
+    },
+    {
+      "epoch": 10.916932907348244,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 6834
+    },
+    {
+      "epoch": 10.918530351437699,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6835
+    },
+    {
+      "epoch": 10.920127795527156,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 6836
+    },
+    {
+      "epoch": 10.921725239616613,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 6837
+    },
+    {
+      "epoch": 10.92332268370607,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6838
+    },
+    {
+      "epoch": 10.924920127795527,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 6839
+    },
+    {
+      "epoch": 10.926517571884984,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 6840
+    },
+    {
+      "epoch": 10.928115015974441,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6841
+    },
+    {
+      "epoch": 10.929712460063898,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 6842
+    },
+    {
+      "epoch": 10.931309904153355,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 6843
+    },
+    {
+      "epoch": 10.93290734824281,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6844
+    },
+    {
+      "epoch": 10.934504792332268,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 6845
+    },
+    {
+      "epoch": 10.936102236421725,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6846
+    },
+    {
+      "epoch": 10.937699680511182,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6847
+    },
+    {
+      "epoch": 10.939297124600639,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 6848
+    },
+    {
+      "epoch": 10.940894568690096,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 6849
+    },
+    {
+      "epoch": 10.942492012779553,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6850
+    },
+    {
+      "epoch": 10.94408945686901,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 6851
+    },
+    {
+      "epoch": 10.945686900958467,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1464,
+      "step": 6852
+    },
+    {
+      "epoch": 10.947284345047922,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 6853
+    },
+    {
+      "epoch": 10.94888178913738,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 6854
+    },
+    {
+      "epoch": 10.950479233226837,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.149,
+      "step": 6855
+    },
+    {
+      "epoch": 10.952076677316294,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 6856
+    },
+    {
+      "epoch": 10.95367412140575,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 6857
+    },
+    {
+      "epoch": 10.955271565495208,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 6858
+    },
+    {
+      "epoch": 10.956869009584665,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 6859
+    },
+    {
+      "epoch": 10.958466453674122,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 6860
+    },
+    {
+      "epoch": 10.960063897763579,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1246,
+      "step": 6861
+    },
+    {
+      "epoch": 10.961661341853034,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 6862
+    },
+    {
+      "epoch": 10.963258785942491,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 6863
+    },
+    {
+      "epoch": 10.964856230031948,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 6864
+    },
+    {
+      "epoch": 10.966453674121405,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 6865
+    },
+    {
+      "epoch": 10.968051118210862,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 6866
+    },
+    {
+      "epoch": 10.96964856230032,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 6867
+    },
+    {
+      "epoch": 10.971246006389777,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 6868
+    },
+    {
+      "epoch": 10.972843450479234,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 6869
+    },
+    {
+      "epoch": 10.97444089456869,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 6870
+    },
+    {
+      "epoch": 10.976038338658148,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 6871
+    },
+    {
+      "epoch": 10.977635782747605,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1262,
+      "step": 6872
+    },
+    {
+      "epoch": 10.97923322683706,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6873
+    },
+    {
+      "epoch": 10.980830670926517,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 6874
+    },
+    {
+      "epoch": 10.982428115015974,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 6875
+    },
+    {
+      "epoch": 10.984025559105431,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 6876
+    },
+    {
+      "epoch": 10.985623003194888,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 6877
+    },
+    {
+      "epoch": 10.987220447284345,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 6878
+    },
+    {
+      "epoch": 10.988817891373802,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 6879
+    },
+    {
+      "epoch": 10.99041533546326,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 6880
+    },
+    {
+      "epoch": 10.992012779552716,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 6881
+    },
+    {
+      "epoch": 10.993610223642172,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 6882
+    },
+    {
+      "epoch": 10.995207667731629,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 6883
+    },
+    {
+      "epoch": 10.996805111821086,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 6884
+    },
+    {
+      "epoch": 10.998402555910543,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 6885
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6886
+    },
+    {
+      "epoch": 11.001597444089457,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 6887
+    },
+    {
+      "epoch": 11.003194888178914,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 6888
+    },
+    {
+      "epoch": 11.004792332268371,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 6889
+    },
+    {
+      "epoch": 11.006389776357828,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 6890
+    },
+    {
+      "epoch": 11.007987220447284,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6891
+    },
+    {
+      "epoch": 11.00958466453674,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 6892
+    },
+    {
+      "epoch": 11.011182108626198,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 6893
+    },
+    {
+      "epoch": 11.012779552715655,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6894
+    },
+    {
+      "epoch": 11.014376996805112,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 6895
+    },
+    {
+      "epoch": 11.015974440894569,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 6896
+    },
+    {
+      "epoch": 11.017571884984026,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 6897
+    },
+    {
+      "epoch": 11.019169329073483,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 6898
+    },
+    {
+      "epoch": 11.02076677316294,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 6899
+    },
+    {
+      "epoch": 11.022364217252397,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 6900
+    },
+    {
+      "epoch": 11.023961661341852,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 6901
+    },
+    {
+      "epoch": 11.02555910543131,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6902
+    },
+    {
+      "epoch": 11.027156549520766,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 6903
+    },
+    {
+      "epoch": 11.028753993610223,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 6904
+    },
+    {
+      "epoch": 11.03035143769968,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 6905
+    },
+    {
+      "epoch": 11.031948881789138,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 6906
+    },
+    {
+      "epoch": 11.033546325878595,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1451,
+      "step": 6907
+    },
+    {
+      "epoch": 11.035143769968052,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 6908
+    },
+    {
+      "epoch": 11.036741214057509,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.119,
+      "step": 6909
+    },
+    {
+      "epoch": 11.038338658146964,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 6910
+    },
+    {
+      "epoch": 11.039936102236421,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 6911
+    },
+    {
+      "epoch": 11.041533546325878,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 6912
+    },
+    {
+      "epoch": 11.043130990415335,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6913
+    },
+    {
+      "epoch": 11.044728434504792,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 6914
+    },
+    {
+      "epoch": 11.04632587859425,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6915
+    },
+    {
+      "epoch": 11.047923322683706,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6916
+    },
+    {
+      "epoch": 11.049520766773163,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 6917
+    },
+    {
+      "epoch": 11.05111821086262,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 6918
+    },
+    {
+      "epoch": 11.052715654952078,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 6919
+    },
+    {
+      "epoch": 11.054313099041533,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 6920
+    },
+    {
+      "epoch": 11.05591054313099,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 6921
+    },
+    {
+      "epoch": 11.057507987220447,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6922
+    },
+    {
+      "epoch": 11.059105431309904,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1431,
+      "step": 6923
+    },
+    {
+      "epoch": 11.060702875399361,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 6924
+    },
+    {
+      "epoch": 11.062300319488818,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 6925
+    },
+    {
+      "epoch": 11.063897763578275,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 6926
+    },
+    {
+      "epoch": 11.065495207667732,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6927
+    },
+    {
+      "epoch": 11.06709265175719,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 6928
+    },
+    {
+      "epoch": 11.068690095846645,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 6929
+    },
+    {
+      "epoch": 11.070287539936102,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 6930
+    },
+    {
+      "epoch": 11.071884984025559,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 6931
+    },
+    {
+      "epoch": 11.073482428115016,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 6932
+    },
+    {
+      "epoch": 11.075079872204473,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 6933
+    },
+    {
+      "epoch": 11.07667731629393,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 6934
+    },
+    {
+      "epoch": 11.078274760383387,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1214,
+      "step": 6935
+    },
+    {
+      "epoch": 11.079872204472844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 6936
+    },
+    {
+      "epoch": 11.081469648562301,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 6937
+    },
+    {
+      "epoch": 11.083067092651758,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 6938
+    },
+    {
+      "epoch": 11.084664536741213,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 6939
+    },
+    {
+      "epoch": 11.08626198083067,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 6940
+    },
+    {
+      "epoch": 11.087859424920127,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 6941
+    },
+    {
+      "epoch": 11.089456869009584,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 6942
+    },
+    {
+      "epoch": 11.091054313099042,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 6943
+    },
+    {
+      "epoch": 11.092651757188499,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 6944
+    },
+    {
+      "epoch": 11.094249201277956,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 6945
+    },
+    {
+      "epoch": 11.095846645367413,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 6946
+    },
+    {
+      "epoch": 11.09744408945687,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 6947
+    },
+    {
+      "epoch": 11.099041533546325,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 6948
+    },
+    {
+      "epoch": 11.100638977635782,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 6949
+    },
+    {
+      "epoch": 11.10223642172524,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 6950
+    },
+    {
+      "epoch": 11.103833865814696,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 6951
+    },
+    {
+      "epoch": 11.105431309904153,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6952
+    },
+    {
+      "epoch": 11.10702875399361,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 6953
+    },
+    {
+      "epoch": 11.108626198083067,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 6954
+    },
+    {
+      "epoch": 11.110223642172524,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 6955
+    },
+    {
+      "epoch": 11.111821086261982,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 6956
+    },
+    {
+      "epoch": 11.113418530351439,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.146,
+      "step": 6957
+    },
+    {
+      "epoch": 11.115015974440894,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 6958
+    },
+    {
+      "epoch": 11.116613418530351,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6959
+    },
+    {
+      "epoch": 11.118210862619808,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 6960
+    },
+    {
+      "epoch": 11.119808306709265,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 6961
+    },
+    {
+      "epoch": 11.121405750798722,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 6962
+    },
+    {
+      "epoch": 11.12300319488818,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 6963
+    },
+    {
+      "epoch": 11.124600638977636,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 6964
+    },
+    {
+      "epoch": 11.126198083067093,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 6965
+    },
+    {
+      "epoch": 11.12779552715655,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 6966
+    },
+    {
+      "epoch": 11.129392971246006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 6967
+    },
+    {
+      "epoch": 11.130990415335463,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 6968
+    },
+    {
+      "epoch": 11.13258785942492,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 6969
+    },
+    {
+      "epoch": 11.134185303514377,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 6970
+    },
+    {
+      "epoch": 11.135782747603834,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 6971
+    },
+    {
+      "epoch": 11.13738019169329,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 6972
+    },
+    {
+      "epoch": 11.138977635782748,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 6973
+    },
+    {
+      "epoch": 11.140575079872205,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 6974
+    },
+    {
+      "epoch": 11.142172523961662,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 6975
+    },
+    {
+      "epoch": 11.143769968051119,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1465,
+      "step": 6976
+    },
+    {
+      "epoch": 11.145367412140574,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 6977
+    },
+    {
+      "epoch": 11.146964856230031,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 6978
+    },
+    {
+      "epoch": 11.148562300319488,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 6979
+    },
+    {
+      "epoch": 11.150159744408946,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 6980
+    },
+    {
+      "epoch": 11.151757188498403,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 6981
+    },
+    {
+      "epoch": 11.15335463258786,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 6982
+    },
+    {
+      "epoch": 11.154952076677317,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 6983
+    },
+    {
+      "epoch": 11.156549520766774,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 6984
+    },
+    {
+      "epoch": 11.15814696485623,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 6985
+    },
+    {
+      "epoch": 11.159744408945686,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 6986
+    },
+    {
+      "epoch": 11.161341853035143,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 6987
+    },
+    {
+      "epoch": 11.1629392971246,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 6988
+    },
+    {
+      "epoch": 11.164536741214057,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 6989
+    },
+    {
+      "epoch": 11.166134185303514,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 6990
+    },
+    {
+      "epoch": 11.167731629392971,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 6991
+    },
+    {
+      "epoch": 11.169329073482428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 6992
+    },
+    {
+      "epoch": 11.170926517571885,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 6993
+    },
+    {
+      "epoch": 11.172523961661343,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 6994
+    },
+    {
+      "epoch": 11.1741214057508,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1236,
+      "step": 6995
+    },
+    {
+      "epoch": 11.175718849840255,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 6996
+    },
+    {
+      "epoch": 11.177316293929712,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 6997
+    },
+    {
+      "epoch": 11.178913738019169,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 6998
+    },
+    {
+      "epoch": 11.180511182108626,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 6999
+    },
+    {
+      "epoch": 11.182108626198083,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 7000
+    },
+    {
+      "epoch": 11.18370607028754,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 7001
+    },
+    {
+      "epoch": 11.185303514376997,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 7002
+    },
+    {
+      "epoch": 11.186900958466454,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7003
+    },
+    {
+      "epoch": 11.188498402555911,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 7004
+    },
+    {
+      "epoch": 11.190095846645367,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 7005
+    },
+    {
+      "epoch": 11.191693290734824,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 7006
+    },
+    {
+      "epoch": 11.19329073482428,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7007
+    },
+    {
+      "epoch": 11.194888178913738,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7008
+    },
+    {
+      "epoch": 11.196485623003195,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 7009
+    },
+    {
+      "epoch": 11.198083067092652,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 7010
+    },
+    {
+      "epoch": 11.199680511182109,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 7011
+    },
+    {
+      "epoch": 11.201277955271566,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 7012
+    },
+    {
+      "epoch": 11.202875399361023,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 7013
+    },
+    {
+      "epoch": 11.204472843450478,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7014
+    },
+    {
+      "epoch": 11.206070287539935,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 7015
+    },
+    {
+      "epoch": 11.207667731629392,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 7016
+    },
+    {
+      "epoch": 11.20926517571885,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7017
+    },
+    {
+      "epoch": 11.210862619808307,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 7018
+    },
+    {
+      "epoch": 11.212460063897764,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 7019
+    },
+    {
+      "epoch": 11.21405750798722,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7020
+    },
+    {
+      "epoch": 11.215654952076678,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7021
+    },
+    {
+      "epoch": 11.217252396166135,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 7022
+    },
+    {
+      "epoch": 11.218849840255592,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 7023
+    },
+    {
+      "epoch": 11.220447284345047,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 7024
+    },
+    {
+      "epoch": 11.222044728434504,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 7025
+    },
+    {
+      "epoch": 11.223642172523961,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7026
+    },
+    {
+      "epoch": 11.225239616613418,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7027
+    },
+    {
+      "epoch": 11.226837060702875,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7028
+    },
+    {
+      "epoch": 11.228434504792332,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 7029
+    },
+    {
+      "epoch": 11.23003194888179,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 7030
+    },
+    {
+      "epoch": 11.231629392971247,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 7031
+    },
+    {
+      "epoch": 11.233226837060704,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 7032
+    },
+    {
+      "epoch": 11.23482428115016,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 7033
+    },
+    {
+      "epoch": 11.236421725239616,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7034
+    },
+    {
+      "epoch": 11.238019169329073,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7035
+    },
+    {
+      "epoch": 11.23961661341853,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1445,
+      "step": 7036
+    },
+    {
+      "epoch": 11.241214057507987,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7037
+    },
+    {
+      "epoch": 11.242811501597444,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 7038
+    },
+    {
+      "epoch": 11.244408945686901,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1449,
+      "step": 7039
+    },
+    {
+      "epoch": 11.246006389776358,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7040
+    },
+    {
+      "epoch": 11.247603833865815,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 7041
+    },
+    {
+      "epoch": 11.249201277955272,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 7042
+    },
+    {
+      "epoch": 11.250798722044728,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7043
+    },
+    {
+      "epoch": 11.252396166134185,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 7044
+    },
+    {
+      "epoch": 11.253993610223642,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 7045
+    },
+    {
+      "epoch": 11.255591054313099,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 7046
+    },
+    {
+      "epoch": 11.257188498402556,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 7047
+    },
+    {
+      "epoch": 11.258785942492013,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 7048
+    },
+    {
+      "epoch": 11.26038338658147,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1217,
+      "step": 7049
+    },
+    {
+      "epoch": 11.261980830670927,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7050
+    },
+    {
+      "epoch": 11.263578274760384,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 7051
+    },
+    {
+      "epoch": 11.26517571884984,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 7052
+    },
+    {
+      "epoch": 11.266773162939296,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 7053
+    },
+    {
+      "epoch": 11.268370607028753,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7054
+    },
+    {
+      "epoch": 11.26996805111821,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 7055
+    },
+    {
+      "epoch": 11.271565495207668,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 7056
+    },
+    {
+      "epoch": 11.273162939297125,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7057
+    },
+    {
+      "epoch": 11.274760383386582,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 7058
+    },
+    {
+      "epoch": 11.276357827476039,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7059
+    },
+    {
+      "epoch": 11.277955271565496,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 7060
+    },
+    {
+      "epoch": 11.279552715654953,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7061
+    },
+    {
+      "epoch": 11.281150159744408,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7062
+    },
+    {
+      "epoch": 11.282747603833865,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 7063
+    },
+    {
+      "epoch": 11.284345047923322,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7064
+    },
+    {
+      "epoch": 11.28594249201278,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7065
+    },
+    {
+      "epoch": 11.287539936102236,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 7066
+    },
+    {
+      "epoch": 11.289137380191693,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7067
+    },
+    {
+      "epoch": 11.29073482428115,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7068
+    },
+    {
+      "epoch": 11.292332268370608,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1239,
+      "step": 7069
+    },
+    {
+      "epoch": 11.293929712460065,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 7070
+    },
+    {
+      "epoch": 11.295527156549522,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 7071
+    },
+    {
+      "epoch": 11.297124600638977,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1201,
+      "step": 7072
+    },
+    {
+      "epoch": 11.298722044728434,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 7073
+    },
+    {
+      "epoch": 11.300319488817891,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 7074
+    },
+    {
+      "epoch": 11.301916932907348,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 7075
+    },
+    {
+      "epoch": 11.303514376996805,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 7076
+    },
+    {
+      "epoch": 11.305111821086262,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7077
+    },
+    {
+      "epoch": 11.30670926517572,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 7078
+    },
+    {
+      "epoch": 11.308306709265176,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 7079
+    },
+    {
+      "epoch": 11.309904153354633,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 7080
+    },
+    {
+      "epoch": 11.311501597444089,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7081
+    },
+    {
+      "epoch": 11.313099041533546,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 7082
+    },
+    {
+      "epoch": 11.314696485623003,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7083
+    },
+    {
+      "epoch": 11.31629392971246,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 7084
+    },
+    {
+      "epoch": 11.317891373801917,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7085
+    },
+    {
+      "epoch": 11.319488817891374,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 7086
+    },
+    {
+      "epoch": 11.321086261980831,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7087
+    },
+    {
+      "epoch": 11.322683706070288,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7088
+    },
+    {
+      "epoch": 11.324281150159745,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7089
+    },
+    {
+      "epoch": 11.3258785942492,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1191,
+      "step": 7090
+    },
+    {
+      "epoch": 11.327476038338657,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7091
+    },
+    {
+      "epoch": 11.329073482428115,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7092
+    },
+    {
+      "epoch": 11.330670926517572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 7093
+    },
+    {
+      "epoch": 11.332268370607029,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 7094
+    },
+    {
+      "epoch": 11.333865814696486,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 7095
+    },
+    {
+      "epoch": 11.335463258785943,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7096
+    },
+    {
+      "epoch": 11.3370607028754,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 7097
+    },
+    {
+      "epoch": 11.338658146964857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 7098
+    },
+    {
+      "epoch": 11.340255591054314,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7099
+    },
+    {
+      "epoch": 11.34185303514377,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7100
+    },
+    {
+      "epoch": 11.343450479233226,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 7101
+    },
+    {
+      "epoch": 11.345047923322683,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 7102
+    },
+    {
+      "epoch": 11.34664536741214,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7103
+    },
+    {
+      "epoch": 11.348242811501597,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7104
+    },
+    {
+      "epoch": 11.349840255591054,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7105
+    },
