diff --git "a/checkpoint-10000/trainer_state.json" "b/checkpoint-10000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-10000/trainer_state.json"
@@ -0,0 +1,70033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9952057530962843,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00039952057530962844,
+      "grad_norm": 7.03125,
+      "learning_rate": 0.0005,
+      "loss": 3.7958,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007990411506192569,
+      "grad_norm": 47.5,
+      "learning_rate": 0.0005,
+      "loss": 16.0438,
+      "step": 2
+    },
+    {
+      "epoch": 0.0011985617259288853,
+      "grad_norm": 41.5,
+      "learning_rate": 0.0005,
+      "loss": 11.6589,
+      "step": 3
+    },
+    {
+      "epoch": 0.0015980823012385138,
+      "grad_norm": 40.0,
+      "learning_rate": 0.0005,
+      "loss": 12.8325,
+      "step": 4
+    },
+    {
+      "epoch": 0.001997602876548142,
+      "grad_norm": 36.75,
+      "learning_rate": 0.0005,
+      "loss": 12.5015,
+      "step": 5
+    },
+    {
+      "epoch": 0.0023971234518577705,
+      "grad_norm": 21.625,
+      "learning_rate": 0.0005,
+      "loss": 7.0332,
+      "step": 6
+    },
+    {
+      "epoch": 0.002796644027167399,
+      "grad_norm": 13.8125,
+      "learning_rate": 0.0005,
+      "loss": 4.1196,
+      "step": 7
+    },
+    {
+      "epoch": 0.0031961646024770275,
+      "grad_norm": 6.5,
+      "learning_rate": 0.0005,
+      "loss": 2.8851,
+      "step": 8
+    },
+    {
+      "epoch": 0.003595685177786656,
+      "grad_norm": 4.5,
+      "learning_rate": 0.0005,
+      "loss": 2.5168,
+      "step": 9
+    },
+    {
+      "epoch": 0.003995205753096284,
+      "grad_norm": 8.1875,
+      "learning_rate": 0.0005,
+      "loss": 2.9302,
+      "step": 10
+    },
+    {
+      "epoch": 0.0043947263284059125,
+      "grad_norm": 5.1875,
+      "learning_rate": 0.0005,
+      "loss": 2.3453,
+      "step": 11
+    },
+    {
+      "epoch": 0.004794246903715541,
+      "grad_norm": 3.5,
+      "learning_rate": 0.0005,
+      "loss": 2.1391,
+      "step": 12
+    },
+    {
+      "epoch": 0.0051937674790251695,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.9119,
+      "step": 13
+    },
+    {
+      "epoch": 0.005593288054334798,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0005,
+      "loss": 1.8872,
+      "step": 14
+    },
+    {
+      "epoch": 0.0059928086296444265,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0005,
+      "loss": 1.8285,
+      "step": 15
+    },
+    {
+      "epoch": 0.006392329204954055,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0005,
+      "loss": 1.8244,
+      "step": 16
+    },
+    {
+      "epoch": 0.0067918497802636835,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0005,
+      "loss": 1.8067,
+      "step": 17
+    },
+    {
+      "epoch": 0.007191370355573312,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.7836,
+      "step": 18
+    },
+    {
+      "epoch": 0.0075908909308829405,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.772,
+      "step": 19
+    },
+    {
+      "epoch": 0.007990411506192568,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.7875,
+      "step": 20
+    },
+    {
+      "epoch": 0.008389932081502197,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.7794,
+      "step": 21
+    },
+    {
+      "epoch": 0.008789452656811825,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.7714,
+      "step": 22
+    },
+    {
+      "epoch": 0.009188973232121454,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.7494,
+      "step": 23
+    },
+    {
+      "epoch": 0.009588493807431082,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.7309,
+      "step": 24
+    },
+    {
+      "epoch": 0.00998801438274071,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.7147,
+      "step": 25
+    },
+    {
+      "epoch": 0.010387534958050339,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.7107,
+      "step": 26
+    },
+    {
+      "epoch": 0.010787055533359968,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.7006,
+      "step": 27
+    },
+    {
+      "epoch": 0.011186576108669596,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0005,
+      "loss": 1.6963,
+      "step": 28
+    },
+    {
+      "epoch": 0.011586096683979225,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005,
+      "loss": 1.6782,
+      "step": 29
+    },
+    {
+      "epoch": 0.011985617259288853,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.6598,
+      "step": 30
+    },
+    {
+      "epoch": 0.012385137834598482,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.646,
+      "step": 31
+    },
+    {
+      "epoch": 0.01278465840990811,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.6078,
+      "step": 32
+    },
+    {
+      "epoch": 0.013184178985217739,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.5961,
+      "step": 33
+    },
+    {
+      "epoch": 0.013583699560527367,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.5956,
+      "step": 34
+    },
+    {
+      "epoch": 0.013983220135836995,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.5822,
+      "step": 35
+    },
+    {
+      "epoch": 0.014382740711146624,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.5709,
+      "step": 36
+    },
+    {
+      "epoch": 0.014782261286456252,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.5436,
+      "step": 37
+    },
+    {
+      "epoch": 0.015181781861765881,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.5349,
+      "step": 38
+    },
+    {
+      "epoch": 0.01558130243707551,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.5121,