+    {
+      "epoch": 11.351437699680512,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 7106
+    },
+    {
+      "epoch": 11.353035143769969,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 7107
+    },
+    {
+      "epoch": 11.354632587859426,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7108
+    },
+    {
+      "epoch": 11.356230031948883,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7109
+    },
+    {
+      "epoch": 11.357827476038338,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 7110
+    },
+    {
+      "epoch": 11.359424920127795,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 7111
+    },
+    {
+      "epoch": 11.361022364217252,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7112
+    },
+    {
+      "epoch": 11.36261980830671,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7113
+    },
+    {
+      "epoch": 11.364217252396166,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 7114
+    },
+    {
+      "epoch": 11.365814696485623,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 7115
+    },
+    {
+      "epoch": 11.36741214057508,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7116
+    },
+    {
+      "epoch": 11.369009584664537,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7117
+    },
+    {
+      "epoch": 11.370607028753994,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7118
+    },
+    {
+      "epoch": 11.37220447284345,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1246,
+      "step": 7119
+    },
+    {
+      "epoch": 11.373801916932907,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 7120
+    },
+    {
+      "epoch": 11.375399361022364,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 7121
+    },
+    {
+      "epoch": 11.37699680511182,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 7122
+    },
+    {
+      "epoch": 11.378594249201278,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 7123
+    },
+    {
+      "epoch": 11.380191693290735,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 7124
+    },
+    {
+      "epoch": 11.381789137380192,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 7125
+    },
+    {
+      "epoch": 11.383386581469649,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1196,
+      "step": 7126
+    },
+    {
+      "epoch": 11.384984025559106,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7127
+    },
+    {
+      "epoch": 11.386581469648561,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7128
+    },
+    {
+      "epoch": 11.388178913738018,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1236,
+      "step": 7129
+    },
+    {
+      "epoch": 11.389776357827476,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7130
+    },
+    {
+      "epoch": 11.391373801916933,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 7131
+    },
+    {
+      "epoch": 11.39297124600639,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7132
+    },
+    {
+      "epoch": 11.394568690095847,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 7133
+    },
+    {
+      "epoch": 11.396166134185304,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 7134
+    },
+    {
+      "epoch": 11.39776357827476,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7135
+    },
+    {
+      "epoch": 11.399361022364218,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 7136
+    },
+    {
+      "epoch": 11.400958466453675,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7137
+    },
+    {
+      "epoch": 11.40255591054313,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7138
+    },
+    {
+      "epoch": 11.404153354632587,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 7139
+    },
+    {
+      "epoch": 11.405750798722044,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1446,
+      "step": 7140
+    },
+    {
+      "epoch": 11.407348242811501,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7141
+    },
+    {
+      "epoch": 11.408945686900958,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7142
+    },
+    {
+      "epoch": 11.410543130990416,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 7143
+    },
+    {
+      "epoch": 11.412140575079873,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 7144
+    },
+    {
+      "epoch": 11.41373801916933,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 7145
+    },
+    {
+      "epoch": 11.415335463258787,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 7146
+    },
+    {
+      "epoch": 11.416932907348242,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7147
+    },
+    {
+      "epoch": 11.418530351437699,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7148
+    },
+    {
+      "epoch": 11.420127795527156,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 7149
+    },
+    {
+      "epoch": 11.421725239616613,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7150
+    },
+    {
+      "epoch": 11.42332268370607,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7151
+    },
+    {
+      "epoch": 11.424920127795527,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7152
+    },
+    {
+      "epoch": 11.426517571884984,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7153
+    },
+    {
+      "epoch": 11.428115015974441,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 7154
+    },
+    {
+      "epoch": 11.429712460063898,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7155
+    },
+    {
+      "epoch": 11.431309904153355,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 7156
+    },
+    {
+      "epoch": 11.43290734824281,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 7157
+    },
+    {
+      "epoch": 11.434504792332268,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7158
+    },
+    {
+      "epoch": 11.436102236421725,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 7159
+    },
+    {
+      "epoch": 11.437699680511182,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 7160
+    },
+    {
+      "epoch": 11.439297124600639,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 7161
+    },
+    {
+      "epoch": 11.440894568690096,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7162
+    },
+    {
+      "epoch": 11.442492012779553,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 7163
+    },
+    {
+      "epoch": 11.44408945686901,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7164
+    },
+    {
+      "epoch": 11.445686900958467,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7165
+    },
+    {
+      "epoch": 11.447284345047922,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 7166
+    },
+    {
+      "epoch": 11.44888178913738,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7167
+    },
+    {
+      "epoch": 11.450479233226837,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7168
+    },
+    {
+      "epoch": 11.452076677316294,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7169
+    },
+    {
+      "epoch": 11.45367412140575,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7170
+    },
+    {
+      "epoch": 11.455271565495208,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 7171
+    },
+    {
+      "epoch": 11.456869009584665,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 7172
+    },
+    {
+      "epoch": 11.458466453674122,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1498,
+      "step": 7173
+    },
+    {
+      "epoch": 11.460063897763579,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1495,
+      "step": 7174
+    },
+    {
+      "epoch": 11.461661341853036,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7175
+    },
+    {
+      "epoch": 11.463258785942491,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 7176
+    },
+    {
+      "epoch": 11.464856230031948,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7177
+    },
+    {
+      "epoch": 11.466453674121405,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 7178
+    },
+    {
+      "epoch": 11.468051118210862,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 7179
+    },
+    {
+      "epoch": 11.46964856230032,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7180
+    },
+    {
+      "epoch": 11.471246006389777,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7181
+    },
+    {
+      "epoch": 11.472843450479234,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 7182
+    },
+    {
+      "epoch": 11.47444089456869,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7183
+    },
+    {
+      "epoch": 11.476038338658148,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7184
+    },
+    {
+      "epoch": 11.477635782747603,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 7185
+    },
+    {
+      "epoch": 11.47923322683706,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 7186
+    },
+    {
+      "epoch": 11.480830670926517,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 7187
+    },
+    {
+      "epoch": 11.482428115015974,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7188
+    },
+    {
+      "epoch": 11.484025559105431,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7189
+    },
+    {
+      "epoch": 11.485623003194888,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7190
+    },
+    {
+      "epoch": 11.487220447284345,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1487,
+      "step": 7191
+    },
+    {
+      "epoch": 11.488817891373802,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7192
+    },
+    {
+      "epoch": 11.49041533546326,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 7193
+    },
+    {
+      "epoch": 11.492012779552716,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7194
+    },
+    {
+      "epoch": 11.493610223642172,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7195
+    },
+    {
+      "epoch": 11.495207667731629,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 7196
+    },
+    {
+      "epoch": 11.496805111821086,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7197
+    },
+    {
+      "epoch": 11.498402555910543,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7198
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7199
+    },
+    {
+      "epoch": 11.501597444089457,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7200
+    },
+    {
+      "epoch": 11.503194888178914,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7201
+    },
+    {
+      "epoch": 11.504792332268371,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 7202
+    },
+    {
+      "epoch": 11.506389776357828,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7203
+    },
+    {
+      "epoch": 11.507987220447284,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 7204
+    },
+    {
+      "epoch": 11.50958466453674,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7205
+    },
+    {
+      "epoch": 11.511182108626198,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 7206
+    },
+    {
+      "epoch": 11.512779552715655,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 7207
+    },
+    {
+      "epoch": 11.514376996805112,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7208
+    },
+    {
+      "epoch": 11.515974440894569,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 7209
+    },
+    {
+      "epoch": 11.517571884984026,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 7210
+    },
+    {
+      "epoch": 11.519169329073483,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 7211
+    },
+    {
+      "epoch": 11.52076677316294,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7212
+    },
+    {
+      "epoch": 11.522364217252395,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1237,
+      "step": 7213
+    },
+    {
+      "epoch": 11.523961661341852,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7214
+    },
+    {
+      "epoch": 11.52555910543131,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1458,
+      "step": 7215
+    },
+    {
+      "epoch": 11.527156549520766,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7216
+    },
+    {
+      "epoch": 11.528753993610223,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 7217
+    },
+    {
+      "epoch": 11.53035143769968,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7218
+    },
+    {
+      "epoch": 11.531948881789138,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 7219
+    },
+    {
+      "epoch": 11.533546325878595,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 7220
+    },
+    {
+      "epoch": 11.535143769968052,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 7221
+    },
+    {
+      "epoch": 11.536741214057509,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7222
+    },
+    {
+      "epoch": 11.538338658146966,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7223
+    },
+    {
+      "epoch": 11.539936102236421,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7224
+    },
+    {
+      "epoch": 11.541533546325878,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7225
+    },
+    {
+      "epoch": 11.543130990415335,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 7226
+    },
+    {
+      "epoch": 11.544728434504792,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 7227
+    },
+    {
+      "epoch": 11.54632587859425,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 7228
+    },
+    {
+      "epoch": 11.547923322683706,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 7229
+    },
+    {
+      "epoch": 11.549520766773163,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 7230
+    },
+    {
+      "epoch": 11.55111821086262,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 7231
+    },
+    {
+      "epoch": 11.552715654952078,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 7232
+    },
+    {
+      "epoch": 11.554313099041533,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 7233
+    },
+    {
+      "epoch": 11.55591054313099,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 7234
+    },
+    {
+      "epoch": 11.557507987220447,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 7235
+    },
+    {
+      "epoch": 11.559105431309904,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7236
+    },
+    {
+      "epoch": 11.560702875399361,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7237
+    },
+    {
+      "epoch": 11.562300319488818,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7238
+    },
+    {
+      "epoch": 11.563897763578275,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7239
+    },
+    {
+      "epoch": 11.565495207667732,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 7240
+    },
+    {
+      "epoch": 11.56709265175719,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 7241
+    },
+    {
+      "epoch": 11.568690095846645,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 7242
+    },
+    {
+      "epoch": 11.570287539936102,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 7243
+    },
+    {
+      "epoch": 11.571884984025559,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7244
+    },
+    {
+      "epoch": 11.573482428115016,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 7245
+    },
+    {
+      "epoch": 11.575079872204473,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7246
+    },
+    {
+      "epoch": 11.57667731629393,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 7247
+    },
+    {
+      "epoch": 11.578274760383387,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7248
+    },
+    {
+      "epoch": 11.579872204472844,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 7249
+    },
+    {
+      "epoch": 11.581469648562301,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7250
+    },
+    {
+      "epoch": 11.583067092651756,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7251
+    },
+    {
+      "epoch": 11.584664536741213,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7252
+    },
+    {
+      "epoch": 11.58626198083067,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 7253
+    },
+    {
+      "epoch": 11.587859424920127,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 7254
+    },
+    {
+      "epoch": 11.589456869009584,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 7255
+    },
+    {
+      "epoch": 11.591054313099042,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7256
+    },
+    {
+      "epoch": 11.592651757188499,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 7257
+    },
+    {
+      "epoch": 11.594249201277956,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 7258
+    },
+    {
+      "epoch": 11.595846645367413,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 7259
+    },
+    {
+      "epoch": 11.59744408945687,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7260
+    },
+    {
+      "epoch": 11.599041533546325,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7261
+    },
+    {
+      "epoch": 11.600638977635782,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7262
+    },
+    {
+      "epoch": 11.60223642172524,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 7263
+    },
+    {
+      "epoch": 11.603833865814696,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7264
+    },
+    {
+      "epoch": 11.605431309904153,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 7265
+    },
+    {
+      "epoch": 11.60702875399361,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7266
+    },
+    {
+      "epoch": 11.608626198083067,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1191,
+      "step": 7267
+    },
+    {
+      "epoch": 11.610223642172524,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7268
+    },
+    {
+      "epoch": 11.611821086261982,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7269
+    },
+    {
+      "epoch": 11.613418530351439,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 7270
+    },
+    {
+      "epoch": 11.615015974440894,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 7271
+    },
+    {
+      "epoch": 11.616613418530351,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 7272
+    },
+    {
+      "epoch": 11.618210862619808,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7273
+    },
+    {
+      "epoch": 11.619808306709265,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 7274
+    },
+    {
+      "epoch": 11.621405750798722,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1196,
+      "step": 7275
+    },
+    {
+      "epoch": 11.62300319488818,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 7276
+    },
+    {
+      "epoch": 11.624600638977636,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 7277
+    },
+    {
+      "epoch": 11.626198083067093,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 7278
+    },
+    {
+      "epoch": 11.62779552715655,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 7279
+    },
+    {
+      "epoch": 11.629392971246006,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 7280
+    },
+    {
+      "epoch": 11.630990415335463,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 7281
+    },
+    {
+      "epoch": 11.63258785942492,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 7282
+    },
+    {
+      "epoch": 11.634185303514377,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 7283
+    },
+    {
+      "epoch": 11.635782747603834,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7284
+    },
+    {
+      "epoch": 11.63738019169329,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 7285
+    },
+    {
+      "epoch": 11.638977635782748,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7286
+    },
+    {
+      "epoch": 11.640575079872205,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 7287
+    },
+    {
+      "epoch": 11.642172523961662,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 7288
+    },
+    {
+      "epoch": 11.643769968051117,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 7289
+    },
+    {
+      "epoch": 11.645367412140574,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 7290
+    },
+    {
+      "epoch": 11.646964856230031,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7291
+    },
+    {
+      "epoch": 11.648562300319488,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7292
+    },
+    {
+      "epoch": 11.650159744408946,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 7293
+    },
+    {
+      "epoch": 11.651757188498403,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7294
+    },
+    {
+      "epoch": 11.65335463258786,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7295
+    },
+    {
+      "epoch": 11.654952076677317,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 7296
+    },
+    {
+      "epoch": 11.656549520766774,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7297
+    },
+    {
+      "epoch": 11.65814696485623,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 7298
+    },
+    {
+      "epoch": 11.659744408945686,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7299
+    },
+    {
+      "epoch": 11.661341853035143,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 7300
+    },
+    {
+      "epoch": 11.6629392971246,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 7301
+    },
+    {
+      "epoch": 11.664536741214057,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7302
+    },
+    {
+      "epoch": 11.666134185303514,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7303
+    },
+    {
+      "epoch": 11.667731629392971,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 7304
+    },
+    {
+      "epoch": 11.669329073482428,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 7305
+    },
+    {
+      "epoch": 11.670926517571885,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 7306
+    },
+    {
+      "epoch": 11.672523961661343,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 7307
+    },
+    {
+      "epoch": 11.6741214057508,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 7308
+    },
+    {
+      "epoch": 11.675718849840255,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 7309
+    },
+    {
+      "epoch": 11.677316293929712,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 7310
+    },
+    {
+      "epoch": 11.678913738019169,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 7311
+    },
+    {
+      "epoch": 11.680511182108626,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1437,
+      "step": 7312
+    },
+    {
+      "epoch": 11.682108626198083,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 7313
+    },
+    {
+      "epoch": 11.68370607028754,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7314
+    },
+    {
+      "epoch": 11.685303514376997,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 7315
+    },
+    {
+      "epoch": 11.686900958466454,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 7316
+    },
+    {
+      "epoch": 11.688498402555911,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7317
+    },
+    {
+      "epoch": 11.690095846645367,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 7318
+    },
+    {
+      "epoch": 11.691693290734824,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7319
+    },
+    {
+      "epoch": 11.69329073482428,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7320
+    },
+    {
+      "epoch": 11.694888178913738,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7321
+    },
+    {
+      "epoch": 11.696485623003195,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 7322
+    },
+    {
+      "epoch": 11.698083067092652,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 7323
+    },
+    {
+      "epoch": 11.699680511182109,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7324
+    },
+    {
+      "epoch": 11.701277955271566,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7325
+    },
+    {
+      "epoch": 11.702875399361023,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 7326
+    },
+    {
+      "epoch": 11.704472843450478,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7327
+    },
+    {
+      "epoch": 11.706070287539935,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7328
+    },
+    {
+      "epoch": 11.707667731629392,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7329
+    },
+    {
+      "epoch": 11.70926517571885,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 7330
+    },
+    {
+      "epoch": 11.710862619808307,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 7331
+    },
+    {
+      "epoch": 11.712460063897764,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 7332
+    },
+    {
+      "epoch": 11.71405750798722,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 7333
+    },
+    {
+      "epoch": 11.715654952076678,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7334
+    },
+    {
+      "epoch": 11.717252396166135,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7335
+    },
+    {
+      "epoch": 11.718849840255592,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 7336
+    },
+    {
+      "epoch": 11.720447284345047,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7337
+    },
+    {
+      "epoch": 11.722044728434504,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 7338
+    },
+    {
+      "epoch": 11.723642172523961,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7339
+    },
+    {
+      "epoch": 11.725239616613418,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 7340
+    },
+    {
+      "epoch": 11.726837060702875,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 7341
+    },
+    {
+      "epoch": 11.728434504792332,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 7342
+    },
+    {
+      "epoch": 11.73003194888179,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7343
+    },
+    {
+      "epoch": 11.731629392971247,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 7344
+    },
+    {
+      "epoch": 11.733226837060704,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 7345
+    },
+    {
+      "epoch": 11.73482428115016,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7346
+    },
+    {
+      "epoch": 11.736421725239616,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7347
+    },
+    {
+      "epoch": 11.738019169329073,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7348
+    },
+    {
+      "epoch": 11.73961661341853,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 7349
+    },
+    {
+      "epoch": 11.741214057507987,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7350
+    },
+    {
+      "epoch": 11.742811501597444,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 7351
+    },
+    {
+      "epoch": 11.744408945686901,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 7352
+    },
+    {
+      "epoch": 11.746006389776358,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7353
+    },
+    {
+      "epoch": 11.747603833865815,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 7354
+    },
+    {
+      "epoch": 11.749201277955272,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7355
+    },
+    {
+      "epoch": 11.750798722044728,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7356
+    },
+    {
+      "epoch": 11.752396166134185,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 7357
+    },
+    {
+      "epoch": 11.753993610223642,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 7358
+    },
+    {
+      "epoch": 11.755591054313099,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 7359
+    },
+    {
+      "epoch": 11.757188498402556,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 7360
+    },
+    {
+      "epoch": 11.758785942492013,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7361
+    },
+    {
+      "epoch": 11.76038338658147,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7362
+    },
+    {
+      "epoch": 11.761980830670927,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 7363
+    },
+    {
+      "epoch": 11.763578274760384,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7364
+    },
+    {
+      "epoch": 11.76517571884984,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7365
+    },
+    {
+      "epoch": 11.766773162939296,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 7366
+    },
+    {
+      "epoch": 11.768370607028753,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 7367
+    },
+    {
+      "epoch": 11.76996805111821,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 7368
+    },
+    {
+      "epoch": 11.771565495207668,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7369
+    },
+    {
+      "epoch": 11.773162939297125,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7370
+    },
+    {
+      "epoch": 11.774760383386582,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7371
+    },
+    {
+      "epoch": 11.776357827476039,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 7372
+    },
+    {
+      "epoch": 11.777955271565496,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7373
+    },
+    {
+      "epoch": 11.779552715654953,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 7374
+    },
+    {
+      "epoch": 11.781150159744408,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 7375
+    },
+    {
+      "epoch": 11.782747603833865,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7376
+    },
+    {
+      "epoch": 11.784345047923322,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 7377
+    },
+    {
+      "epoch": 11.78594249201278,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 7378
+    },
+    {