+      "step": 39
+    },
+    {
+      "epoch": 0.015980823012385136,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.5008,
+      "step": 40
+    },
+    {
+      "epoch": 0.016380343587694766,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.5031,
+      "step": 41
+    },
+    {
+      "epoch": 0.016779864163004393,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.4996,
+      "step": 42
+    },
+    {
+      "epoch": 0.017179384738314023,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.4847,
+      "step": 43
+    },
+    {
+      "epoch": 0.01757890531362365,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.4572,
+      "step": 44
+    },
+    {
+      "epoch": 0.01797842588893328,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.4555,
+      "step": 45
+    },
+    {
+      "epoch": 0.018377946464242907,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.4598,
+      "step": 46
+    },
+    {
+      "epoch": 0.018777467039552537,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.4292,
+      "step": 47
+    },
+    {
+      "epoch": 0.019176987614862164,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0005,
+      "loss": 1.4271,
+      "step": 48
+    },
+    {
+      "epoch": 0.019576508190171794,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.4139,
+      "step": 49
+    },
+    {
+      "epoch": 0.01997602876548142,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.4094,
+      "step": 50
+    },
+    {
+      "epoch": 0.02037554934079105,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.4106,
+      "step": 51
+    },
+    {
+      "epoch": 0.020775069916100678,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.399,
+      "step": 52
+    },
+    {
+      "epoch": 0.02117459049141031,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.3949,
+      "step": 53
+    },
+    {
+      "epoch": 0.021574111066719935,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3915,
+      "step": 54
+    },
+    {
+      "epoch": 0.021973631642029565,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.3763,
+      "step": 55
+    },
+    {
+      "epoch": 0.022373152217339192,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.3636,
+      "step": 56
+    },
+    {
+      "epoch": 0.022772672792648822,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3694,
+      "step": 57
+    },
+    {
+      "epoch": 0.02317219336795845,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.3668,
+      "step": 58
+    },
+    {
+      "epoch": 0.02357171394326808,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.3664,
+      "step": 59
+    },
+    {
+      "epoch": 0.023971234518577706,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.3588,
+      "step": 60
+    },
+    {
+      "epoch": 0.024370755093887336,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0005,
+      "loss": 1.356,
+      "step": 61
+    },
+    {
+      "epoch": 0.024770275669196963,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.3621,
+      "step": 62
+    },
+    {
+      "epoch": 0.025169796244506593,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3493,
+      "step": 63
+    },
+    {
+      "epoch": 0.02556931681981622,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.3488,
+      "step": 64
+    },
+    {
+      "epoch": 0.02596883739512585,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.3591,
+      "step": 65
+    },
+    {
+      "epoch": 0.026368357970435477,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0005,
+      "loss": 1.3558,
+      "step": 66
+    },
+    {
+      "epoch": 0.026767878545745107,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0005,
+      "loss": 1.3593,
+      "step": 67
+    },
+    {
+      "epoch": 0.027167399121054734,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.3227,
+      "step": 68
+    },
+    {
+      "epoch": 0.027566919696364364,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.3502,
+      "step": 69
+    },
+    {
+      "epoch": 0.02796644027167399,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.3437,
+      "step": 70
+    },
+    {
+      "epoch": 0.02836596084698362,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0005,
+      "loss": 1.3553,
+      "step": 71
+    },
+    {
+      "epoch": 0.028765481422293248,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3279,
+      "step": 72
+    },
+    {
+      "epoch": 0.029165001997602878,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0005,
+      "loss": 1.3416,
+      "step": 73
+    },
+    {
+      "epoch": 0.029564522572912505,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.3436,
+      "step": 74
+    },
+    {
+      "epoch": 0.02996404314822213,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3348,
+      "step": 75
+    },
+    {
+      "epoch": 0.030363563723531762,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.3171,
+      "step": 76
+    },
+    {
+      "epoch": 0.03076308429884139,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.3143,
+      "step": 77
+    },
+    {
+      "epoch": 0.03116260487415102,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3066,
+      "step": 78
+    },
+    {
+      "epoch": 0.031562125449460646,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.3274,
+      "step": 79
+    },
+    {
+      "epoch": 0.03196164602477027,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.3102,
+      "step": 80
+    },
+    {
+      "epoch": 0.032361166600079906,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.312,
+      "step": 81
+    },
+    {
+      "epoch": 0.03276068717538953,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.3134,