+      "epoch": 11.787539936102236,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 7379
+    },
+    {
+      "epoch": 11.789137380191693,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7380
+    },
+    {
+      "epoch": 11.79073482428115,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 7381
+    },
+    {
+      "epoch": 11.792332268370608,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7382
+    },
+    {
+      "epoch": 11.793929712460065,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7383
+    },
+    {
+      "epoch": 11.795527156549522,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7384
+    },
+    {
+      "epoch": 11.797124600638977,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 7385
+    },
+    {
+      "epoch": 11.798722044728434,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7386
+    },
+    {
+      "epoch": 11.800319488817891,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.142,
+      "step": 7387
+    },
+    {
+      "epoch": 11.801916932907348,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 7388
+    },
+    {
+      "epoch": 11.803514376996805,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7389
+    },
+    {
+      "epoch": 11.805111821086262,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7390
+    },
+    {
+      "epoch": 11.80670926517572,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7391
+    },
+    {
+      "epoch": 11.808306709265176,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7392
+    },
+    {
+      "epoch": 11.809904153354633,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7393
+    },
+    {
+      "epoch": 11.811501597444089,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7394
+    },
+    {
+      "epoch": 11.813099041533546,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7395
+    },
+    {
+      "epoch": 11.814696485623003,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7396
+    },
+    {
+      "epoch": 11.81629392971246,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 7397
+    },
+    {
+      "epoch": 11.817891373801917,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7398
+    },
+    {
+      "epoch": 11.819488817891374,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 7399
+    },
+    {
+      "epoch": 11.821086261980831,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7400
+    },
+    {
+      "epoch": 11.822683706070288,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7401
+    },
+    {
+      "epoch": 11.824281150159745,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 7402
+    },
+    {
+      "epoch": 11.8258785942492,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7403
+    },
+    {
+      "epoch": 11.827476038338657,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 7404
+    },
+    {
+      "epoch": 11.829073482428115,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 7405
+    },
+    {
+      "epoch": 11.830670926517572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 7406
+    },
+    {
+      "epoch": 11.832268370607029,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7407
+    },
+    {
+      "epoch": 11.833865814696486,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 7408
+    },
+    {
+      "epoch": 11.835463258785943,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1198,
+      "step": 7409
+    },
+    {
+      "epoch": 11.8370607028754,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 7410
+    },
+    {
+      "epoch": 11.838658146964857,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 7411
+    },
+    {
+      "epoch": 11.840255591054314,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7412
+    },
+    {
+      "epoch": 11.84185303514377,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7413
+    },
+    {
+      "epoch": 11.843450479233226,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.148,
+      "step": 7414
+    },
+    {
+      "epoch": 11.845047923322683,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 7415
+    },
+    {
+      "epoch": 11.84664536741214,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7416
+    },
+    {
+      "epoch": 11.848242811501597,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7417
+    },
+    {
+      "epoch": 11.849840255591054,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 7418
+    },
+    {
+      "epoch": 11.851437699680512,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 7419
+    },
+    {
+      "epoch": 11.853035143769969,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 7420
+    },
+    {
+      "epoch": 11.854632587859426,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7421
+    },
+    {
+      "epoch": 11.856230031948883,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7422
+    },
+    {
+      "epoch": 11.857827476038338,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7423
+    },
+    {
+      "epoch": 11.859424920127795,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 7424
+    },
+    {
+      "epoch": 11.861022364217252,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7425
+    },
+    {
+      "epoch": 11.86261980830671,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7426
+    },
+    {
+      "epoch": 11.864217252396166,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7427
+    },
+    {
+      "epoch": 11.865814696485623,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7428
+    },
+    {
+      "epoch": 11.86741214057508,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 7429
+    },
+    {
+      "epoch": 11.869009584664537,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 7430
+    },
+    {
+      "epoch": 11.870607028753994,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 7431
+    },
+    {
+      "epoch": 11.87220447284345,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 7432
+    },
+    {
+      "epoch": 11.873801916932907,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 7433
+    },
+    {
+      "epoch": 11.875399361022364,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1204,
+      "step": 7434
+    },
+    {
+      "epoch": 11.87699680511182,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7435
+    },
+    {
+      "epoch": 11.878594249201278,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7436
+    },
+    {
+      "epoch": 11.880191693290735,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 7437
+    },
+    {
+      "epoch": 11.881789137380192,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 7438
+    },
+    {
+      "epoch": 11.883386581469649,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 7439
+    },
+    {
+      "epoch": 11.884984025559106,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 7440
+    },
+    {
+      "epoch": 11.886581469648561,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 7441
+    },
+    {
+      "epoch": 11.888178913738018,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7442
+    },
+    {
+      "epoch": 11.889776357827476,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 7443
+    },
+    {
+      "epoch": 11.891373801916933,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7444
+    },
+    {
+      "epoch": 11.89297124600639,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 7445
+    },
+    {
+      "epoch": 11.894568690095847,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 7446
+    },
+    {
+      "epoch": 11.896166134185304,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7447
+    },
+    {
+      "epoch": 11.89776357827476,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 7448
+    },
+    {
+      "epoch": 11.899361022364218,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 7449
+    },
+    {
+      "epoch": 11.900958466453675,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 7450
+    },
+    {
+      "epoch": 11.90255591054313,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1225,
+      "step": 7451
+    },
+    {
+      "epoch": 11.904153354632587,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7452
+    },
+    {
+      "epoch": 11.905750798722044,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7453
+    },
+    {
+      "epoch": 11.907348242811501,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 7454
+    },
+    {
+      "epoch": 11.908945686900958,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7455
+    },
+    {
+      "epoch": 11.910543130990416,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7456
+    },
+    {
+      "epoch": 11.912140575079873,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7457
+    },
+    {
+      "epoch": 11.91373801916933,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 7458
+    },
+    {
+      "epoch": 11.915335463258787,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 7459
+    },
+    {
+      "epoch": 11.916932907348244,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7460
+    },
+    {
+      "epoch": 11.918530351437699,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7461
+    },
+    {
+      "epoch": 11.920127795527156,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 7462
+    },
+    {
+      "epoch": 11.921725239616613,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 7463
+    },
+    {
+      "epoch": 11.92332268370607,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1204,
+      "step": 7464
+    },
+    {
+      "epoch": 11.924920127795527,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 7465
+    },
+    {
+      "epoch": 11.926517571884984,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7466
+    },
+    {
+      "epoch": 11.928115015974441,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 7467
+    },
+    {
+      "epoch": 11.929712460063898,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 7468
+    },
+    {
+      "epoch": 11.931309904153355,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 7469
+    },
+    {
+      "epoch": 11.93290734824281,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7470
+    },
+    {
+      "epoch": 11.934504792332268,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7471
+    },
+    {
+      "epoch": 11.936102236421725,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7472
+    },
+    {
+      "epoch": 11.937699680511182,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7473
+    },
+    {
+      "epoch": 11.939297124600639,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 7474
+    },
+    {
+      "epoch": 11.940894568690096,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7475
+    },
+    {
+      "epoch": 11.942492012779553,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 7476
+    },
+    {
+      "epoch": 11.94408945686901,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7477
+    },
+    {
+      "epoch": 11.945686900958467,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7478
+    },
+    {
+      "epoch": 11.947284345047922,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7479
+    },
+    {
+      "epoch": 11.94888178913738,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 7480
+    },
+    {
+      "epoch": 11.950479233226837,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 7481
+    },
+    {
+      "epoch": 11.952076677316294,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7482
+    },
+    {
+      "epoch": 11.95367412140575,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7483
+    },
+    {
+      "epoch": 11.955271565495208,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 7484
+    },
+    {
+      "epoch": 11.956869009584665,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 7485
+    },
+    {
+      "epoch": 11.958466453674122,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 7486
+    },
+    {
+      "epoch": 11.960063897763579,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7487
+    },
+    {
+      "epoch": 11.961661341853034,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 7488
+    },
+    {
+      "epoch": 11.963258785942491,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7489
+    },
+    {
+      "epoch": 11.964856230031948,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7490
+    },
+    {
+      "epoch": 11.966453674121405,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7491
+    },
+    {
+      "epoch": 11.968051118210862,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7492
+    },
+    {
+      "epoch": 11.96964856230032,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 7493
+    },
+    {
+      "epoch": 11.971246006389777,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7494
+    },
+    {
+      "epoch": 11.972843450479234,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 7495
+    },
+    {
+      "epoch": 11.97444089456869,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7496
+    },
+    {
+      "epoch": 11.976038338658148,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 7497
+    },
+    {
+      "epoch": 11.977635782747605,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 7498
+    },
+    {
+      "epoch": 11.97923322683706,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 7499
+    },
+    {
+      "epoch": 11.980830670926517,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 7500
+    },
+    {
+      "epoch": 11.982428115015974,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7501
+    },
+    {
+      "epoch": 11.984025559105431,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 7502
+    },
+    {
+      "epoch": 11.985623003194888,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7503
+    },
+    {
+      "epoch": 11.987220447284345,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 7504
+    },
+    {
+      "epoch": 11.988817891373802,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 7505
+    },
+    {
+      "epoch": 11.99041533546326,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7506
+    },
+    {
+      "epoch": 11.992012779552716,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7507
+    },
+    {
+      "epoch": 11.993610223642172,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 7508
+    },
+    {
+      "epoch": 11.995207667731629,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 7509
+    },
+    {
+      "epoch": 11.996805111821086,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7510
+    },
+    {
+      "epoch": 11.998402555910543,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1196,
+      "step": 7511
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 7512
+    },
+    {
+      "epoch": 12.001597444089457,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 7513
+    },
+    {
+      "epoch": 12.003194888178914,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 7514
+    },
+    {
+      "epoch": 12.004792332268371,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7515
+    },
+    {
+      "epoch": 12.006389776357828,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7516
+    },
+    {
+      "epoch": 12.007987220447284,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7517
+    },
+    {
+      "epoch": 12.00958466453674,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 7518
+    },
+    {
+      "epoch": 12.011182108626198,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 7519
+    },
+    {
+      "epoch": 12.012779552715655,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1161,
+      "step": 7520
+    },
+    {
+      "epoch": 12.014376996805112,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 7521
+    },
+    {
+      "epoch": 12.015974440894569,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7522
+    },
+    {
+      "epoch": 12.017571884984026,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 7523
+    },
+    {
+      "epoch": 12.019169329073483,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7524
+    },
+    {
+      "epoch": 12.02076677316294,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7525
+    },
+    {
+      "epoch": 12.022364217252397,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7526
+    },
+    {
+      "epoch": 12.023961661341852,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7527
+    },
+    {
+      "epoch": 12.02555910543131,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7528
+    },
+    {
+      "epoch": 12.027156549520766,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 7529
+    },
+    {
+      "epoch": 12.028753993610223,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 7530
+    },
+    {
+      "epoch": 12.03035143769968,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 7531
+    },
+    {
+      "epoch": 12.031948881789138,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7532
+    },
+    {
+      "epoch": 12.033546325878595,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 7533
+    },
+    {
+      "epoch": 12.035143769968052,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 7534
+    },
+    {
+      "epoch": 12.036741214057509,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7535
+    },
+    {
+      "epoch": 12.038338658146964,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7536
+    },
+    {
+      "epoch": 12.039936102236421,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 7537
+    },
+    {
+      "epoch": 12.041533546325878,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 7538
+    },
+    {
+      "epoch": 12.043130990415335,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 7539
+    },
+    {
+      "epoch": 12.044728434504792,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7540
+    },
+    {
+      "epoch": 12.04632587859425,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 7541
+    },
+    {
+      "epoch": 12.047923322683706,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 7542
+    },
+    {
+      "epoch": 12.049520766773163,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 7543
+    },
+    {
+      "epoch": 12.05111821086262,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 7544
+    },
+    {
+      "epoch": 12.052715654952078,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7545
+    },
+    {
+      "epoch": 12.054313099041533,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 7546
+    },
+    {
+      "epoch": 12.05591054313099,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7547
+    },
+    {
+      "epoch": 12.057507987220447,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 7548
+    },
+    {
+      "epoch": 12.059105431309904,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 7549
+    },
+    {
+      "epoch": 12.060702875399361,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7550
+    },
+    {
+      "epoch": 12.062300319488818,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 7551
+    },
+    {
+      "epoch": 12.063897763578275,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 7552
+    },
+    {
+      "epoch": 12.065495207667732,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 7553
+    },
+    {
+      "epoch": 12.06709265175719,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7554
+    },
+    {
+      "epoch": 12.068690095846645,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 7555
+    },
+    {
+      "epoch": 12.070287539936102,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7556
+    },
+    {
+      "epoch": 12.071884984025559,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 7557
+    },
+    {
+      "epoch": 12.073482428115016,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7558
+    },
+    {
+      "epoch": 12.075079872204473,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 7559
+    },
+    {
+      "epoch": 12.07667731629393,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7560
+    },
+    {
+      "epoch": 12.078274760383387,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 7561
+    },
+    {
+      "epoch": 12.079872204472844,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 7562
+    },
+    {
+      "epoch": 12.081469648562301,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 7563
+    },
+    {
+      "epoch": 12.083067092651758,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 7564
+    },
+    {
+      "epoch": 12.084664536741213,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 7565
+    },
+    {
+      "epoch": 12.08626198083067,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7566
+    },
+    {
+      "epoch": 12.087859424920127,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 7567
+    },
+    {
+      "epoch": 12.089456869009584,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 7568
+    },
+    {
+      "epoch": 12.091054313099042,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 7569
+    },
+    {
+      "epoch": 12.092651757188499,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7570
+    },
+    {
+      "epoch": 12.094249201277956,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 7571
+    },
+    {
+      "epoch": 12.095846645367413,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 7572
+    },
+    {
+      "epoch": 12.09744408945687,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 7573
+    },
+    {
+      "epoch": 12.099041533546325,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1205,
+      "step": 7574
+    },
+    {
+      "epoch": 12.100638977635782,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 7575
+    },
+    {
+      "epoch": 12.10223642172524,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7576
+    },
+    {
+      "epoch": 12.103833865814696,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7577
+    },
+    {
+      "epoch": 12.105431309904153,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 7578
+    },
+    {
+      "epoch": 12.10702875399361,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 7579
+    },
+    {
+      "epoch": 12.108626198083067,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7580
+    },
+    {
+      "epoch": 12.110223642172524,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 7581
+    },
+    {
+      "epoch": 12.111821086261982,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 7582
+    },
+    {
+      "epoch": 12.113418530351439,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7583
+    },
+    {
+      "epoch": 12.115015974440894,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7584
+    },
+    {
+      "epoch": 12.116613418530351,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 7585
+    },
+    {
+      "epoch": 12.118210862619808,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7586
+    },
+    {
+      "epoch": 12.119808306709265,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 7587
+    },
+    {
+      "epoch": 12.121405750798722,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7588
+    },
+    {
+      "epoch": 12.12300319488818,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 7589
+    },
+    {
+      "epoch": 12.124600638977636,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7590
+    },
+    {
+      "epoch": 12.126198083067093,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 7591
+    },
+    {
+      "epoch": 12.12779552715655,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7592
+    },
+    {
+      "epoch": 12.129392971246006,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 7593
+    },
+    {
+      "epoch": 12.130990415335463,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7594
+    },
+    {
+      "epoch": 12.13258785942492,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 7595
+    },
+    {
+      "epoch": 12.134185303514377,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 7596
+    },
+    {
+      "epoch": 12.135782747603834,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7597
+    },
+    {
+      "epoch": 12.13738019169329,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 7598
+    },
+    {
+      "epoch": 12.138977635782748,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 7599
+    },
+    {
+      "epoch": 12.140575079872205,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 7600
+    },
+    {
+      "epoch": 12.142172523961662,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7601
+    },
+    {
+      "epoch": 12.143769968051119,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7602
+    },
+    {
+      "epoch": 12.145367412140574,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 7603
+    },
+    {
+      "epoch": 12.146964856230031,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 7604
+    },
+    {
+      "epoch": 12.148562300319488,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 7605
+    },
+    {
+      "epoch": 12.150159744408946,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7606
+    },
+    {
+      "epoch": 12.151757188498403,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 7607
+    },
+    {
+      "epoch": 12.15335463258786,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 7608
+    },
+    {
+      "epoch": 12.154952076677317,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7609
+    },
+    {
+      "epoch": 12.156549520766774,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7610
+    },
+    {
+      "epoch": 12.15814696485623,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 7611
+    },
+    {
+      "epoch": 12.159744408945686,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7612
+    },
+    {
+      "epoch": 12.161341853035143,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1224,
+      "step": 7613
+    },
+    {
+      "epoch": 12.1629392971246,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7614
+    },
+    {
+      "epoch": 12.164536741214057,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7615
+    },
+    {
+      "epoch": 12.166134185303514,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7616
+    },
+    {
+      "epoch": 12.167731629392971,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 7617
+    },
+    {
+      "epoch": 12.169329073482428,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 7618
+    },
+    {
+      "epoch": 12.170926517571885,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 7619
+    },
+    {
+      "epoch": 12.172523961661343,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 7620
+    },
+    {
+      "epoch": 12.1741214057508,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7621
+    },
+    {
+      "epoch": 12.175718849840255,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 7622
+    },
+    {
+      "epoch": 12.177316293929712,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 7623
+    },
+    {
+      "epoch": 12.178913738019169,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7624
+    },
+    {
+      "epoch": 12.180511182108626,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 7625
+    },
+    {
+      "epoch": 12.182108626198083,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 7626
+    },
+    {
+      "epoch": 12.18370607028754,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1216,
+      "step": 7627
+    },
+    {
+      "epoch": 12.185303514376997,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7628
+    },
+    {
+      "epoch": 12.186900958466454,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7629
+    },
+    {
+      "epoch": 12.188498402555911,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7630
+    },
+    {
+      "epoch": 12.190095846645367,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7631
+    },
+    {
+      "epoch": 12.191693290734824,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 7632
+    },
+    {
+      "epoch": 12.19329073482428,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 7633
+    },
+    {
+      "epoch": 12.194888178913738,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 7634
+    },
+    {
+      "epoch": 12.196485623003195,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 7635
+    },
+    {
+      "epoch": 12.198083067092652,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7636
+    },
+    {
+      "epoch": 12.199680511182109,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7637
+    },
+    {
+      "epoch": 12.201277955271566,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 7638
+    },
+    {
+      "epoch": 12.202875399361023,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7639
+    },
+    {
+      "epoch": 12.204472843450478,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 7640
+    },
+    {
+      "epoch": 12.206070287539935,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7641
+    },
+    {
+      "epoch": 12.207667731629392,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 7642
+    },
+    {
+      "epoch": 12.20926517571885,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 7643
+    },
+    {
+      "epoch": 12.210862619808307,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 7644
+    },
+    {
+      "epoch": 12.212460063897764,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7645
+    },
+    {
+      "epoch": 12.21405750798722,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 7646
+    },
+    {
+      "epoch": 12.215654952076678,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7647
+    },
+    {
+      "epoch": 12.217252396166135,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 7648
+    },
+    {
+      "epoch": 12.218849840255592,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 7649
+    },
+    {
+      "epoch": 12.220447284345047,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7650
+    },
+    {
+      "epoch": 12.222044728434504,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7651
+    },
+    {
+      "epoch": 12.223642172523961,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 7652
+    },
+    {
+      "epoch": 12.225239616613418,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 7653
+    },
+    {
+      "epoch": 12.226837060702875,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7654
+    },
+    {
+      "epoch": 12.228434504792332,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 7655
+    },
+    {
+      "epoch": 12.23003194888179,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7656
+    },
+    {
+      "epoch": 12.231629392971247,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 7657
+    },
+    {
+      "epoch": 12.233226837060704,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7658
+    },
+    {
+      "epoch": 12.23482428115016,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 7659
+    },
+    {
+      "epoch": 12.236421725239616,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 7660
+    },
+    {
+      "epoch": 12.238019169329073,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7661
+    },
+    {
+      "epoch": 12.23961661341853,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 7662
+    },
+    {
+      "epoch": 12.241214057507987,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7663
+    },
+    {
+      "epoch": 12.242811501597444,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 7664
+    },
+    {
+      "epoch": 12.244408945686901,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 7665
+    },
+    {
+      "epoch": 12.246006389776358,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 7666
+    },
+    {
+      "epoch": 12.247603833865815,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7667
+    },
+    {
+      "epoch": 12.249201277955272,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7668
+    },
+    {
+      "epoch": 12.250798722044728,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7669
+    },
+    {
+      "epoch": 12.252396166134185,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7670
+    },
+    {
+      "epoch": 12.253993610223642,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 7671
+    },
+    {
+      "epoch": 12.255591054313099,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7672
+    },
+    {
+      "epoch": 12.257188498402556,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7673
+    },
+    {
+      "epoch": 12.258785942492013,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7674
+    },
+    {
+      "epoch": 12.26038338658147,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 7675
+    },
+    {
+      "epoch": 12.261980830670927,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7676
+    },
+    {
+      "epoch": 12.263578274760384,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 7677
+    },
+    {
+      "epoch": 12.26517571884984,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 7678
+    },
+    {
+      "epoch": 12.266773162939296,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 7679
+    },
+    {
+      "epoch": 12.268370607028753,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 7680
+    },
+    {
+      "epoch": 12.26996805111821,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7681
+    },
+    {
+      "epoch": 12.271565495207668,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 7682
+    },
+    {
+      "epoch": 12.273162939297125,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7683
+    },
+    {
+      "epoch": 12.274760383386582,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 7684
+    },
+    {
+      "epoch": 12.276357827476039,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7685
+    },
+    {
+      "epoch": 12.277955271565496,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7686
+    },
+    {
+      "epoch": 12.279552715654953,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 7687
+    },
+    {
+      "epoch": 12.281150159744408,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7688
+    },
+    {
+      "epoch": 12.282747603833865,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 7689
+    },
+    {
+      "epoch": 12.284345047923322,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7690
+    },
+    {
+      "epoch": 12.28594249201278,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 7691
+    },
+    {
+      "epoch": 12.287539936102236,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 7692
+    },
+    {
+      "epoch": 12.289137380191693,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 7693
+    },
+    {
+      "epoch": 12.29073482428115,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 7694
+    },
+    {
+      "epoch": 12.292332268370608,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 7695
+    },
+    {
+      "epoch": 12.293929712460065,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 7696
+    },
+    {
+      "epoch": 12.295527156549522,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 7697
+    },
+    {
+      "epoch": 12.297124600638977,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7698
+    },
+    {
+      "epoch": 12.298722044728434,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1438,
+      "step": 7699
+    },
+    {
+      "epoch": 12.300319488817891,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 7700
+    },
+    {
+      "epoch": 12.301916932907348,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 7701
+    },
+    {
+      "epoch": 12.303514376996805,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7702
+    },
+    {
+      "epoch": 12.305111821086262,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 7703
+    },
+    {
+      "epoch": 12.30670926517572,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1262,
+      "step": 7704
+    },
+    {
+      "epoch": 12.308306709265176,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 7705
+    },
+    {
+      "epoch": 12.309904153354633,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7706
+    },
+    {
+      "epoch": 12.311501597444089,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7707
+    },
+    {
+      "epoch": 12.313099041533546,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7708
+    },
+    {
+      "epoch": 12.314696485623003,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7709
+    },
+    {
+      "epoch": 12.31629392971246,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 7710
+    },
+    {
+      "epoch": 12.317891373801917,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7711
+    },
+    {
+      "epoch": 12.319488817891374,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7712
+    },
+    {
+      "epoch": 12.321086261980831,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 7713
+    },
+    {
+      "epoch": 12.322683706070288,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7714
+    },
+    {
+      "epoch": 12.324281150159745,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 7715
+    },
+    {
+      "epoch": 12.3258785942492,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7716
+    },
+    {
+      "epoch": 12.327476038338657,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7717
+    },
+    {
+      "epoch": 12.329073482428115,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 7718
+    },
+    {
+      "epoch": 12.330670926517572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7719
+    },
+    {
+      "epoch": 12.332268370607029,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 7720
+    },
+    {
+      "epoch": 12.333865814696486,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 7721
+    },
+    {
+      "epoch": 12.335463258785943,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7722
+    },
+    {
+      "epoch": 12.3370607028754,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 7723
+    },
+    {
+      "epoch": 12.338658146964857,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7724
+    },
+    {
+      "epoch": 12.340255591054314,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7725
+    },
+    {
+      "epoch": 12.34185303514377,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7726
+    },
+    {
+      "epoch": 12.343450479233226,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 7727
+    },
+    {
+      "epoch": 12.345047923322683,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 7728
+    },
+    {
+      "epoch": 12.34664536741214,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 7729
+    },
+    {
+      "epoch": 12.348242811501597,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 7730
+    },
+    {
+      "epoch": 12.349840255591054,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7731
+    },
+    {
+      "epoch": 12.351437699680512,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7732
+    },
+    {
+      "epoch": 12.353035143769969,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 7733
+    },
+    {
+      "epoch": 12.354632587859426,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7734
+    },
+    {
+      "epoch": 12.356230031948883,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 7735
+    },
+    {
+      "epoch": 12.357827476038338,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7736
+    },
+    {
+      "epoch": 12.359424920127795,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 7737
+    },
+    {
+      "epoch": 12.361022364217252,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 7738
+    },
+    {
+      "epoch": 12.36261980830671,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 7739
+    },
+    {
+      "epoch": 12.364217252396166,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7740
+    },
+    {
+      "epoch": 12.365814696485623,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7741
+    },
+    {
+      "epoch": 12.36741214057508,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 7742
+    },
+    {
+      "epoch": 12.369009584664537,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 7743
+    },
+    {
+      "epoch": 12.370607028753994,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 7744
+    },
+    {
+      "epoch": 12.37220447284345,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7745
+    },
+    {
+      "epoch": 12.373801916932907,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7746
+    },
+    {
+      "epoch": 12.375399361022364,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7747
+    },
+    {
+      "epoch": 12.37699680511182,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7748
+    },
+    {
+      "epoch": 12.378594249201278,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7749
+    },
+    {
+      "epoch": 12.380191693290735,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7750
+    },
+    {
+      "epoch": 12.381789137380192,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 7751
+    },
+    {
+      "epoch": 12.383386581469649,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 7752
+    },
+    {
+      "epoch": 12.384984025559106,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 7753
+    },
+    {
+      "epoch": 12.386581469648561,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 7754
+    },
+    {
+      "epoch": 12.388178913738018,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7755
+    },
+    {
+      "epoch": 12.389776357827476,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7756
+    },
+    {
+      "epoch": 12.391373801916933,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 7757
+    },
+    {
+      "epoch": 12.39297124600639,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 7758
+    },
+    {
+      "epoch": 12.394568690095847,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 7759
+    },
+    {
+      "epoch": 12.396166134185304,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 7760
+    },
+    {
+      "epoch": 12.39776357827476,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7761
+    },
+    {
+      "epoch": 12.399361022364218,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7762
+    },
+    {
+      "epoch": 12.400958466453675,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7763
+    },
+    {
+      "epoch": 12.40255591054313,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7764
+    },
+    {
+      "epoch": 12.404153354632587,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 7765
+    },
+    {
+      "epoch": 12.405750798722044,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7766
+    },
+    {
+      "epoch": 12.407348242811501,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 7767
+    },
+    {
+      "epoch": 12.408945686900958,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1383,
+      "step": 7768
+    },
+    {
+      "epoch": 12.410543130990416,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 7769
+    },
+    {
+      "epoch": 12.412140575079873,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 7770
+    },
+    {
+      "epoch": 12.41373801916933,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 7771
+    },
+    {
+      "epoch": 12.415335463258787,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 7772
+    },
+    {
+      "epoch": 12.416932907348242,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 7773
+    },
+    {
+      "epoch": 12.418530351437699,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 7774
+    },
+    {
+      "epoch": 12.420127795527156,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7775
+    },
+    {
+      "epoch": 12.421725239616613,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7776
+    },
+    {
+      "epoch": 12.42332268370607,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7777
+    },
+    {
+      "epoch": 12.424920127795527,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 7778
+    },
+    {
+      "epoch": 12.426517571884984,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 7779
+    },
+    {
+      "epoch": 12.428115015974441,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 7780
+    },
+    {
+      "epoch": 12.429712460063898,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 7781
+    },
+    {
+      "epoch": 12.431309904153355,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7782
+    },
+    {
+      "epoch": 12.43290734824281,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 7783
+    },
+    {
+      "epoch": 12.434504792332268,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1226,
+      "step": 7784
+    },
+    {
+      "epoch": 12.436102236421725,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 7785
+    },
+    {
+      "epoch": 12.437699680511182,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7786
+    },
+    {
+      "epoch": 12.439297124600639,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7787
+    },
+    {
+      "epoch": 12.440894568690096,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 7788
+    },
+    {
+      "epoch": 12.442492012779553,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7789
+    },
+    {
+      "epoch": 12.44408945686901,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 7790
+    },
+    {
+      "epoch": 12.445686900958467,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7791
+    },
+    {
+      "epoch": 12.447284345047922,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7792
+    },
+    {
+      "epoch": 12.44888178913738,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7793
+    },
+    {
+      "epoch": 12.450479233226837,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 7794
+    },
+    {
+      "epoch": 12.452076677316294,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7795
+    },
+    {
+      "epoch": 12.45367412140575,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 7796
+    },
+    {
+      "epoch": 12.455271565495208,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7797
+    },
+    {
+      "epoch": 12.456869009584665,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 7798
+    },
+    {
+      "epoch": 12.458466453674122,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 7799
+    },
+    {
+      "epoch": 12.460063897763579,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 7800
+    },
+    {
+      "epoch": 12.461661341853036,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1409,
+      "step": 7801
+    },
+    {
+      "epoch": 12.463258785942491,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 7802
+    },
+    {
+      "epoch": 12.464856230031948,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7803
+    },
+    {
+      "epoch": 12.466453674121405,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 7804
+    },
+    {
+      "epoch": 12.468051118210862,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 7805
+    },
+    {
+      "epoch": 12.46964856230032,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 7806
+    },
+    {
+      "epoch": 12.471246006389777,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 7807
+    },
+    {
+      "epoch": 12.472843450479234,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 7808
+    },
+    {
+      "epoch": 12.47444089456869,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 7809
+    },
+    {
+      "epoch": 12.476038338658148,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 7810
+    },
+    {
+      "epoch": 12.477635782747603,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7811
+    },
+    {
+      "epoch": 12.47923322683706,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 7812
+    },
+    {
+      "epoch": 12.480830670926517,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 7813
+    },
+    {
+      "epoch": 12.482428115015974,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7814
+    },
+    {
+      "epoch": 12.484025559105431,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7815
+    },
+    {
+      "epoch": 12.485623003194888,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 7816
+    },
+    {
+      "epoch": 12.487220447284345,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 7817
+    },
+    {
+      "epoch": 12.488817891373802,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7818
+    },
+    {
+      "epoch": 12.49041533546326,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7819
+    },
+    {
+      "epoch": 12.492012779552716,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 7820
+    },
+    {
+      "epoch": 12.493610223642172,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7821
+    },
+    {
+      "epoch": 12.495207667731629,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 7822
+    },
+    {
+      "epoch": 12.496805111821086,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 7823
+    },
+    {
+      "epoch": 12.498402555910543,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 7824
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 7825
+    },
+    {
+      "epoch": 12.501597444089457,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7826
+    },
+    {
+      "epoch": 12.503194888178914,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 7827
+    },
+    {
+      "epoch": 12.504792332268371,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7828
+    },
+    {
+      "epoch": 12.506389776357828,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7829
+    },
+    {
+      "epoch": 12.507987220447284,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 7830
+    },
+    {
+      "epoch": 12.50958466453674,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7831
+    },
+    {
+      "epoch": 12.511182108626198,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7832
+    },
+    {
+      "epoch": 12.512779552715655,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 7833
+    },
+    {
+      "epoch": 12.514376996805112,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 7834
+    },
+    {
+      "epoch": 12.515974440894569,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 7835
+    },
+    {
+      "epoch": 12.517571884984026,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7836
+    },
+    {
+      "epoch": 12.519169329073483,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 7837
+    },
+    {
+      "epoch": 12.52076677316294,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 7838
+    },
+    {
+      "epoch": 12.522364217252395,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 7839
+    },
+    {
+      "epoch": 12.523961661341852,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 7840
+    },
+    {
+      "epoch": 12.52555910543131,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 7841
+    },
+    {
+      "epoch": 12.527156549520766,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 7842
+    },
+    {
+      "epoch": 12.528753993610223,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7843
+    },
+    {
+      "epoch": 12.53035143769968,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7844
+    },
+    {
+      "epoch": 12.531948881789138,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 7845
+    },
+    {
+      "epoch": 12.533546325878595,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 7846
+    },
+    {
+      "epoch": 12.535143769968052,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7847
+    },
+    {
+      "epoch": 12.536741214057509,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 7848
+    },
+    {
+      "epoch": 12.538338658146966,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7849
+    },
+    {
+      "epoch": 12.539936102236421,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 7850
+    },
+    {
+      "epoch": 12.541533546325878,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 7851
+    },
+    {
+      "epoch": 12.543130990415335,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 7852
+    },
+    {
+      "epoch": 12.544728434504792,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7853
+    },
+    {
+      "epoch": 12.54632587859425,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 7854
+    },
+    {
+      "epoch": 12.547923322683706,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 7855
+    },
+    {
+      "epoch": 12.549520766773163,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 7856
+    },
+    {
+      "epoch": 12.55111821086262,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7857
+    },
+    {
+      "epoch": 12.552715654952078,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7858
+    },
+    {
+      "epoch": 12.554313099041533,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7859
+    },
+    {
+      "epoch": 12.55591054313099,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 7860
+    },
+    {
+      "epoch": 12.557507987220447,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 7861
+    },
+    {
+      "epoch": 12.559105431309904,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1219,
+      "step": 7862
+    },
+    {
+      "epoch": 12.560702875399361,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7863
+    },
+    {
+      "epoch": 12.562300319488818,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 7864
+    },
+    {
+      "epoch": 12.563897763578275,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 7865
+    },
+    {
+      "epoch": 12.565495207667732,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 7866
+    },
+    {
+      "epoch": 12.56709265175719,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 7867
+    },
+    {
+      "epoch": 12.568690095846645,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 7868
+    },
+    {
+      "epoch": 12.570287539936102,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7869
+    },
+    {
+      "epoch": 12.571884984025559,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7870
+    },
+    {
+      "epoch": 12.573482428115016,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 7871
+    },
+    {
+      "epoch": 12.575079872204473,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7872
+    },
+    {
+      "epoch": 12.57667731629393,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7873
+    },
+    {
+      "epoch": 12.578274760383387,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 7874
+    },
+    {
+      "epoch": 12.579872204472844,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 7875
+    },
+    {
+      "epoch": 12.581469648562301,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7876
+    },
+    {
+      "epoch": 12.583067092651756,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 7877
+    },
+    {
+      "epoch": 12.584664536741213,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7878
+    },
+    {
+      "epoch": 12.58626198083067,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 7879
+    },
+    {
+      "epoch": 12.587859424920127,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 7880
+    },
+    {
+      "epoch": 12.589456869009584,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 7881
+    },
+    {
+      "epoch": 12.591054313099042,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 7882
+    },
+    {
+      "epoch": 12.592651757188499,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 7883
+    },
+    {
+      "epoch": 12.594249201277956,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7884
+    },
+    {
+      "epoch": 12.595846645367413,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 7885
+    },
+    {
+      "epoch": 12.59744408945687,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7886
+    },
+    {
+      "epoch": 12.599041533546325,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 7887
+    },
+    {
+      "epoch": 12.600638977635782,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 7888
+    },
+    {
+      "epoch": 12.60223642172524,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 7889
+    },
+    {
+      "epoch": 12.603833865814696,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 7890
+    },
+    {
+      "epoch": 12.605431309904153,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 7891
+    },
+    {
+      "epoch": 12.60702875399361,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 7892
+    },
+    {
+      "epoch": 12.608626198083067,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7893
+    },
+    {
+      "epoch": 12.610223642172524,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 7894
+    },
+    {
+      "epoch": 12.611821086261982,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 7895
+    },
+    {
+      "epoch": 12.613418530351439,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7896
+    },
+    {
+      "epoch": 12.615015974440894,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 7897
+    },
+    {
+      "epoch": 12.616613418530351,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 7898
+    },
+    {
+      "epoch": 12.618210862619808,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.122,
+      "step": 7899
+    },
+    {
+      "epoch": 12.619808306709265,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 7900
+    },
+    {
+      "epoch": 12.621405750798722,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7901
+    },
+    {
+      "epoch": 12.62300319488818,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 7902
+    },
+    {
+      "epoch": 12.624600638977636,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1217,
+      "step": 7903
+    },
+    {
+      "epoch": 12.626198083067093,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7904
+    },
+    {
+      "epoch": 12.62779552715655,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 7905
+    },
+    {
+      "epoch": 12.629392971246006,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 7906
+    },
+    {
+      "epoch": 12.630990415335463,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 7907
+    },
+    {
+      "epoch": 12.63258785942492,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7908
+    },
+    {
+      "epoch": 12.634185303514377,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 7909
+    },
+    {
+      "epoch": 12.635782747603834,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 7910
+    },
+    {
+      "epoch": 12.63738019169329,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 7911
+    },
+    {
+      "epoch": 12.638977635782748,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 7912
+    },
+    {
+      "epoch": 12.640575079872205,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 7913
+    },
+    {
+      "epoch": 12.642172523961662,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 7914
+    },
+    {
+      "epoch": 12.643769968051117,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 7915
+    },
+    {
+      "epoch": 12.645367412140574,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 7916
+    },
+    {
+      "epoch": 12.646964856230031,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7917
+    },
+    {
+      "epoch": 12.648562300319488,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 7918
+    },
+    {
+      "epoch": 12.650159744408946,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 7919
+    },
+    {
+      "epoch": 12.651757188498403,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 7920
+    },
+    {
+      "epoch": 12.65335463258786,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 7921
+    },
+    {
+      "epoch": 12.654952076677317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 7922
+    },
+    {
+      "epoch": 12.656549520766774,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 7923
+    },
+    {
+      "epoch": 12.65814696485623,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7924
+    },
+    {
+      "epoch": 12.659744408945686,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 7925
+    },
+    {
+      "epoch": 12.661341853035143,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 7926
+    },
+    {