+      "step": 82
+    },
+    {
+      "epoch": 0.03316020775069916,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2988,
+      "step": 83
+    },
+    {
+      "epoch": 0.033559728326008786,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.3042,
+      "step": 84
+    },
+    {
+      "epoch": 0.03395924890131842,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.3062,
+      "step": 85
+    },
+    {
+      "epoch": 0.03435876947662805,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2982,
+      "step": 86
+    },
+    {
+      "epoch": 0.034758290051937674,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.3077,
+      "step": 87
+    },
+    {
+      "epoch": 0.0351578106272473,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.3042,
+      "step": 88
+    },
+    {
+      "epoch": 0.035557331202556934,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.2904,
+      "step": 89
+    },
+    {
+      "epoch": 0.03595685177786656,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2845,
+      "step": 90
+    },
+    {
+      "epoch": 0.03635637235317619,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.3028,
+      "step": 91
+    },
+    {
+      "epoch": 0.036755892928485814,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2846,
+      "step": 92
+    },
+    {
+      "epoch": 0.03715541350379545,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.281,
+      "step": 93
+    },
+    {
+      "epoch": 0.037554934079105075,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2841,
+      "step": 94
+    },
+    {
+      "epoch": 0.0379544546544147,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2928,
+      "step": 95
+    },
+    {
+      "epoch": 0.03835397522972433,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2807,
+      "step": 96
+    },
+    {
+      "epoch": 0.03875349580503396,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2837,
+      "step": 97
+    },
+    {
+      "epoch": 0.03915301638034359,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2684,
+      "step": 98
+    },
+    {
+      "epoch": 0.039552536955653216,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2694,
+      "step": 99
+    },
+    {
+      "epoch": 0.03995205753096284,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2613,
+      "step": 100
+    },
+    {
+      "epoch": 0.040351578106272476,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.2594,
+      "step": 101
+    },
+    {
+      "epoch": 0.0407510986815821,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2697,
+      "step": 102
+    },
+    {
+      "epoch": 0.04115061925689173,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2636,
+      "step": 103
+    },
+    {
+      "epoch": 0.041550139832201356,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2633,
+      "step": 104
+    },
+    {
+      "epoch": 0.04194966040751099,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2651,
+      "step": 105
+    },
+    {
+      "epoch": 0.04234918098282062,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2696,
+      "step": 106
+    },
+    {
+      "epoch": 0.042748701558130243,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2726,
+      "step": 107
+    },
+    {
+      "epoch": 0.04314822213343987,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2519,
+      "step": 108
+    },
+    {
+      "epoch": 0.043547742708749504,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2695,
+      "step": 109
+    },
+    {
+      "epoch": 0.04394726328405913,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2667,
+      "step": 110
+    },
+    {
+      "epoch": 0.04434678385936876,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2507,
+      "step": 111
+    },
+    {
+      "epoch": 0.044746304434678384,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2676,
+      "step": 112
+    },
+    {
+      "epoch": 0.04514582500998801,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.2574,
+      "step": 113
+    },
+    {
+      "epoch": 0.045545345585297645,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2484,
+      "step": 114
+    },
+    {
+      "epoch": 0.04594486616060727,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2514,
+      "step": 115
+    },
+    {
+      "epoch": 0.0463443867359169,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2374,
+      "step": 116
+    },
+    {
+      "epoch": 0.046743907311226525,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2643,
+      "step": 117
+    },
+    {
+      "epoch": 0.04714342788653616,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2525,
+      "step": 118
+    },
+    {
+      "epoch": 0.047542948461845785,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2457,
+      "step": 119
+    },
+    {
+      "epoch": 0.04794246903715541,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2422,
+      "step": 120
+    },
+    {
+      "epoch": 0.04834198961246504,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2451,
+      "step": 121
+    },
+    {
+      "epoch": 0.04874151018777467,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.2453,
+      "step": 122
+    },
+    {
+      "epoch": 0.0491410307630843,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2419,
+      "step": 123
+    },
+    {
+      "epoch": 0.049540551338393926,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2361,
+      "step": 124
+    },
+    {
+      "epoch": 0.04994007191370355,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.221,
+      "step": 125
+    },
+    {
+      "epoch": 0.05033959248901319,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.238,