+      "epoch": 12.6629392971246,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1166,
+      "step": 7927
+    },
+    {
+      "epoch": 12.664536741214057,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 7928
+    },
+    {
+      "epoch": 12.666134185303514,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 7929
+    },
+    {
+      "epoch": 12.667731629392971,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 7930
+    },
+    {
+      "epoch": 12.669329073482428,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 7931
+    },
+    {
+      "epoch": 12.670926517571885,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 7932
+    },
+    {
+      "epoch": 12.672523961661343,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 7933
+    },
+    {
+      "epoch": 12.6741214057508,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 7934
+    },
+    {
+      "epoch": 12.675718849840255,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7935
+    },
+    {
+      "epoch": 12.677316293929712,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 7936
+    },
+    {
+      "epoch": 12.678913738019169,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 7937
+    },
+    {
+      "epoch": 12.680511182108626,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7938
+    },
+    {
+      "epoch": 12.682108626198083,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 7939
+    },
+    {
+      "epoch": 12.68370607028754,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 7940
+    },
+    {
+      "epoch": 12.685303514376997,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 7941
+    },
+    {
+      "epoch": 12.686900958466454,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 7942
+    },
+    {
+      "epoch": 12.688498402555911,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 7943
+    },
+    {
+      "epoch": 12.690095846645367,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 7944
+    },
+    {
+      "epoch": 12.691693290734824,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7945
+    },
+    {
+      "epoch": 12.69329073482428,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 7946
+    },
+    {
+      "epoch": 12.694888178913738,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 7947
+    },
+    {
+      "epoch": 12.696485623003195,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7948
+    },
+    {
+      "epoch": 12.698083067092652,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 7949
+    },
+    {
+      "epoch": 12.699680511182109,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 7950
+    },
+    {
+      "epoch": 12.701277955271566,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7951
+    },
+    {
+      "epoch": 12.702875399361023,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 7952
+    },
+    {
+      "epoch": 12.704472843450478,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 7953
+    },
+    {
+      "epoch": 12.706070287539935,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 7954
+    },
+    {
+      "epoch": 12.707667731629392,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 7955
+    },
+    {
+      "epoch": 12.70926517571885,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 7956
+    },
+    {
+      "epoch": 12.710862619808307,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 7957
+    },
+    {
+      "epoch": 12.712460063897764,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 7958
+    },
+    {
+      "epoch": 12.71405750798722,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 7959
+    },
+    {
+      "epoch": 12.715654952076678,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 7960
+    },
+    {
+      "epoch": 12.717252396166135,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 7961
+    },
+    {
+      "epoch": 12.718849840255592,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 7962
+    },
+    {
+      "epoch": 12.720447284345047,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 7963
+    },
+    {
+      "epoch": 12.722044728434504,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7964
+    },
+    {
+      "epoch": 12.723642172523961,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 7965
+    },
+    {
+      "epoch": 12.725239616613418,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1236,
+      "step": 7966
+    },
+    {
+      "epoch": 12.726837060702875,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 7967
+    },
+    {
+      "epoch": 12.728434504792332,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7968
+    },
+    {
+      "epoch": 12.73003194888179,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 7969
+    },
+    {
+      "epoch": 12.731629392971247,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 7970
+    },
+    {
+      "epoch": 12.733226837060704,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 7971
+    },
+    {
+      "epoch": 12.73482428115016,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 7972
+    },
+    {
+      "epoch": 12.736421725239616,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 7973
+    },
+    {
+      "epoch": 12.738019169329073,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 7974
+    },
+    {
+      "epoch": 12.73961661341853,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 7975
+    },
+    {
+      "epoch": 12.741214057507987,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 7976
+    },
+    {
+      "epoch": 12.742811501597444,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 7977
+    },
+    {
+      "epoch": 12.744408945686901,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 7978
+    },
+    {
+      "epoch": 12.746006389776358,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7979
+    },
+    {
+      "epoch": 12.747603833865815,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 7980
+    },
+    {
+      "epoch": 12.749201277955272,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 7981
+    },
+    {
+      "epoch": 12.750798722044728,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 7982
+    },
+    {
+      "epoch": 12.752396166134185,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7983
+    },
+    {
+      "epoch": 12.753993610223642,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 7984
+    },
+    {
+      "epoch": 12.755591054313099,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1262,
+      "step": 7985
+    },
+    {
+      "epoch": 12.757188498402556,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 7986
+    },
+    {
+      "epoch": 12.758785942492013,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 7987
+    },
+    {
+      "epoch": 12.76038338658147,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 7988
+    },
+    {
+      "epoch": 12.761980830670927,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1181,
+      "step": 7989
+    },
+    {
+      "epoch": 12.763578274760384,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 7990
+    },
+    {
+      "epoch": 12.76517571884984,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 7991
+    },
+    {
+      "epoch": 12.766773162939296,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 7992
+    },
+    {
+      "epoch": 12.768370607028753,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 7993
+    },
+    {
+      "epoch": 12.76996805111821,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1223,
+      "step": 7994
+    },
+    {
+      "epoch": 12.771565495207668,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 7995
+    },
+    {
+      "epoch": 12.773162939297125,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 7996
+    },
+    {
+      "epoch": 12.774760383386582,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 7997
+    },
+    {
+      "epoch": 12.776357827476039,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 7998
+    },
+    {
+      "epoch": 12.777955271565496,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 7999
+    },
+    {
+      "epoch": 12.779552715654953,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8000
+    },
+    {
+      "epoch": 12.781150159744408,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 8001
+    },
+    {
+      "epoch": 12.782747603833865,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8002
+    },
+    {
+      "epoch": 12.784345047923322,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8003
+    },
+    {
+      "epoch": 12.78594249201278,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8004
+    },
+    {
+      "epoch": 12.787539936102236,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 8005
+    },
+    {
+      "epoch": 12.789137380191693,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 8006
+    },
+    {
+      "epoch": 12.79073482428115,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 8007
+    },
+    {
+      "epoch": 12.792332268370608,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8008
+    },
+    {
+      "epoch": 12.793929712460065,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 8009
+    },
+    {
+      "epoch": 12.795527156549522,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 8010
+    },
+    {
+      "epoch": 12.797124600638977,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8011
+    },
+    {
+      "epoch": 12.798722044728434,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 8012
+    },
+    {
+      "epoch": 12.800319488817891,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8013
+    },
+    {
+      "epoch": 12.801916932907348,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 8014
+    },
+    {
+      "epoch": 12.803514376996805,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8015
+    },
+    {
+      "epoch": 12.805111821086262,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 8016
+    },
+    {
+      "epoch": 12.80670926517572,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 8017
+    },
+    {
+      "epoch": 12.808306709265176,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8018
+    },
+    {
+      "epoch": 12.809904153354633,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 8019
+    },
+    {
+      "epoch": 12.811501597444089,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8020
+    },
+    {
+      "epoch": 12.813099041533546,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 8021
+    },
+    {
+      "epoch": 12.814696485623003,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 8022
+    },
+    {
+      "epoch": 12.81629392971246,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 8023
+    },
+    {
+      "epoch": 12.817891373801917,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 8024
+    },
+    {
+      "epoch": 12.819488817891374,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 8025
+    },
+    {
+      "epoch": 12.821086261980831,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 8026
+    },
+    {
+      "epoch": 12.822683706070288,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1219,
+      "step": 8027
+    },
+    {
+      "epoch": 12.824281150159745,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 8028
+    },
+    {
+      "epoch": 12.8258785942492,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8029
+    },
+    {
+      "epoch": 12.827476038338657,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 8030
+    },
+    {
+      "epoch": 12.829073482428115,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8031
+    },
+    {
+      "epoch": 12.830670926517572,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1194,
+      "step": 8032
+    },
+    {
+      "epoch": 12.832268370607029,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8033
+    },
+    {
+      "epoch": 12.833865814696486,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 8034
+    },
+    {
+      "epoch": 12.835463258785943,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 8035
+    },
+    {
+      "epoch": 12.8370607028754,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 8036
+    },
+    {
+      "epoch": 12.838658146964857,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8037
+    },
+    {
+      "epoch": 12.840255591054314,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 8038
+    },
+    {
+      "epoch": 12.84185303514377,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 8039
+    },
+    {
+      "epoch": 12.843450479233226,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8040
+    },
+    {
+      "epoch": 12.845047923322683,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 8041
+    },
+    {
+      "epoch": 12.84664536741214,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 8042
+    },
+    {
+      "epoch": 12.848242811501597,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8043
+    },
+    {
+      "epoch": 12.849840255591054,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8044
+    },
+    {
+      "epoch": 12.851437699680512,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 8045
+    },
+    {
+      "epoch": 12.853035143769969,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 8046
+    },
+    {
+      "epoch": 12.854632587859426,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 8047
+    },
+    {
+      "epoch": 12.856230031948883,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 8048
+    },
+    {
+      "epoch": 12.857827476038338,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 8049
+    },
+    {
+      "epoch": 12.859424920127795,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8050
+    },
+    {
+      "epoch": 12.861022364217252,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 8051
+    },
+    {
+      "epoch": 12.86261980830671,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 8052
+    },
+    {
+      "epoch": 12.864217252396166,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 8053
+    },
+    {
+      "epoch": 12.865814696485623,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8054
+    },
+    {
+      "epoch": 12.86741214057508,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8055
+    },
+    {
+      "epoch": 12.869009584664537,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 8056
+    },
+    {
+      "epoch": 12.870607028753994,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 8057
+    },
+    {
+      "epoch": 12.87220447284345,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 8058
+    },
+    {
+      "epoch": 12.873801916932907,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 8059
+    },
+    {
+      "epoch": 12.875399361022364,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 8060
+    },
+    {
+      "epoch": 12.87699680511182,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8061
+    },
+    {
+      "epoch": 12.878594249201278,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8062
+    },
+    {
+      "epoch": 12.880191693290735,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 8063
+    },
+    {
+      "epoch": 12.881789137380192,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8064
+    },
+    {
+      "epoch": 12.883386581469649,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8065
+    },
+    {
+      "epoch": 12.884984025559106,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 8066
+    },
+    {
+      "epoch": 12.886581469648561,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8067
+    },
+    {
+      "epoch": 12.888178913738018,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1389,
+      "step": 8068
+    },
+    {
+      "epoch": 12.889776357827476,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1173,
+      "step": 8069
+    },
+    {
+      "epoch": 12.891373801916933,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 8070
+    },
+    {
+      "epoch": 12.89297124600639,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 8071
+    },
+    {
+      "epoch": 12.894568690095847,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8072
+    },
+    {
+      "epoch": 12.896166134185304,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8073
+    },
+    {
+      "epoch": 12.89776357827476,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1239,
+      "step": 8074
+    },
+    {
+      "epoch": 12.899361022364218,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 8075
+    },
+    {
+      "epoch": 12.900958466453675,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8076
+    },
+    {
+      "epoch": 12.90255591054313,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 8077
+    },
+    {
+      "epoch": 12.904153354632587,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8078
+    },
+    {
+      "epoch": 12.905750798722044,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8079
+    },
+    {
+      "epoch": 12.907348242811501,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 8080
+    },
+    {
+      "epoch": 12.908945686900958,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 8081
+    },
+    {
+      "epoch": 12.910543130990416,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8082
+    },
+    {
+      "epoch": 12.912140575079873,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 8083
+    },
+    {
+      "epoch": 12.91373801916933,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8084
+    },
+    {
+      "epoch": 12.915335463258787,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 8085
+    },
+    {
+      "epoch": 12.916932907348244,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 8086
+    },
+    {
+      "epoch": 12.918530351437699,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8087
+    },
+    {
+      "epoch": 12.920127795527156,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 8088
+    },
+    {
+      "epoch": 12.921725239616613,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8089
+    },
+    {
+      "epoch": 12.92332268370607,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 8090
+    },
+    {
+      "epoch": 12.924920127795527,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 8091
+    },
+    {
+      "epoch": 12.926517571884984,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 8092
+    },
+    {
+      "epoch": 12.928115015974441,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 8093
+    },
+    {
+      "epoch": 12.929712460063898,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 8094
+    },
+    {
+      "epoch": 12.931309904153355,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8095
+    },
+    {
+      "epoch": 12.93290734824281,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8096
+    },
+    {
+      "epoch": 12.934504792332268,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 8097
+    },
+    {
+      "epoch": 12.936102236421725,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8098
+    },
+    {
+      "epoch": 12.937699680511182,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 8099
+    },
+    {
+      "epoch": 12.939297124600639,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 8100
+    },
+    {
+      "epoch": 12.940894568690096,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1217,
+      "step": 8101
+    },
+    {
+      "epoch": 12.942492012779553,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8102
+    },
+    {
+      "epoch": 12.94408945686901,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 8103
+    },
+    {
+      "epoch": 12.945686900958467,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 8104
+    },
+    {
+      "epoch": 12.947284345047922,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8105
+    },
+    {
+      "epoch": 12.94888178913738,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 8106
+    },
+    {
+      "epoch": 12.950479233226837,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 8107
+    },
+    {
+      "epoch": 12.952076677316294,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 8108
+    },
+    {
+      "epoch": 12.95367412140575,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8109
+    },
+    {
+      "epoch": 12.955271565495208,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8110
+    },
+    {
+      "epoch": 12.956869009584665,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8111
+    },
+    {
+      "epoch": 12.958466453674122,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 8112
+    },
+    {
+      "epoch": 12.960063897763579,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 8113
+    },
+    {
+      "epoch": 12.961661341853034,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 8114
+    },
+    {
+      "epoch": 12.963258785942491,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 8115
+    },
+    {
+      "epoch": 12.964856230031948,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 8116
+    },
+    {
+      "epoch": 12.966453674121405,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8117
+    },
+    {
+      "epoch": 12.968051118210862,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8118
+    },
+    {
+      "epoch": 12.96964856230032,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8119
+    },
+    {
+      "epoch": 12.971246006389777,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8120
+    },
+    {
+      "epoch": 12.972843450479234,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8121
+    },
+    {
+      "epoch": 12.97444089456869,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 8122
+    },
+    {
+      "epoch": 12.976038338658148,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8123
+    },
+    {
+      "epoch": 12.977635782747605,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 8124
+    },
+    {
+      "epoch": 12.97923322683706,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8125
+    },
+    {
+      "epoch": 12.980830670926517,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8126
+    },
+    {
+      "epoch": 12.982428115015974,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8127
+    },
+    {
+      "epoch": 12.984025559105431,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8128
+    },
+    {
+      "epoch": 12.985623003194888,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 8129
+    },
+    {
+      "epoch": 12.987220447284345,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8130
+    },
+    {
+      "epoch": 12.988817891373802,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 8131
+    },
+    {
+      "epoch": 12.99041533546326,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8132
+    },
+    {
+      "epoch": 12.992012779552716,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8133
+    },
+    {
+      "epoch": 12.993610223642172,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 8134
+    },
+    {
+      "epoch": 12.995207667731629,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8135
+    },
+    {
+      "epoch": 12.996805111821086,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8136
+    },
+    {
+      "epoch": 12.998402555910543,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8137
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8138
+    },
+    {
+      "epoch": 13.001597444089457,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 8139
+    },
+    {
+      "epoch": 13.003194888178914,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8140
+    },
+    {
+      "epoch": 13.004792332268371,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 8141
+    },
+    {
+      "epoch": 13.006389776357828,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 8142
+    },
+    {
+      "epoch": 13.007987220447284,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8143
+    },
+    {
+      "epoch": 13.00958466453674,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8144
+    },
+    {
+      "epoch": 13.011182108626198,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1211,
+      "step": 8145
+    },
+    {
+      "epoch": 13.012779552715655,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8146
+    },
+    {
+      "epoch": 13.014376996805112,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 8147
+    },
+    {
+      "epoch": 13.015974440894569,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 8148
+    },
+    {
+      "epoch": 13.017571884984026,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 8149
+    },
+    {
+      "epoch": 13.019169329073483,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 8150
+    },
+    {
+      "epoch": 13.02076677316294,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 8151
+    },
+    {
+      "epoch": 13.022364217252397,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 8152
+    },
+    {
+      "epoch": 13.023961661341852,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 8153
+    },
+    {
+      "epoch": 13.02555910543131,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8154
+    },
+    {
+      "epoch": 13.027156549520766,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 8155
+    },
+    {
+      "epoch": 13.028753993610223,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1452,
+      "step": 8156
+    },
+    {
+      "epoch": 13.03035143769968,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1228,
+      "step": 8157
+    },
+    {
+      "epoch": 13.031948881789138,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 8158
+    },
+    {
+      "epoch": 13.033546325878595,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 8159
+    },
+    {
+      "epoch": 13.035143769968052,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 8160
+    },
+    {
+      "epoch": 13.036741214057509,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 8161
+    },
+    {
+      "epoch": 13.038338658146964,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 8162
+    },
+    {
+      "epoch": 13.039936102236421,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8163
+    },
+    {
+      "epoch": 13.041533546325878,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 8164
+    },
+    {
+      "epoch": 13.043130990415335,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 8165
+    },
+    {
+      "epoch": 13.044728434504792,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8166
+    },
+    {
+      "epoch": 13.04632587859425,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 8167
+    },
+    {
+      "epoch": 13.047923322683706,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1237,
+      "step": 8168
+    },
+    {
+      "epoch": 13.049520766773163,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8169
+    },
+    {
+      "epoch": 13.05111821086262,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8170
+    },
+    {
+      "epoch": 13.052715654952078,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 8171
+    },
+    {
+      "epoch": 13.054313099041533,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8172
+    },
+    {
+      "epoch": 13.05591054313099,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 8173
+    },
+    {
+      "epoch": 13.057507987220447,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 8174
+    },
+    {
+      "epoch": 13.059105431309904,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 8175
+    },
+    {
+      "epoch": 13.060702875399361,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8176
+    },
+    {
+      "epoch": 13.062300319488818,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8177
+    },
+    {
+      "epoch": 13.063897763578275,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1414,
+      "step": 8178
+    },
+    {
+      "epoch": 13.065495207667732,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 8179
+    },
+    {
+      "epoch": 13.06709265175719,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 8180
+    },
+    {
+      "epoch": 13.068690095846645,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 8181
+    },
+    {
+      "epoch": 13.070287539936102,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8182
+    },
+    {
+      "epoch": 13.071884984025559,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1439,
+      "step": 8183
+    },
+    {
+      "epoch": 13.073482428115016,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 8184
+    },
+    {
+      "epoch": 13.075079872204473,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1429,
+      "step": 8185
+    },
+    {
+      "epoch": 13.07667731629393,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.122,
+      "step": 8186
+    },
+    {
+      "epoch": 13.078274760383387,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 8187
+    },
+    {
+      "epoch": 13.079872204472844,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8188
+    },
+    {
+      "epoch": 13.081469648562301,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8189
+    },
+    {
+      "epoch": 13.083067092651758,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 8190
+    },
+    {
+      "epoch": 13.084664536741213,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1175,
+      "step": 8191
+    },
+    {
+      "epoch": 13.08626198083067,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8192
+    },
+    {
+      "epoch": 13.087859424920127,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8193
+    },
+    {
+      "epoch": 13.089456869009584,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8194
+    },
+    {
+      "epoch": 13.091054313099042,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8195
+    },
+    {
+      "epoch": 13.092651757188499,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8196
+    },
+    {
+      "epoch": 13.094249201277956,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 8197
+    },
+    {
+      "epoch": 13.095846645367413,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8198