+      "step": 126
+    },
+    {
+      "epoch": 0.05073911306432281,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2425,
+      "step": 127
+    },
+    {
+      "epoch": 0.05113863363963244,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2299,
+      "step": 128
+    },
+    {
+      "epoch": 0.05153815421494207,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2447,
+      "step": 129
+    },
+    {
+      "epoch": 0.0519376747902517,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2401,
+      "step": 130
+    },
+    {
+      "epoch": 0.05233719536556133,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.2332,
+      "step": 131
+    },
+    {
+      "epoch": 0.052736715940870954,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2308,
+      "step": 132
+    },
+    {
+      "epoch": 0.05313623651618058,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2323,
+      "step": 133
+    },
+    {
+      "epoch": 0.053535757091490214,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2371,
+      "step": 134
+    },
+    {
+      "epoch": 0.05393527766679984,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2282,
+      "step": 135
+    },
+    {
+      "epoch": 0.05433479824210947,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2391,
+      "step": 136
+    },
+    {
+      "epoch": 0.054734318817419095,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.2467,
+      "step": 137
+    },
+    {
+      "epoch": 0.05513383939272873,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2266,
+      "step": 138
+    },
+    {
+      "epoch": 0.055533359968038355,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2468,
+      "step": 139
+    },
+    {
+      "epoch": 0.05593288054334798,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2299,
+      "step": 140
+    },
+    {
+      "epoch": 0.05633240111865761,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2147,
+      "step": 141
+    },
+    {
+      "epoch": 0.05673192169396724,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.234,
+      "step": 142
+    },
+    {
+      "epoch": 0.05713144226927687,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2296,
+      "step": 143
+    },
+    {
+      "epoch": 0.057530962844586496,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2334,
+      "step": 144
+    },
+    {
+      "epoch": 0.05793048341989612,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2367,
+      "step": 145
+    },
+    {
+      "epoch": 0.058330003995205756,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2225,
+      "step": 146
+    },
+    {
+      "epoch": 0.05872952457051538,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2325,
+      "step": 147
+    },
+    {
+      "epoch": 0.05912904514582501,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2237,
+      "step": 148
+    },
+    {
+      "epoch": 0.05952856572113464,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2226,
+      "step": 149
+    },
+    {
+      "epoch": 0.05992808629644426,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.225,
+      "step": 150
+    },
+    {
+      "epoch": 0.0603276068717539,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2315,
+      "step": 151
+    },
+    {
+      "epoch": 0.060727127447063524,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2224,
+      "step": 152
+    },
+    {
+      "epoch": 0.06112664802237315,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2152,
+      "step": 153
+    },
+    {
+      "epoch": 0.06152616859768278,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2128,
+      "step": 154
+    },
+    {
+      "epoch": 0.06192568917299241,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.2218,
+      "step": 155
+    },
+    {
+      "epoch": 0.06232520974830204,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2318,
+      "step": 156
+    },
+    {
+      "epoch": 0.06272473032361167,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2094,
+      "step": 157
+    },
+    {
+      "epoch": 0.06312425089892129,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2367,
+      "step": 158
+    },
+    {
+      "epoch": 0.06352377147423093,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2254,
+      "step": 159
+    },
+    {
+      "epoch": 0.06392329204954054,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.23,
+      "step": 160
+    },
+    {
+      "epoch": 0.06432281262485018,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2304,
+      "step": 161
+    },
+    {
+      "epoch": 0.06472233320015981,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2216,
+      "step": 162
+    },
+    {
+      "epoch": 0.06512185377546943,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2211,
+      "step": 163
+    },
+    {
+      "epoch": 0.06552137435077907,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2095,
+      "step": 164
+    },
+    {
+      "epoch": 0.0659208949260887,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2175,
+      "step": 165
+    },
+    {
+      "epoch": 0.06632041550139832,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.201,
+      "step": 166
+    },
+    {
+      "epoch": 0.06671993607670795,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.214,
+      "step": 167
+    },
+    {
+      "epoch": 0.06711945665201757,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2096,
+      "step": 168
+    },
+    {
+      "epoch": 0.0675189772273272,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.2102,
+      "step": 169
+    },
+    {
+      "epoch": 0.06791849780263684,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2061,
+      "step": 170
+    },
+    {
+      "epoch": 0.06831801837794646,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.2017,
+      "step": 171
+    },
+    {
+      "epoch": 0.0687175389532561,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1911,
+      "step": 172
+    },
+    {
+      "epoch": 0.06911705952856573,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.2196,
+      "step": 173
+    },
+    {
+      "epoch": 0.06951658010387535,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2002,
+      "step": 174
+    },
+    {
+      "epoch": 0.06991610067918498,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2146,
+      "step": 175
+    },
+    {
+      "epoch": 0.0703156212544946,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.217,
+      "step": 176
+    },
+    {
+      "epoch": 0.07071514182980423,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1998,
+      "step": 177
+    },
+    {
+      "epoch": 0.07111466240511387,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.2189,
+      "step": 178
+    },
+    {
+      "epoch": 0.07151418298042349,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2081,
+      "step": 179
+    },
+    {
+      "epoch": 0.07191370355573312,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2105,
+      "step": 180
+    },
+    {
+      "epoch": 0.07231322413104276,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2148,
+      "step": 181
+    },
+    {
+      "epoch": 0.07271274470635238,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1977,
+      "step": 182
+    },
+    {
+      "epoch": 0.07311226528166201,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.2027,
+      "step": 183
+    },
+    {
+      "epoch": 0.07351178585697163,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2038,
+      "step": 184
+    },
+    {
+      "epoch": 0.07391130643228126,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.207,
+      "step": 185
+    },
+    {
+      "epoch": 0.0743108270075909,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2025,
+      "step": 186
+    },
+    {
+      "epoch": 0.07471034758290052,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1999,
+      "step": 187
+    },
+    {
+      "epoch": 0.07510986815821015,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2108,
+      "step": 188
+    },
+    {
+      "epoch": 0.07550938873351977,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2082,
+      "step": 189
+    },
+    {
+      "epoch": 0.0759089093088294,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.2076,
+      "step": 190
+    },
+    {
+      "epoch": 0.07630842988413904,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2073,
+      "step": 191
+    },
+    {
+      "epoch": 0.07670795045944866,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1904,
+      "step": 192
+    },
+    {
+      "epoch": 0.07710747103475829,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.2029,
+      "step": 193
+    },
+    {
+      "epoch": 0.07750699161006792,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1945,
+      "step": 194
+    },
+    {
+      "epoch": 0.07790651218537754,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2095,
+      "step": 195
+    },
+    {
+      "epoch": 0.07830603276068718,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2029,
+      "step": 196
+    },
+    {
+      "epoch": 0.0787055533359968,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2025,
+      "step": 197
+    },
+    {
+      "epoch": 0.07910507391130643,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1891,
+      "step": 198
+    },
+    {
+      "epoch": 0.07950459448661606,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 199
+    },
+    {
+      "epoch": 0.07990411506192568,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2055,
+      "step": 200
+    },
+    {
+      "epoch": 0.08030363563723532,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2098,
+      "step": 201
+    },
+    {
+      "epoch": 0.08070315621254495,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1928,
+      "step": 202
+    },
+    {
+      "epoch": 0.08110267678785457,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1964,
+      "step": 203
+    },
+    {
+      "epoch": 0.0815021973631642,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1919,
+      "step": 204
+    },
+    {
+      "epoch": 0.08190171793847383,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1907,
+      "step": 205
+    },
+    {
+      "epoch": 0.08230123851378346,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1992,
+      "step": 206
+    },
+    {
+      "epoch": 0.08270075908909309,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.2075,
+      "step": 207
+    },
+    {
+      "epoch": 0.08310027966440271,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2005,
+      "step": 208
+    },
+    {
+      "epoch": 0.08349980023971235,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1882,
+      "step": 209
+    },
+    {
+      "epoch": 0.08389932081502198,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1984,
+      "step": 210
+    },
+    {
+      "epoch": 0.0842988413903316,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.2138,
+      "step": 211
+    },
+    {
+      "epoch": 0.08469836196564123,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1859,
+      "step": 212
+    },
+    {
+      "epoch": 0.08509788254095085,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1894,
+      "step": 213
+    },
+    {
+      "epoch": 0.08549740311626049,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.1867,