+    },
+    {
+      "epoch": 13.09744408945687,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8199
+    },
+    {
+      "epoch": 13.099041533546325,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 8200
+    },
+    {
+      "epoch": 13.100638977635782,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 8201
+    },
+    {
+      "epoch": 13.10223642172524,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 8202
+    },
+    {
+      "epoch": 13.103833865814696,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 8203
+    },
+    {
+      "epoch": 13.105431309904153,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 8204
+    },
+    {
+      "epoch": 13.10702875399361,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8205
+    },
+    {
+      "epoch": 13.108626198083067,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 8206
+    },
+    {
+      "epoch": 13.110223642172524,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1211,
+      "step": 8207
+    },
+    {
+      "epoch": 13.111821086261982,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8208
+    },
+    {
+      "epoch": 13.113418530351439,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8209
+    },
+    {
+      "epoch": 13.115015974440894,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8210
+    },
+    {
+      "epoch": 13.116613418530351,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8211
+    },
+    {
+      "epoch": 13.118210862619808,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8212
+    },
+    {
+      "epoch": 13.119808306709265,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8213
+    },
+    {
+      "epoch": 13.121405750798722,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8214
+    },
+    {
+      "epoch": 13.12300319488818,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 8215
+    },
+    {
+      "epoch": 13.124600638977636,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8216
+    },
+    {
+      "epoch": 13.126198083067093,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8217
+    },
+    {
+      "epoch": 13.12779552715655,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8218
+    },
+    {
+      "epoch": 13.129392971246006,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8219
+    },
+    {
+      "epoch": 13.130990415335463,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 8220
+    },
+    {
+      "epoch": 13.13258785942492,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 8221
+    },
+    {
+      "epoch": 13.134185303514377,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8222
+    },
+    {
+      "epoch": 13.135782747603834,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 8223
+    },
+    {
+      "epoch": 13.13738019169329,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 8224
+    },
+    {
+      "epoch": 13.138977635782748,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8225
+    },
+    {
+      "epoch": 13.140575079872205,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 8226
+    },
+    {
+      "epoch": 13.142172523961662,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 8227
+    },
+    {
+      "epoch": 13.143769968051119,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8228
+    },
+    {
+      "epoch": 13.145367412140574,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8229
+    },
+    {
+      "epoch": 13.146964856230031,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8230
+    },
+    {
+      "epoch": 13.148562300319488,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 8231
+    },
+    {
+      "epoch": 13.150159744408946,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8232
+    },
+    {
+      "epoch": 13.151757188498403,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8233
+    },
+    {
+      "epoch": 13.15335463258786,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8234
+    },
+    {
+      "epoch": 13.154952076677317,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8235
+    },
+    {
+      "epoch": 13.156549520766774,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 8236
+    },
+    {
+      "epoch": 13.15814696485623,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8237
+    },
+    {
+      "epoch": 13.159744408945686,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8238
+    },
+    {
+      "epoch": 13.161341853035143,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 8239
+    },
+    {
+      "epoch": 13.1629392971246,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 8240
+    },
+    {
+      "epoch": 13.164536741214057,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8241
+    },
+    {
+      "epoch": 13.166134185303514,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8242
+    },
+    {
+      "epoch": 13.167731629392971,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 8243
+    },
+    {
+      "epoch": 13.169329073482428,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8244
+    },
+    {
+      "epoch": 13.170926517571885,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 8245
+    },
+    {
+      "epoch": 13.172523961661343,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 8246
+    },
+    {
+      "epoch": 13.1741214057508,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1398,
+      "step": 8247
+    },
+    {
+      "epoch": 13.175718849840255,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8248
+    },
+    {
+      "epoch": 13.177316293929712,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8249
+    },
+    {
+      "epoch": 13.178913738019169,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8250
+    },
+    {
+      "epoch": 13.180511182108626,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 8251
+    },
+    {
+      "epoch": 13.182108626198083,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 8252
+    },
+    {
+      "epoch": 13.18370607028754,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 8253
+    },
+    {
+      "epoch": 13.185303514376997,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8254
+    },
+    {
+      "epoch": 13.186900958466454,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8255
+    },
+    {
+      "epoch": 13.188498402555911,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 8256
+    },
+    {
+      "epoch": 13.190095846645367,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8257
+    },
+    {
+      "epoch": 13.191693290734824,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8258
+    },
+    {
+      "epoch": 13.19329073482428,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 8259
+    },
+    {
+      "epoch": 13.194888178913738,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8260
+    },
+    {
+      "epoch": 13.196485623003195,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 8261
+    },
+    {
+      "epoch": 13.198083067092652,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8262
+    },
+    {
+      "epoch": 13.199680511182109,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8263
+    },
+    {
+      "epoch": 13.201277955271566,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8264
+    },
+    {
+      "epoch": 13.202875399361023,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1151,
+      "step": 8265
+    },
+    {
+      "epoch": 13.204472843450478,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 8266
+    },
+    {
+      "epoch": 13.206070287539935,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8267
+    },
+    {
+      "epoch": 13.207667731629392,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8268
+    },
+    {
+      "epoch": 13.20926517571885,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 8269
+    },
+    {
+      "epoch": 13.210862619808307,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8270
+    },
+    {
+      "epoch": 13.212460063897764,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 8271
+    },
+    {
+      "epoch": 13.21405750798722,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8272
+    },
+    {
+      "epoch": 13.215654952076678,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1204,
+      "step": 8273
+    },
+    {
+      "epoch": 13.217252396166135,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 8274
+    },
+    {
+      "epoch": 13.218849840255592,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 8275
+    },
+    {
+      "epoch": 13.220447284345047,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1175,
+      "step": 8276
+    },
+    {
+      "epoch": 13.222044728434504,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 8277
+    },
+    {
+      "epoch": 13.223642172523961,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 8278
+    },
+    {
+      "epoch": 13.225239616613418,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8279
+    },
+    {
+      "epoch": 13.226837060702875,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 8280
+    },
+    {
+      "epoch": 13.228434504792332,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 8281
+    },
+    {
+      "epoch": 13.23003194888179,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 8282
+    },
+    {
+      "epoch": 13.231629392971247,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 8283
+    },
+    {
+      "epoch": 13.233226837060704,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 8284
+    },
+    {
+      "epoch": 13.23482428115016,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 8285
+    },
+    {
+      "epoch": 13.236421725239616,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8286
+    },
+    {
+      "epoch": 13.238019169329073,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8287
+    },
+    {
+      "epoch": 13.23961661341853,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 8288
+    },
+    {
+      "epoch": 13.241214057507987,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8289
+    },
+    {
+      "epoch": 13.242811501597444,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8290
+    },
+    {
+      "epoch": 13.244408945686901,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8291
+    },
+    {
+      "epoch": 13.246006389776358,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 8292
+    },
+    {
+      "epoch": 13.247603833865815,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 8293
+    },
+    {
+      "epoch": 13.249201277955272,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 8294
+    },
+    {
+      "epoch": 13.250798722044728,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8295
+    },
+    {
+      "epoch": 13.252396166134185,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 8296
+    },
+    {
+      "epoch": 13.253993610223642,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8297
+    },
+    {
+      "epoch": 13.255591054313099,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8298
+    },
+    {
+      "epoch": 13.257188498402556,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8299
+    },
+    {
+      "epoch": 13.258785942492013,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8300
+    },
+    {
+      "epoch": 13.26038338658147,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 8301
+    },
+    {
+      "epoch": 13.261980830670927,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8302
+    },
+    {
+      "epoch": 13.263578274760384,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 8303
+    },
+    {
+      "epoch": 13.26517571884984,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 8304
+    },
+    {
+      "epoch": 13.266773162939296,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8305
+    },
+    {
+      "epoch": 13.268370607028753,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8306
+    },
+    {
+      "epoch": 13.26996805111821,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 8307
+    },
+    {
+      "epoch": 13.271565495207668,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 8308
+    },
+    {
+      "epoch": 13.273162939297125,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1416,
+      "step": 8309
+    },
+    {
+      "epoch": 13.274760383386582,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8310
+    },
+    {
+      "epoch": 13.276357827476039,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 8311
+    },
+    {
+      "epoch": 13.277955271565496,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 8312
+    },
+    {
+      "epoch": 13.279552715654953,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 8313
+    },
+    {
+      "epoch": 13.281150159744408,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 8314
+    },
+    {
+      "epoch": 13.282747603833865,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1214,
+      "step": 8315
+    },
+    {
+      "epoch": 13.284345047923322,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8316
+    },
+    {
+      "epoch": 13.28594249201278,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1246,
+      "step": 8317
+    },
+    {
+      "epoch": 13.287539936102236,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 8318
+    },
+    {
+      "epoch": 13.289137380191693,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.122,
+      "step": 8319
+    },
+    {
+      "epoch": 13.29073482428115,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 8320
+    },
+    {
+      "epoch": 13.292332268370608,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 8321
+    },
+    {
+      "epoch": 13.293929712460065,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 8322
+    },
+    {
+      "epoch": 13.295527156549522,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 8323
+    },
+    {
+      "epoch": 13.297124600638977,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8324
+    },
+    {
+      "epoch": 13.298722044728434,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 8325
+    },
+    {
+      "epoch": 13.300319488817891,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8326
+    },
+    {
+      "epoch": 13.301916932907348,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 8327
+    },
+    {
+      "epoch": 13.303514376996805,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 8328
+    },
+    {
+      "epoch": 13.305111821086262,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 8329
+    },
+    {
+      "epoch": 13.30670926517572,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 8330
+    },
+    {
+      "epoch": 13.308306709265176,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 8331
+    },
+    {
+      "epoch": 13.309904153354633,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8332
+    },
+    {
+      "epoch": 13.311501597444089,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 8333
+    },
+    {
+      "epoch": 13.313099041533546,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8334
+    },
+    {
+      "epoch": 13.314696485623003,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 8335
+    },
+    {
+      "epoch": 13.31629392971246,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1211,
+      "step": 8336
+    },
+    {
+      "epoch": 13.317891373801917,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8337
+    },
+    {
+      "epoch": 13.319488817891374,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 8338
+    },
+    {
+      "epoch": 13.321086261980831,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 8339
+    },
+    {
+      "epoch": 13.322683706070288,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8340
+    },
+    {
+      "epoch": 13.324281150159745,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1224,
+      "step": 8341
+    },
+    {
+      "epoch": 13.3258785942492,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 8342
+    },
+    {
+      "epoch": 13.327476038338657,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 8343
+    },
+    {
+      "epoch": 13.329073482428115,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8344
+    },
+    {
+      "epoch": 13.330670926517572,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 8345
+    },
+    {
+      "epoch": 13.332268370607029,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 8346
+    },
+    {
+      "epoch": 13.333865814696486,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8347
+    },
+    {
+      "epoch": 13.335463258785943,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8348
+    },
+    {
+      "epoch": 13.3370607028754,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 8349
+    },
+    {
+      "epoch": 13.338658146964857,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8350
+    },
+    {
+      "epoch": 13.340255591054314,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8351
+    },
+    {
+      "epoch": 13.34185303514377,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8352
+    },
+    {
+      "epoch": 13.343450479233226,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8353
+    },
+    {
+      "epoch": 13.345047923322683,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8354
+    },
+    {
+      "epoch": 13.34664536741214,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8355
+    },
+    {
+      "epoch": 13.348242811501597,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8356
+    },
+    {
+      "epoch": 13.349840255591054,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8357
+    },
+    {
+      "epoch": 13.351437699680512,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 8358
+    },
+    {
+      "epoch": 13.353035143769969,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8359
+    },
+    {
+      "epoch": 13.354632587859426,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8360
+    },
+    {
+      "epoch": 13.356230031948883,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 8361
+    },
+    {
+      "epoch": 13.357827476038338,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8362
+    },
+    {
+      "epoch": 13.359424920127795,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 8363
+    },
+    {
+      "epoch": 13.361022364217252,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 8364
+    },
+    {
+      "epoch": 13.36261980830671,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 8365
+    },
+    {
+      "epoch": 13.364217252396166,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8366
+    },
+    {
+      "epoch": 13.365814696485623,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 8367
+    },
+    {
+      "epoch": 13.36741214057508,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1297,
+      "step": 8368
+    },
+    {
+      "epoch": 13.369009584664537,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8369
+    },
+    {
+      "epoch": 13.370607028753994,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8370
+    },
+    {
+      "epoch": 13.37220447284345,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 8371
+    },
+    {
+      "epoch": 13.373801916932907,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8372
+    },
+    {
+      "epoch": 13.375399361022364,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8373
+    },
+    {
+      "epoch": 13.37699680511182,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 8374
+    },
+    {
+      "epoch": 13.378594249201278,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 8375
+    },
+    {
+      "epoch": 13.380191693290735,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8376
+    },
+    {
+      "epoch": 13.381789137380192,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.139,
+      "step": 8377
+    },
+    {
+      "epoch": 13.383386581469649,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8378
+    },
+    {
+      "epoch": 13.384984025559106,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 8379
+    },
+    {
+      "epoch": 13.386581469648561,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8380
+    },
+    {
+      "epoch": 13.388178913738018,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 8381
+    },
+    {
+      "epoch": 13.389776357827476,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 8382
+    },
+    {
+      "epoch": 13.391373801916933,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 8383
+    },
+    {
+      "epoch": 13.39297124600639,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 8384
+    },
+    {
+      "epoch": 13.394568690095847,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 8385
+    },
+    {
+      "epoch": 13.396166134185304,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1387,
+      "step": 8386
+    },
+    {
+      "epoch": 13.39776357827476,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 8387
+    },
+    {
+      "epoch": 13.399361022364218,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1262,
+      "step": 8388
+    },
+    {
+      "epoch": 13.400958466453675,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 8389
+    },
+    {
+      "epoch": 13.40255591054313,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 8390
+    },
+    {
+      "epoch": 13.404153354632587,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 8391
+    },
+    {
+      "epoch": 13.405750798722044,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1224,
+      "step": 8392
+    },
+    {
+      "epoch": 13.407348242811501,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 8393
+    },
+    {
+      "epoch": 13.408945686900958,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 8394
+    },
+    {
+      "epoch": 13.410543130990416,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8395
+    },
+    {
+      "epoch": 13.412140575079873,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8396
+    },
+    {
+      "epoch": 13.41373801916933,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1182,
+      "step": 8397
+    },
+    {
+      "epoch": 13.415335463258787,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8398
+    },
+    {
+      "epoch": 13.416932907348242,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 8399
+    },
+    {
+      "epoch": 13.418530351437699,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 8400
+    },
+    {
+      "epoch": 13.420127795527156,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8401
+    },
+    {
+      "epoch": 13.421725239616613,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8402
+    },
+    {
+      "epoch": 13.42332268370607,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 8403
+    },
+    {
+      "epoch": 13.424920127795527,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8404
+    },
+    {
+      "epoch": 13.426517571884984,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8405
+    },
+    {
+      "epoch": 13.428115015974441,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 8406
+    },
+    {
+      "epoch": 13.429712460063898,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 8407
+    },
+    {
+      "epoch": 13.431309904153355,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1218,
+      "step": 8408
+    },
+    {
+      "epoch": 13.43290734824281,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 8409
+    },
+    {
+      "epoch": 13.434504792332268,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8410
+    },
+    {
+      "epoch": 13.436102236421725,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 8411
+    },
+    {
+      "epoch": 13.437699680511182,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8412
+    },
+    {
+      "epoch": 13.439297124600639,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 8413
+    },
+    {
+      "epoch": 13.440894568690096,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8414
+    },
+    {
+      "epoch": 13.442492012779553,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 8415
+    },
+    {
+      "epoch": 13.44408945686901,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8416
+    },
+    {
+      "epoch": 13.445686900958467,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 8417
+    },
+    {
+      "epoch": 13.447284345047922,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 8418
+    },
+    {
+      "epoch": 13.44888178913738,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.123,
+      "step": 8419
+    },
+    {
+      "epoch": 13.450479233226837,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 8420
+    },
+    {
+      "epoch": 13.452076677316294,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8421
+    },
+    {
+      "epoch": 13.45367412140575,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 8422
+    },
+    {
+      "epoch": 13.455271565495208,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8423
+    },
+    {
+      "epoch": 13.456869009584665,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 8424
+    },
+    {
+      "epoch": 13.458466453674122,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 8425
+    },
+    {
+      "epoch": 13.460063897763579,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8426
+    },
+    {
+      "epoch": 13.461661341853036,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8427
+    },
+    {
+      "epoch": 13.463258785942491,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 8428
+    },
+    {
+      "epoch": 13.464856230031948,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8429
+    },
+    {
+      "epoch": 13.466453674121405,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8430
+    },
+    {
+      "epoch": 13.468051118210862,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8431
+    },
+    {
+      "epoch": 13.46964856230032,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1262,
+      "step": 8432
+    },
+    {
+      "epoch": 13.471246006389777,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 8433
+    },
+    {
+      "epoch": 13.472843450479234,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 8434
+    },
+    {
+      "epoch": 13.47444089456869,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8435
+    },
+    {
+      "epoch": 13.476038338658148,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1254,
+      "step": 8436
+    },
+    {
+      "epoch": 13.477635782747603,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 8437
+    },
+    {
+      "epoch": 13.47923322683706,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8438
+    },
+    {
+      "epoch": 13.480830670926517,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 8439
+    },
+    {
+      "epoch": 13.482428115015974,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 8440
+    },
+    {
+      "epoch": 13.484025559105431,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8441
+    },
+    {
+      "epoch": 13.485623003194888,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 8442
+    },
+    {
+      "epoch": 13.487220447284345,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8443
+    },
+    {
+      "epoch": 13.488817891373802,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8444
+    },
+    {
+      "epoch": 13.49041533546326,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 8445
+    },
+    {
+      "epoch": 13.492012779552716,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1403,
+      "step": 8446
+    },
+    {
+      "epoch": 13.493610223642172,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 8447
+    },
+    {
+      "epoch": 13.495207667731629,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 8448
+    },
+    {
+      "epoch": 13.496805111821086,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 8449
+    },
+    {
+      "epoch": 13.498402555910543,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 8450
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8451
+    },
+    {
+      "epoch": 13.501597444089457,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8452
+    },
+    {
+      "epoch": 13.503194888178914,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8453
+    },
+    {
+      "epoch": 13.504792332268371,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 8454
+    },
+    {
+      "epoch": 13.506389776357828,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1219,
+      "step": 8455
+    },
+    {
+      "epoch": 13.507987220447284,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 8456
+    },
+    {
+      "epoch": 13.50958466453674,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1204,
+      "step": 8457
+    },
+    {
+      "epoch": 13.511182108626198,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8458
+    },
+    {
+      "epoch": 13.512779552715655,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8459
+    },
+    {
+      "epoch": 13.514376996805112,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 8460
+    },
+    {
+      "epoch": 13.515974440894569,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8461
+    },
+    {
+      "epoch": 13.517571884984026,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 8462
+    },
+    {
+      "epoch": 13.519169329073483,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8463
+    },
+    {
+      "epoch": 13.52076677316294,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8464
+    },
+    {
+      "epoch": 13.522364217252395,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8465
+    },
+    {
+      "epoch": 13.523961661341852,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8466
+    },
+    {
+      "epoch": 13.52555910543131,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 8467
+    },
+    {
+      "epoch": 13.527156549520766,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1217,
+      "step": 8468
+    },
+    {
+      "epoch": 13.528753993610223,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1225,
+      "step": 8469