+      "step": 214
+    },
+    {
+      "epoch": 0.08589692369157012,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0005,
+      "loss": 1.2184,
+      "step": 215
+    },
+    {
+      "epoch": 0.08629644426687974,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2526,
+      "step": 216
+    },
+    {
+      "epoch": 0.08669596484218937,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2766,
+      "step": 217
+    },
+    {
+      "epoch": 0.08709548541749901,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2506,
+      "step": 218
+    },
+    {
+      "epoch": 0.08749500599280863,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.255,
+      "step": 219
+    },
+    {
+      "epoch": 0.08789452656811826,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.2583,
+      "step": 220
+    },
+    {
+      "epoch": 0.08829404714342788,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2486,
+      "step": 221
+    },
+    {
+      "epoch": 0.08869356771873751,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.2367,
+      "step": 222
+    },
+    {
+      "epoch": 0.08909308829404715,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.2234,
+      "step": 223
+    },
+    {
+      "epoch": 0.08949260886935677,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2375,
+      "step": 224
+    },
+    {
+      "epoch": 0.0898921294446664,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.236,
+      "step": 225
+    },
+    {
+      "epoch": 0.09029165001997602,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2155,
+      "step": 226
+    },
+    {
+      "epoch": 0.09069117059528566,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2178,
+      "step": 227
+    },
+    {
+      "epoch": 0.09109069117059529,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2117,
+      "step": 228
+    },
+    {
+      "epoch": 0.09149021174590491,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2235,
+      "step": 229
+    },
+    {
+      "epoch": 0.09188973232121454,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2189,
+      "step": 230
+    },
+    {
+      "epoch": 0.09228925289652418,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.2079,
+      "step": 231
+    },
+    {
+      "epoch": 0.0926887734718338,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.2094,
+      "step": 232
+    },
+    {
+      "epoch": 0.09308829404714343,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1981,
+      "step": 233
+    },
+    {
+      "epoch": 0.09348781462245305,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.2178,
+      "step": 234
+    },
+    {
+      "epoch": 0.09388733519776268,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.2071,
+      "step": 235
+    },
+    {
+      "epoch": 0.09428685577307232,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1894,
+      "step": 236
+    },
+    {
+      "epoch": 0.09468637634838194,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.192,
+      "step": 237
+    },
+    {
+      "epoch": 0.09508589692369157,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.199,
+      "step": 238
+    },
+    {
+      "epoch": 0.0954854174990012,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.2165,
+      "step": 239
+    },
+    {
+      "epoch": 0.09588493807431082,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1975,
+      "step": 240
+    },
+    {
+      "epoch": 0.09628445864962046,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1948,
+      "step": 241
+    },
+    {
+      "epoch": 0.09668397922493008,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1932,
+      "step": 242
+    },
+    {
+      "epoch": 0.09708349980023971,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1978,
+      "step": 243
+    },
+    {
+      "epoch": 0.09748302037554935,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1932,
+      "step": 244
+    },
+    {
+      "epoch": 0.09788254095085897,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1988,
+      "step": 245
+    },
+    {
+      "epoch": 0.0982820615261686,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2009,
+      "step": 246
+    },
+    {
+      "epoch": 0.09868158210147823,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.2003,
+      "step": 247
+    },
+    {
+      "epoch": 0.09908110267678785,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.2001,
+      "step": 248
+    },
+    {
+      "epoch": 0.09948062325209749,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.19,
+      "step": 249
+    },
+    {
+      "epoch": 0.0998801438274071,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1888,
+      "step": 250
+    },
+    {
+      "epoch": 0.10027966440271674,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1851,
+      "step": 251
+    },
+    {
+      "epoch": 0.10067918497802637,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.179,
+      "step": 252
+    },
+    {
+      "epoch": 0.10107870555333599,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1999,
+      "step": 253
+    },
+    {
+      "epoch": 0.10147822612864563,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.2023,
+      "step": 254
+    },
+    {
+      "epoch": 0.10187774670395526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1998,
+      "step": 255
+    },
+    {
+      "epoch": 0.10227726727926488,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1926,
+      "step": 256
+    },
+    {
+      "epoch": 0.10267678785457451,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1924,
+      "step": 257
+    },
+    {
+      "epoch": 0.10307630842988413,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1782,
+      "step": 258
+    },
+    {
+      "epoch": 0.10347582900519377,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.198,
+      "step": 259
+    },
+    {
+      "epoch": 0.1038753495805034,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1839,
+      "step": 260
+    },
+    {
+      "epoch": 0.10427487015581302,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1991,
+      "step": 261
+    },
+    {
+      "epoch": 0.10467439073112265,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.185,
+      "step": 262
+    },
+    {
+      "epoch": 0.10507391130643227,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1815,
+      "step": 263
+    },
+    {
+      "epoch": 0.10547343188174191,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1967,
+      "step": 264
+    },
+    {
+      "epoch": 0.10587295245705154,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1831,
+      "step": 265
+    },
+    {
+      "epoch": 0.10627247303236116,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1872,
+      "step": 266
+    },
+    {
+      "epoch": 0.1066719936076708,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1988,
+      "step": 267
+    },
+    {
+      "epoch": 0.10707151418298043,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1834,
+      "step": 268
+    },
+    {
+      "epoch": 0.10747103475829005,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1782,
+      "step": 269
+    },
+    {
+      "epoch": 0.10787055533359968,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1779,
+      "step": 270
+    },
+    {
+      "epoch": 0.1082700759089093,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1743,
+      "step": 271
+    },
+    {
+      "epoch": 0.10866959648421894,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1861,
+      "step": 272
+    },
+    {
+      "epoch": 0.10906911705952857,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1875,
+      "step": 273
+    },
+    {
+      "epoch": 0.10946863763483819,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 274
+    },
+    {
+      "epoch": 0.10986815821014782,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1796,
+      "step": 275
+    },
+    {
+      "epoch": 0.11026767878545746,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1808,
+      "step": 276
+    },
+    {
+      "epoch": 0.11066719936076708,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0005,
+      "loss": 1.2016,
+      "step": 277
+    },
+    {
+      "epoch": 0.11106671993607671,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0005,
+      "loss": 1.2221,
+      "step": 278
+    },
+    {
+      "epoch": 0.11146624051138633,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.2108,
+      "step": 279
+    },
+    {
+      "epoch": 0.11186576108669596,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1984,
+      "step": 280
+    },
+    {
+      "epoch": 0.1122652816620056,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.2073,
+      "step": 281
+    },
+    {
+      "epoch": 0.11266480223731522,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1945,
+      "step": 282
+    },
+    {
+      "epoch": 0.11306432281262485,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.184,
+      "step": 283
+    },
+    {
+      "epoch": 0.11346384338793448,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1903,
+      "step": 284
+    },
+    {
+      "epoch": 0.1138633639632441,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1902,
+      "step": 285
+    },
+    {
+      "epoch": 0.11426288453855374,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1941,
+      "step": 286
+    },
+    {
+      "epoch": 0.11466240511386336,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.2027,
+      "step": 287
+    },
+    {
+      "epoch": 0.11506192568917299,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 288
+    },
+    {
+      "epoch": 0.11546144626448263,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1766,
+      "step": 289
+    },
+    {
+      "epoch": 0.11586096683979225,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1671,
+      "step": 290
+    },
+    {
+      "epoch": 0.11626048741510188,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1756,
+      "step": 291
+    },
+    {
+      "epoch": 0.11666000799041151,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.2036,
+      "step": 292
+    },
+    {
+      "epoch": 0.11705952856572113,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1797,
+      "step": 293
+    },
+    {
+      "epoch": 0.11745904914103077,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1793,
+      "step": 294
+    },
+    {
+      "epoch": 0.11785856971634039,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.177,
+      "step": 295
+    },
+    {
+      "epoch": 0.11825809029165002,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.178,
+      "step": 296
+    },
+    {
+      "epoch": 0.11865761086695965,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 297
+    },
+    {
+      "epoch": 0.11905713144226927,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1791,
+      "step": 298
+    },
+    {
+      "epoch": 0.11945665201757891,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1904,
+      "step": 299
+    },
+    {
+      "epoch": 0.11985617259288853,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1701,
+      "step": 300
+    },
+    {
+      "epoch": 0.12025569316819816,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1809,
+      "step": 301
+    },
+    {
+      "epoch": 0.1206552137435078,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 302
+    },