+    },
+    {
+      "epoch": 13.53035143769968,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 8470
+    },
+    {
+      "epoch": 13.531948881789138,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8471
+    },
+    {
+      "epoch": 13.533546325878595,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8472
+    },
+    {
+      "epoch": 13.535143769968052,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.119,
+      "step": 8473
+    },
+    {
+      "epoch": 13.536741214057509,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8474
+    },
+    {
+      "epoch": 13.538338658146966,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 8475
+    },
+    {
+      "epoch": 13.539936102236421,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 8476
+    },
+    {
+      "epoch": 13.541533546325878,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 8477
+    },
+    {
+      "epoch": 13.543130990415335,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 8478
+    },
+    {
+      "epoch": 13.544728434504792,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 8479
+    },
+    {
+      "epoch": 13.54632587859425,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8480
+    },
+    {
+      "epoch": 13.547923322683706,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 8481
+    },
+    {
+      "epoch": 13.549520766773163,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 8482
+    },
+    {
+      "epoch": 13.55111821086262,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8483
+    },
+    {
+      "epoch": 13.552715654952078,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1192,
+      "step": 8484
+    },
+    {
+      "epoch": 13.554313099041533,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8485
+    },
+    {
+      "epoch": 13.55591054313099,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8486
+    },
+    {
+      "epoch": 13.557507987220447,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8487
+    },
+    {
+      "epoch": 13.559105431309904,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8488
+    },
+    {
+      "epoch": 13.560702875399361,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 8489
+    },
+    {
+      "epoch": 13.562300319488818,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1207,
+      "step": 8490
+    },
+    {
+      "epoch": 13.563897763578275,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8491
+    },
+    {
+      "epoch": 13.565495207667732,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 8492
+    },
+    {
+      "epoch": 13.56709265175719,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8493
+    },
+    {
+      "epoch": 13.568690095846645,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 8494
+    },
+    {
+      "epoch": 13.570287539936102,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8495
+    },
+    {
+      "epoch": 13.571884984025559,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8496
+    },
+    {
+      "epoch": 13.573482428115016,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1151,
+      "step": 8497
+    },
+    {
+      "epoch": 13.575079872204473,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 8498
+    },
+    {
+      "epoch": 13.57667731629393,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8499
+    },
+    {
+      "epoch": 13.578274760383387,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 8500
+    },
+    {
+      "epoch": 13.579872204472844,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8501
+    },
+    {
+      "epoch": 13.581469648562301,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 8502
+    },
+    {
+      "epoch": 13.583067092651756,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 8503
+    },
+    {
+      "epoch": 13.584664536741213,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 8504
+    },
+    {
+      "epoch": 13.58626198083067,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8505
+    },
+    {
+      "epoch": 13.587859424920127,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 8506
+    },
+    {
+      "epoch": 13.589456869009584,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8507
+    },
+    {
+      "epoch": 13.591054313099042,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8508
+    },
+    {
+      "epoch": 13.592651757188499,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8509
+    },
+    {
+      "epoch": 13.594249201277956,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 8510
+    },
+    {
+      "epoch": 13.595846645367413,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1314,
+      "step": 8511
+    },
+    {
+      "epoch": 13.59744408945687,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 8512
+    },
+    {
+      "epoch": 13.599041533546325,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 8513
+    },
+    {
+      "epoch": 13.600638977635782,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 8514
+    },
+    {
+      "epoch": 13.60223642172524,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 8515
+    },
+    {
+      "epoch": 13.603833865814696,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8516
+    },
+    {
+      "epoch": 13.605431309904153,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8517
+    },
+    {
+      "epoch": 13.60702875399361,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 8518
+    },
+    {
+      "epoch": 13.608626198083067,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 8519
+    },
+    {
+      "epoch": 13.610223642172524,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1228,
+      "step": 8520
+    },
+    {
+      "epoch": 13.611821086261982,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8521
+    },
+    {
+      "epoch": 13.613418530351439,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8522
+    },
+    {
+      "epoch": 13.615015974440894,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8523
+    },
+    {
+      "epoch": 13.616613418530351,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 8524
+    },
+    {
+      "epoch": 13.618210862619808,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1239,
+      "step": 8525
+    },
+    {
+      "epoch": 13.619808306709265,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 8526
+    },
+    {
+      "epoch": 13.621405750798722,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8527
+    },
+    {
+      "epoch": 13.62300319488818,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8528
+    },
+    {
+      "epoch": 13.624600638977636,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 8529
+    },
+    {
+      "epoch": 13.626198083067093,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 8530
+    },
+    {
+      "epoch": 13.62779552715655,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8531
+    },
+    {
+      "epoch": 13.629392971246006,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 8532
+    },
+    {
+      "epoch": 13.630990415335463,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8533
+    },
+    {
+      "epoch": 13.63258785942492,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8534
+    },
+    {
+      "epoch": 13.634185303514377,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8535
+    },
+    {
+      "epoch": 13.635782747603834,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8536
+    },
+    {
+      "epoch": 13.63738019169329,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8537
+    },
+    {
+      "epoch": 13.638977635782748,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 8538
+    },
+    {
+      "epoch": 13.640575079872205,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1162,
+      "step": 8539
+    },
+    {
+      "epoch": 13.642172523961662,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8540
+    },
+    {
+      "epoch": 13.643769968051117,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 8541
+    },
+    {
+      "epoch": 13.645367412140574,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 8542
+    },
+    {
+      "epoch": 13.646964856230031,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 8543
+    },
+    {
+      "epoch": 13.648562300319488,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 8544
+    },
+    {
+      "epoch": 13.650159744408946,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8545
+    },
+    {
+      "epoch": 13.651757188498403,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 8546
+    },
+    {
+      "epoch": 13.65335463258786,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 8547
+    },
+    {
+      "epoch": 13.654952076677317,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8548
+    },
+    {
+      "epoch": 13.656549520766774,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 8549
+    },
+    {
+      "epoch": 13.65814696485623,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8550
+    },
+    {
+      "epoch": 13.659744408945686,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 8551
+    },
+    {
+      "epoch": 13.661341853035143,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 8552
+    },
+    {
+      "epoch": 13.6629392971246,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 8553
+    },
+    {
+      "epoch": 13.664536741214057,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8554
+    },
+    {
+      "epoch": 13.666134185303514,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1216,
+      "step": 8555
+    },
+    {
+      "epoch": 13.667731629392971,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8556
+    },
+    {
+      "epoch": 13.669329073482428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 8557
+    },
+    {
+      "epoch": 13.670926517571885,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 8558
+    },
+    {
+      "epoch": 13.672523961661343,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8559
+    },
+    {
+      "epoch": 13.6741214057508,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8560
+    },
+    {
+      "epoch": 13.675718849840255,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8561
+    },
+    {
+      "epoch": 13.677316293929712,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1172,
+      "step": 8562
+    },
+    {
+      "epoch": 13.678913738019169,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 8563
+    },
+    {
+      "epoch": 13.680511182108626,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 8564
+    },
+    {
+      "epoch": 13.682108626198083,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 8565
+    },
+    {
+      "epoch": 13.68370607028754,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 8566
+    },
+    {
+      "epoch": 13.685303514376997,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 8567
+    },
+    {
+      "epoch": 13.686900958466454,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.118,
+      "step": 8568
+    },
+    {
+      "epoch": 13.688498402555911,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8569
+    },
+    {
+      "epoch": 13.690095846645367,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 8570
+    },
+    {
+      "epoch": 13.691693290734824,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8571
+    },
+    {
+      "epoch": 13.69329073482428,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8572
+    },
+    {
+      "epoch": 13.694888178913738,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8573
+    },
+    {
+      "epoch": 13.696485623003195,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 8574
+    },
+    {
+      "epoch": 13.698083067092652,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 8575
+    },
+    {
+      "epoch": 13.699680511182109,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 8576
+    },
+    {
+      "epoch": 13.701277955271566,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8577
+    },
+    {
+      "epoch": 13.702875399361023,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8578
+    },
+    {
+      "epoch": 13.704472843450478,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 8579
+    },
+    {
+      "epoch": 13.706070287539935,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 8580
+    },
+    {
+      "epoch": 13.707667731629392,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 8581
+    },
+    {
+      "epoch": 13.70926517571885,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 8582
+    },
+    {
+      "epoch": 13.710862619808307,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1228,
+      "step": 8583
+    },
+    {
+      "epoch": 13.712460063897764,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 8584
+    },
+    {
+      "epoch": 13.71405750798722,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8585
+    },
+    {
+      "epoch": 13.715654952076678,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1138,
+      "step": 8586
+    },
+    {
+      "epoch": 13.717252396166135,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8587
+    },
+    {
+      "epoch": 13.718849840255592,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 8588
+    },
+    {
+      "epoch": 13.720447284345047,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 8589
+    },
+    {
+      "epoch": 13.722044728434504,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8590
+    },
+    {
+      "epoch": 13.723642172523961,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 8591
+    },
+    {
+      "epoch": 13.725239616613418,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8592
+    },
+    {
+      "epoch": 13.726837060702875,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8593
+    },
+    {
+      "epoch": 13.728434504792332,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8594
+    },
+    {
+      "epoch": 13.73003194888179,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1237,
+      "step": 8595
+    },
+    {
+      "epoch": 13.731629392971247,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8596
+    },
+    {
+      "epoch": 13.733226837060704,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 8597
+    },
+    {
+      "epoch": 13.73482428115016,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8598
+    },
+    {
+      "epoch": 13.736421725239616,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 8599
+    },
+    {
+      "epoch": 13.738019169329073,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 8600
+    },
+    {
+      "epoch": 13.73961661341853,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8601
+    },
+    {
+      "epoch": 13.741214057507987,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8602
+    },
+    {
+      "epoch": 13.742811501597444,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8603
+    },
+    {
+      "epoch": 13.744408945686901,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8604
+    },
+    {
+      "epoch": 13.746006389776358,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8605
+    },
+    {
+      "epoch": 13.747603833865815,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 8606
+    },
+    {
+      "epoch": 13.749201277955272,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 8607
+    },
+    {
+      "epoch": 13.750798722044728,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 8608
+    },
+    {
+      "epoch": 13.752396166134185,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 8609
+    },
+    {
+      "epoch": 13.753993610223642,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 8610
+    },
+    {
+      "epoch": 13.755591054313099,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 8611
+    },
+    {
+      "epoch": 13.757188498402556,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 8612
+    },
+    {
+      "epoch": 13.758785942492013,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8613
+    },
+    {
+      "epoch": 13.76038338658147,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8614
+    },
+    {
+      "epoch": 13.761980830670927,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1168,
+      "step": 8615
+    },
+    {
+      "epoch": 13.763578274760384,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8616
+    },
+    {
+      "epoch": 13.76517571884984,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 8617
+    },
+    {
+      "epoch": 13.766773162939296,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8618
+    },
+    {
+      "epoch": 13.768370607028753,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8619
+    },
+    {
+      "epoch": 13.76996805111821,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8620
+    },
+    {
+      "epoch": 13.771565495207668,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8621
+    },
+    {
+      "epoch": 13.773162939297125,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8622
+    },
+    {
+      "epoch": 13.774760383386582,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8623
+    },
+    {
+      "epoch": 13.776357827476039,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8624
+    },
+    {
+      "epoch": 13.777955271565496,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8625
+    },
+    {
+      "epoch": 13.779552715654953,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8626
+    },
+    {
+      "epoch": 13.781150159744408,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 8627
+    },
+    {
+      "epoch": 13.782747603833865,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8628
+    },
+    {
+      "epoch": 13.784345047923322,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.123,
+      "step": 8629
+    },
+    {
+      "epoch": 13.78594249201278,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1197,
+      "step": 8630
+    },
+    {
+      "epoch": 13.787539936102236,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 8631
+    },
+    {
+      "epoch": 13.789137380191693,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 8632
+    },
+    {
+      "epoch": 13.79073482428115,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8633
+    },
+    {
+      "epoch": 13.792332268370608,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 8634
+    },
+    {
+      "epoch": 13.793929712460065,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8635
+    },
+    {
+      "epoch": 13.795527156549522,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8636
+    },
+    {
+      "epoch": 13.797124600638977,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8637
+    },
+    {
+      "epoch": 13.798722044728434,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8638
+    },
+    {
+      "epoch": 13.800319488817891,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1179,
+      "step": 8639
+    },
+    {
+      "epoch": 13.801916932907348,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 8640
+    },
+    {
+      "epoch": 13.803514376996805,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8641
+    },
+    {
+      "epoch": 13.805111821086262,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8642
+    },
+    {
+      "epoch": 13.80670926517572,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1276,
+      "step": 8643
+    },
+    {
+      "epoch": 13.808306709265176,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1205,
+      "step": 8644
+    },
+    {
+      "epoch": 13.809904153354633,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 8645
+    },
+    {
+      "epoch": 13.811501597444089,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 8646
+    },
+    {
+      "epoch": 13.813099041533546,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1198,
+      "step": 8647
+    },
+    {
+      "epoch": 13.814696485623003,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8648
+    },
+    {
+      "epoch": 13.81629392971246,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 8649
+    },
+    {
+      "epoch": 13.817891373801917,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 8650
+    },
+    {
+      "epoch": 13.819488817891374,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 8651
+    },
+    {
+      "epoch": 13.821086261980831,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 8652
+    },
+    {
+      "epoch": 13.822683706070288,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8653
+    },
+    {
+      "epoch": 13.824281150159745,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1376,
+      "step": 8654
+    },
+    {
+      "epoch": 13.8258785942492,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 8655
+    },
+    {
+      "epoch": 13.827476038338657,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 8656
+    },
+    {
+      "epoch": 13.829073482428115,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 8657
+    },
+    {
+      "epoch": 13.830670926517572,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8658
+    },
+    {
+      "epoch": 13.832268370607029,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8659
+    },
+    {
+      "epoch": 13.833865814696486,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 8660
+    },
+    {
+      "epoch": 13.835463258785943,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8661
+    },
+    {
+      "epoch": 13.8370607028754,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1378,
+      "step": 8662
+    },
+    {
+      "epoch": 13.838658146964857,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 8663
+    },
+    {
+      "epoch": 13.840255591054314,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 8664
+    },
+    {
+      "epoch": 13.84185303514377,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8665
+    },
+    {
+      "epoch": 13.843450479233226,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8666
+    },
+    {
+      "epoch": 13.845047923322683,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8667
+    },
+    {
+      "epoch": 13.84664536741214,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8668
+    },
+    {
+      "epoch": 13.848242811501597,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8669
+    },
+    {
+      "epoch": 13.849840255591054,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 8670
+    },
+    {
+      "epoch": 13.851437699680512,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8671
+    },
+    {
+      "epoch": 13.853035143769969,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8672
+    },
+    {
+      "epoch": 13.854632587859426,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8673
+    },
+    {
+      "epoch": 13.856230031948883,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.116,
+      "step": 8674
+    },
+    {
+      "epoch": 13.857827476038338,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.123,
+      "step": 8675
+    },
+    {
+      "epoch": 13.859424920127795,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8676
+    },
+    {
+      "epoch": 13.861022364217252,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 8677
+    },
+    {
+      "epoch": 13.86261980830671,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1195,
+      "step": 8678
+    },
+    {
+      "epoch": 13.864217252396166,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8679
+    },
+    {
+      "epoch": 13.865814696485623,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 8680
+    },
+    {
+      "epoch": 13.86741214057508,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 8681
+    },
+    {
+      "epoch": 13.869009584664537,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8682
+    },
+    {
+      "epoch": 13.870607028753994,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1256,
+      "step": 8683
+    },
+    {
+      "epoch": 13.87220447284345,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8684
+    },
+    {
+      "epoch": 13.873801916932907,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8685
+    },
+    {
+      "epoch": 13.875399361022364,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8686
+    },
+    {
+      "epoch": 13.87699680511182,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8687
+    },
+    {
+      "epoch": 13.878594249201278,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8688
+    },
+    {
+      "epoch": 13.880191693290735,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 8689
+    },
+    {
+      "epoch": 13.881789137380192,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8690
+    },
+    {
+      "epoch": 13.883386581469649,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8691
+    },
+    {
+      "epoch": 13.884984025559106,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8692
+    },
+    {
+      "epoch": 13.886581469648561,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 8693
+    },
+    {
+      "epoch": 13.888178913738018,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1223,
+      "step": 8694
+    },
+    {
+      "epoch": 13.889776357827476,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8695
+    },
+    {
+      "epoch": 13.891373801916933,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 8696
+    },
+    {
+      "epoch": 13.89297124600639,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1388,
+      "step": 8697
+    },
+    {
+      "epoch": 13.894568690095847,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1211,
+      "step": 8698
+    },
+    {
+      "epoch": 13.896166134185304,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 8699
+    },
+    {
+      "epoch": 13.89776357827476,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8700
+    },
+    {
+      "epoch": 13.899361022364218,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8701
+    },
+    {
+      "epoch": 13.900958466453675,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 8702
+    },
+    {
+      "epoch": 13.90255591054313,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1236,
+      "step": 8703
+    },
+    {
+      "epoch": 13.904153354632587,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 8704
+    },
+    {
+      "epoch": 13.905750798722044,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1397,
+      "step": 8705
+    },
+    {
+      "epoch": 13.907348242811501,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 8706
+    },
+    {
+      "epoch": 13.908945686900958,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 8707
+    },
+    {
+      "epoch": 13.910543130990416,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8708
+    },
+    {
+      "epoch": 13.912140575079873,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8709
+    },
+    {
+      "epoch": 13.91373801916933,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8710
+    },
+    {
+      "epoch": 13.915335463258787,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 8711
+    },
+    {
+      "epoch": 13.916932907348244,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 8712
+    },
+    {
+      "epoch": 13.918530351437699,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 8713
+    },
+    {
+      "epoch": 13.920127795527156,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8714
+    },
+    {
+      "epoch": 13.921725239616613,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 8715
+    },
+    {
+      "epoch": 13.92332268370607,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 8716
+    },
+    {
+      "epoch": 13.924920127795527,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 8717
+    },
+    {
+      "epoch": 13.926517571884984,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8718
+    },
+    {
+      "epoch": 13.928115015974441,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8719
+    },
+    {
+      "epoch": 13.929712460063898,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8720
+    },
+    {
+      "epoch": 13.931309904153355,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8721
+    },
+    {
+      "epoch": 13.93290734824281,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 8722
+    },
+    {
+      "epoch": 13.934504792332268,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 8723
+    },
+    {
+      "epoch": 13.936102236421725,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8724
+    },
+    {
+      "epoch": 13.937699680511182,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 8725
+    },
+    {
+      "epoch": 13.939297124600639,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 8726
+    },
+    {
+      "epoch": 13.940894568690096,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8727
+    },
+    {
+      "epoch": 13.942492012779553,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8728
+    },
+    {
+      "epoch": 13.94408945686901,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1357,
+      "step": 8729
+    },
+    {
+      "epoch": 13.945686900958467,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 8730
+    },
+    {
+      "epoch": 13.947284345047922,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8731
+    },
+    {
+      "epoch": 13.94888178913738,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8732
+    },
+    {
+      "epoch": 13.950479233226837,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 8733
+    },
+    {
+      "epoch": 13.952076677316294,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1219,
+      "step": 8734
+    },
+    {
+      "epoch": 13.95367412140575,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8735
+    },
+    {
+      "epoch": 13.955271565495208,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 8736
+    },
+    {
+      "epoch": 13.956869009584665,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 8737
+    },
+    {
+      "epoch": 13.958466453674122,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8738
+    },
+    {
+      "epoch": 13.960063897763579,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 8739
+    },
+    {
+      "epoch": 13.961661341853034,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 8740
+    },
+    {
+      "epoch": 13.963258785942491,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1228,
+      "step": 8741
+    },
+    {
+      "epoch": 13.964856230031948,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 8742
+    },
+    {
+      "epoch": 13.966453674121405,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 8743
+    },
+    {
+      "epoch": 13.968051118210862,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8744
+    },
+    {
+      "epoch": 13.96964856230032,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 8745
+    },
+    {
+      "epoch": 13.971246006389777,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8746
+    },
+    {
+      "epoch": 13.972843450479234,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 8747
+    },
+    {
+      "epoch": 13.97444089456869,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 8748
+    },
+    {
+      "epoch": 13.976038338658148,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8749
+    },
+    {
+      "epoch": 13.977635782747605,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8750
+    },
+    {
+      "epoch": 13.97923322683706,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8751
+    },
+    {
+      "epoch": 13.980830670926517,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8752
+    },
+    {
+      "epoch": 13.982428115015974,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 8753
+    },
+    {
+      "epoch": 13.984025559105431,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 8754
+    },
+    {
+      "epoch": 13.985623003194888,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8755
+    },
+    {
+      "epoch": 13.987220447284345,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8756
+    },
+    {
+      "epoch": 13.988817891373802,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 8757
+    },
+    {
+      "epoch": 13.99041533546326,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8758
+    },
+    {
+      "epoch": 13.992012779552716,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 8759
+    },
+    {
+      "epoch": 13.993610223642172,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1302,
+      "step": 8760
+    },
+    {
+      "epoch": 13.995207667731629,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 8761
+    },
+    {
+      "epoch": 13.996805111821086,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8762
+    },
+    {
+      "epoch": 13.998402555910543,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1189,
+      "step": 8763
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8764
+    },
+    {
+      "epoch": 14.001597444089457,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 8765
+    },
+    {
+      "epoch": 14.003194888178914,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.129,
+      "step": 8766
+    },
+    {
+      "epoch": 14.004792332268371,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 8767
+    },
+    {
+      "epoch": 14.006389776357828,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 8768
+    },
+    {
+      "epoch": 14.007987220447284,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8769
+    },
+    {
+      "epoch": 14.00958466453674,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8770
+    },
+    {
+      "epoch": 14.011182108626198,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8771
+    },
+    {
+      "epoch": 14.012779552715655,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 8772
+    },
+    {
+      "epoch": 14.014376996805112,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8773
+    },
+    {
+      "epoch": 14.015974440894569,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8774
+    },
+    {
+      "epoch": 14.017571884984026,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 8775
+    },
+    {
+      "epoch": 14.019169329073483,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8776
+    },
+    {
+      "epoch": 14.02076677316294,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 8777
+    },
+    {
+      "epoch": 14.022364217252397,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 8778
+    },
+    {
+      "epoch": 14.023961661341852,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1176,
+      "step": 8779
+    },
+    {
+      "epoch": 14.02555910543131,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8780
+    },
+    {
+      "epoch": 14.027156549520766,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1286,
+      "step": 8781
+    },
+    {
+      "epoch": 14.028753993610223,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8782
+    },
+    {
+      "epoch": 14.03035143769968,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8783
+    },
+    {
+      "epoch": 14.031948881789138,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 8784
+    },
+    {
+      "epoch": 14.033546325878595,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 8785
+    },
+    {
+      "epoch": 14.035143769968052,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 8786
+    },
+    {
+      "epoch": 14.036741214057509,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1415,
+      "step": 8787
+    },
+    {
+      "epoch": 14.038338658146964,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 8788
+    },
+    {
+      "epoch": 14.039936102236421,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8789
+    },
+    {
+      "epoch": 14.041533546325878,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8790
+    },
+    {
+      "epoch": 14.043130990415335,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8791
+    },
+    {
+      "epoch": 14.044728434504792,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8792
+    },
+    {
+      "epoch": 14.04632587859425,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 8793
+    },
+    {
+      "epoch": 14.047923322683706,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1381,
+      "step": 8794
+    },
+    {
+      "epoch": 14.049520766773163,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1257,
+      "step": 8795
+    },
+    {
+      "epoch": 14.05111821086262,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8796
+    },
+    {
+      "epoch": 14.052715654952078,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8797
+    },
+    {
+      "epoch": 14.054313099041533,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 8798
+    },
+    {
+      "epoch": 14.05591054313099,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8799
+    },
+    {
+      "epoch": 14.057507987220447,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8800
+    },
+    {
+      "epoch": 14.059105431309904,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8801
+    },
+    {
+      "epoch": 14.060702875399361,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 8802
+    },
+    {
+      "epoch": 14.062300319488818,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1208,
+      "step": 8803
+    },
+    {
+      "epoch": 14.063897763578275,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1242,
+      "step": 8804
+    },
+    {
+      "epoch": 14.065495207667732,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 8805
+    },
+    {
+      "epoch": 14.06709265175719,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 8806
+    },
+    {
+      "epoch": 14.068690095846645,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8807
+    },
+    {
+      "epoch": 14.070287539936102,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 8808
+    },
+    {
+      "epoch": 14.071884984025559,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 8809
+    },
+    {
+      "epoch": 14.073482428115016,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1411,
+      "step": 8810
+    },
+    {
+      "epoch": 14.075079872204473,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8811
+    },
+    {
+      "epoch": 14.07667731629393,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 8812
+    },
+    {
+      "epoch": 14.078274760383387,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1382,
+      "step": 8813
+    },
+    {
+      "epoch": 14.079872204472844,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1578,
+      "step": 8814
+    },
+    {
+      "epoch": 14.081469648562301,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1352,
+      "step": 8815
+    },
+    {
+      "epoch": 14.083067092651758,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8816
+    },
+    {
+      "epoch": 14.084664536741213,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1428,
+      "step": 8817
+    },
+    {
+      "epoch": 14.08626198083067,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8818
+    },
+    {
+      "epoch": 14.087859424920127,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 8819
+    },
+    {
+      "epoch": 14.089456869009584,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8820
+    },
+    {
+      "epoch": 14.091054313099042,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 8821
+    },
+    {
+      "epoch": 14.092651757188499,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8822
+    },
+    {
+      "epoch": 14.094249201277956,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 8823
+    },
+    {
+      "epoch": 14.095846645367413,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8824
+    },
+    {
+      "epoch": 14.09744408945687,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 8825
+    },
+    {
+      "epoch": 14.099041533546325,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1367,
+      "step": 8826
+    },
+    {
+      "epoch": 14.100638977635782,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 8827
+    },
+    {
+      "epoch": 14.10223642172524,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 8828
+    },
+    {
+      "epoch": 14.103833865814696,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8829
+    },
+    {
+      "epoch": 14.105431309904153,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1419,
+      "step": 8830
+    },
+    {
+      "epoch": 14.10702875399361,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 8831
+    },
+    {
+      "epoch": 14.108626198083067,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 8832
+    },
+    {
+      "epoch": 14.110223642172524,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8833
+    },
+    {
+      "epoch": 14.111821086261982,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8834
+    },
+    {
+      "epoch": 14.113418530351439,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 8835
+    },
+    {
+      "epoch": 14.115015974440894,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 8836
+    },
+    {
+      "epoch": 14.116613418530351,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8837
+    },
+    {
+      "epoch": 14.118210862619808,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 8838
+    },
+    {
+      "epoch": 14.119808306709265,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8839
+    },
+    {
+      "epoch": 14.121405750798722,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1365,
+      "step": 8840
+    },
+    {
+      "epoch": 14.12300319488818,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 8841
+    },
+    {
+      "epoch": 14.124600638977636,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8842
+    },
+    {
+      "epoch": 14.126198083067093,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 8843
+    },
+    {
+      "epoch": 14.12779552715655,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 8844
+    },
+    {
+      "epoch": 14.129392971246006,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8845
+    },
+    {
+      "epoch": 14.130990415335463,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8846
+    },
+    {
+      "epoch": 14.13258785942492,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 8847
+    },
+    {
+      "epoch": 14.134185303514377,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8848
+    },
+    {
+      "epoch": 14.135782747603834,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1303,
+      "step": 8849
+    },
+    {
+      "epoch": 14.13738019169329,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 8850
+    },
+    {
+      "epoch": 14.138977635782748,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8851
+    },
+    {
+      "epoch": 14.140575079872205,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 8852
+    },
+    {
+      "epoch": 14.142172523961662,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1208,
+      "step": 8853
+    },
+    {
+      "epoch": 14.143769968051119,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 8854
+    },
+    {
+      "epoch": 14.145367412140574,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8855
+    },
+    {
+      "epoch": 14.146964856230031,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8856
+    },
+    {
+      "epoch": 14.148562300319488,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8857
+    },
+    {
+      "epoch": 14.150159744408946,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8858
+    },
+    {
+      "epoch": 14.151757188498403,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8859
+    },
+    {
+      "epoch": 14.15335463258786,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 8860
+    },
+    {
+      "epoch": 14.154952076677317,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.122,
+      "step": 8861
+    },
+    {
+      "epoch": 14.156549520766774,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1224,
+      "step": 8862
+    },
+    {
+      "epoch": 14.15814696485623,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 8863
+    },
+    {
+      "epoch": 14.159744408945686,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8864
+    },
+    {
+      "epoch": 14.161341853035143,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 8865
+    },
+    {
+      "epoch": 14.1629392971246,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 8866
+    },
+    {
+      "epoch": 14.164536741214057,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8867
+    },
+    {
+      "epoch": 14.166134185303514,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 8868
+    },
+    {
+      "epoch": 14.167731629392971,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 8869
+    },
+    {
+      "epoch": 14.169329073482428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1237,
+      "step": 8870
+    },
+    {
+      "epoch": 14.170926517571885,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 8871
+    },
+    {
+      "epoch": 14.172523961661343,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 8872
+    },
+    {
+      "epoch": 14.1741214057508,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8873
+    },
+    {
+      "epoch": 14.175718849840255,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 8874
+    },
+    {
+      "epoch": 14.177316293929712,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.137,
+      "step": 8875
+    },
+    {
+      "epoch": 14.178913738019169,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8876
+    },
+    {
+      "epoch": 14.180511182108626,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 8877
+    },
+    {
+      "epoch": 14.182108626198083,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8878
+    },
+    {
+      "epoch": 14.18370607028754,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8879
+    },
+    {
+      "epoch": 14.185303514376997,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 8880
+    },
+    {
+      "epoch": 14.186900958466454,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1285,
+      "step": 8881
+    },
+    {
+      "epoch": 14.188498402555911,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8882
+    },
+    {
+      "epoch": 14.190095846645367,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1274,
+      "step": 8883
+    },
+    {
+      "epoch": 14.191693290734824,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 8884
+    },
+    {
+      "epoch": 14.19329073482428,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8885
+    },
+    {
+      "epoch": 14.194888178913738,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 8886
+    },
+    {
+      "epoch": 14.196485623003195,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 8887
+    },
+    {
+      "epoch": 14.198083067092652,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8888
+    },
+    {
+      "epoch": 14.199680511182109,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 8889
+    },
+    {
+      "epoch": 14.201277955271566,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1205,
+      "step": 8890
+    },
+    {
+      "epoch": 14.202875399361023,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 8891
+    },
+    {
+      "epoch": 14.204472843450478,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 8892
+    },
+    {
+      "epoch": 14.206070287539935,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 8893
+    },
+    {
+      "epoch": 14.207667731629392,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 8894
+    },
+    {
+      "epoch": 14.20926517571885,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 8895
+    },
+    {
+      "epoch": 14.210862619808307,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 8896
+    },
+    {
+      "epoch": 14.212460063897764,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8897
+    },
+    {
+      "epoch": 14.21405750798722,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 8898
+    },
+    {
+      "epoch": 14.215654952076678,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 8899
+    },
+    {
+      "epoch": 14.217252396166135,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8900
+    },
+    {
+      "epoch": 14.218849840255592,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8901
+    },
+    {
+      "epoch": 14.220447284345047,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 8902
+    },
+    {
+      "epoch": 14.222044728434504,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 8903
+    },
+    {
+      "epoch": 14.223642172523961,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 8904
+    },
+    {
+      "epoch": 14.225239616613418,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8905
+    },
+    {
+      "epoch": 14.226837060702875,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 8906
+    },
+    {
+      "epoch": 14.228434504792332,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 8907
+    },
+    {
+      "epoch": 14.23003194888179,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8908
+    },
+    {
+      "epoch": 14.231629392971247,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 8909
+    },
+    {
+      "epoch": 14.233226837060704,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8910
+    },
+    {
+      "epoch": 14.23482428115016,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1263,
+      "step": 8911
+    },
+    {
+      "epoch": 14.236421725239616,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 8912
+    },
+    {
+      "epoch": 14.238019169329073,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 8913
+    },
+    {
+      "epoch": 14.23961661341853,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 8914
+    },
+    {
+      "epoch": 14.241214057507987,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1219,
+      "step": 8915
+    },
+    {
+      "epoch": 14.242811501597444,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1321,
+      "step": 8916
+    },
+    {
+      "epoch": 14.244408945686901,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 8917
+    },
+    {
+      "epoch": 14.246006389776358,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 8918
+    },
+    {
+      "epoch": 14.247603833865815,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 8919
+    },
+    {
+      "epoch": 14.249201277955272,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1323,
+      "step": 8920
+    },
+    {
+      "epoch": 14.250798722044728,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 8921
+    },
+    {
+      "epoch": 14.252396166134185,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 8922
+    },
+    {
+      "epoch": 14.253993610223642,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 8923
+    },
+    {
+      "epoch": 14.255591054313099,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 8924
+    },
+    {
+      "epoch": 14.257188498402556,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 8925
+    },
+    {
+      "epoch": 14.258785942492013,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 8926
+    },
+    {
+      "epoch": 14.26038338658147,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1191,
+      "step": 8927
+    },
+    {
+      "epoch": 14.261980830670927,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8928
+    },
+    {
+      "epoch": 14.263578274760384,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 8929
+    },
+    {
+      "epoch": 14.26517571884984,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1319,
+      "step": 8930
+    },
+    {
+      "epoch": 14.266773162939296,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8931
+    },
+    {
+      "epoch": 14.268370607028753,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 8932
+    },
+    {
+      "epoch": 14.26996805111821,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8933
+    },
+    {
+      "epoch": 14.271565495207668,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1396,
+      "step": 8934
+    },
+    {
+      "epoch": 14.273162939297125,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8935
+    },
+    {
+      "epoch": 14.274760383386582,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 8936
+    },
+    {
+      "epoch": 14.276357827476039,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 8937
+    },
+    {
+      "epoch": 14.277955271565496,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 8938
+    },
+    {
+      "epoch": 14.279552715654953,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.116,
+      "step": 8939
+    },
+    {
+      "epoch": 14.281150159744408,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1344,
+      "step": 8940
+    },
+    {
+      "epoch": 14.282747603833865,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1322,
+      "step": 8941
+    },
+    {
+      "epoch": 14.284345047923322,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1427,
+      "step": 8942
+    },
+    {
+      "epoch": 14.28594249201278,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 8943
+    },
+    {
+      "epoch": 14.287539936102236,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8944
+    },
+    {
+      "epoch": 14.289137380191693,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1342,
+      "step": 8945
+    },
+    {
+      "epoch": 14.29073482428115,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1179,
+      "step": 8946
+    },
+    {
+      "epoch": 14.292332268370608,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 8947
+    },
+    {
+      "epoch": 14.293929712460065,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 8948
+    },
+    {
+      "epoch": 14.295527156549522,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 8949
+    },
+    {
+      "epoch": 14.297124600638977,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1308,
+      "step": 8950
+    },
+    {
+      "epoch": 14.298722044728434,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 8951
+    },
+    {
+      "epoch": 14.300319488817891,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8952
+    },
+    {
+      "epoch": 14.301916932907348,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 8953
+    },
+    {
+      "epoch": 14.303514376996805,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1291,
+      "step": 8954
+    },
+    {
+      "epoch": 14.305111821086262,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 8955
+    },
+    {
+      "epoch": 14.30670926517572,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 8956
+    },
+    {
+      "epoch": 14.308306709265176,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1197,
+      "step": 8957
+    },
+    {
+      "epoch": 14.309904153354633,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 8958
+    },
+    {
+      "epoch": 14.311501597444089,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 8959
+    },
+    {
+      "epoch": 14.313099041533546,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 8960
+    },
+    {
+      "epoch": 14.314696485623003,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1407,
+      "step": 8961
+    },
+    {
+      "epoch": 14.31629392971246,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1304,
+      "step": 8962
+    },
+    {
+      "epoch": 14.317891373801917,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 8963
+    },
+    {
+      "epoch": 14.319488817891374,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1226,
+      "step": 8964
+    },
+    {
+      "epoch": 14.321086261980831,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8965
+    },
+    {
+      "epoch": 14.322683706070288,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8966
+    },
+    {
+      "epoch": 14.324281150159745,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1149,
+      "step": 8967
+    },
+    {
+      "epoch": 14.3258785942492,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 8968
+    },
+    {
+      "epoch": 14.327476038338657,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1214,
+      "step": 8969
+    },
+    {
+      "epoch": 14.329073482428115,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1374,
+      "step": 8970
+    },
+    {
+      "epoch": 14.330670926517572,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 8971
+    },
+    {
+      "epoch": 14.332268370607029,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 8972
+    },
+    {
+      "epoch": 14.333865814696486,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8973
+    },
+    {
+      "epoch": 14.335463258785943,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1301,
+      "step": 8974
+    },
+    {
+      "epoch": 14.3370607028754,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1368,
+      "step": 8975
+    },
+    {
+      "epoch": 14.338658146964857,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 8976
+    },
+    {
+      "epoch": 14.340255591054314,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1345,
+      "step": 8977
+    },
+    {
+      "epoch": 14.34185303514377,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 8978
+    },
+    {
+      "epoch": 14.343450479233226,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1228,
+      "step": 8979
+    },
+    {
+      "epoch": 14.345047923322683,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 8980
+    },
+    {
+      "epoch": 14.34664536741214,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 8981
+    },
+    {
+      "epoch": 14.348242811501597,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.133,
+      "step": 8982
+    },
+    {
+      "epoch": 14.349840255591054,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1222,
+      "step": 8983
+    },
+    {
+      "epoch": 14.351437699680512,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 8984
+    },
+    {
+      "epoch": 14.353035143769969,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1279,
+      "step": 8985
+    },
+    {
+      "epoch": 14.354632587859426,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 8986
+    },
+    {
+      "epoch": 14.356230031948883,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 8987
+    },
+    {
+      "epoch": 14.357827476038338,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 8988
+    },
+    {
+      "epoch": 14.359424920127795,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 8989
+    },
+    {
+      "epoch": 14.361022364217252,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1318,
+      "step": 8990
+    },
+    {
+      "epoch": 14.36261980830671,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1195,
+      "step": 8991
+    },
+    {
+      "epoch": 14.364217252396166,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 8992
+    },
+    {
+      "epoch": 14.365814696485623,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 8993
+    },
+    {
+      "epoch": 14.36741214057508,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 8994
+    },
+    {
+      "epoch": 14.369009584664537,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 8995
+    },
+    {
+      "epoch": 14.370607028753994,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 8996
+    },
+    {
+      "epoch": 14.37220447284345,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 8997
+    },
+    {
+      "epoch": 14.373801916932907,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 8998
+    },
+    {
+      "epoch": 14.375399361022364,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 8999
+    },
+    {
+      "epoch": 14.37699680511182,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1192,
+      "step": 9000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1001600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1600,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.82292735410674e+18,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}