+    {
+      "epoch": 0.12105473431881741,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 303
+    },
+    {
+      "epoch": 0.12145425489412705,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1712,
+      "step": 304
+    },
+    {
+      "epoch": 0.12185377546943668,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1784,
+      "step": 305
+    },
+    {
+      "epoch": 0.1222532960447463,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1764,
+      "step": 306
+    },
+    {
+      "epoch": 0.12265281662005593,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 307
+    },
+    {
+      "epoch": 0.12305233719536555,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1663,
+      "step": 308
+    },
+    {
+      "epoch": 0.12345185777067519,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 309
+    },
+    {
+      "epoch": 0.12385137834598482,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1696,
+      "step": 310
+    },
+    {
+      "epoch": 0.12425089892129444,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.164,
+      "step": 311
+    },
+    {
+      "epoch": 0.12465041949660408,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1665,
+      "step": 312
+    },
+    {
+      "epoch": 0.1250499400719137,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1808,
+      "step": 313
+    },
+    {
+      "epoch": 0.12544946064722334,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1648,
+      "step": 314
+    },
+    {
+      "epoch": 0.12584898122253296,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1614,
+      "step": 315
+    },
+    {
+      "epoch": 0.12624850179784258,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1653,
+      "step": 316
+    },
+    {
+      "epoch": 0.12664802237315223,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 317
+    },
+    {
+      "epoch": 0.12704754294846185,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 318
+    },
+    {
+      "epoch": 0.12744706352377147,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1654,
+      "step": 319
+    },
+    {
+      "epoch": 0.1278465840990811,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1677,
+      "step": 320
+    },
+    {
+      "epoch": 0.12824610467439074,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 321
+    },
+    {
+      "epoch": 0.12864562524970036,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1775,
+      "step": 322
+    },
+    {
+      "epoch": 0.12904514582500998,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0005,
+      "loss": 1.186,
+      "step": 323
+    },
+    {
+      "epoch": 0.12944466640031962,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1871,
+      "step": 324
+    },
+    {
+      "epoch": 0.12984418697562924,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1934,
+      "step": 325
+    },
+    {
+      "epoch": 0.13024370755093886,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1796,
+      "step": 326
+    },
+    {
+      "epoch": 0.1306432281262485,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.1741,
+      "step": 327
+    },
+    {
+      "epoch": 0.13104274870155813,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1744,
+      "step": 328
+    },
+    {
+      "epoch": 0.13144226927686775,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1802,
+      "step": 329
+    },
+    {
+      "epoch": 0.1318417898521774,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.172,
+      "step": 330
+    },
+    {
+      "epoch": 0.13224131042748702,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1667,
+      "step": 331
+    },
+    {
+      "epoch": 0.13264083100279664,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 332
+    },
+    {
+      "epoch": 0.1330403515781063,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1617,
+      "step": 333
+    },
+    {
+      "epoch": 0.1334398721534159,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1657,
+      "step": 334
+    },
+    {
+      "epoch": 0.13383939272872553,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1628,
+      "step": 335
+    },
+    {
+      "epoch": 0.13423891330403515,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.17,
+      "step": 336
+    },
+    {
+      "epoch": 0.1346384338793448,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1576,
+      "step": 337
+    },
+    {
+      "epoch": 0.1350379544546544,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.17,
+      "step": 338
+    },
+    {
+      "epoch": 0.13543747502996403,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1678,
+      "step": 339
+    },
+    {
+      "epoch": 0.13583699560527368,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1596,
+      "step": 340
+    },
+    {
+      "epoch": 0.1362365161805833,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1674,
+      "step": 341
+    },
+    {
+      "epoch": 0.13663603675589292,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1605,
+      "step": 342
+    },
+    {
+      "epoch": 0.13703555733120257,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1593,
+      "step": 343
+    },
+    {
+      "epoch": 0.1374350779065122,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1692,
+      "step": 344
+    },
+    {
+      "epoch": 0.1378345984818218,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1635,
+      "step": 345
+    },
+    {
+      "epoch": 0.13823411905713145,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 346
+    },
+    {
+      "epoch": 0.13863363963244107,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1606,
+      "step": 347
+    },
+    {
+      "epoch": 0.1390331602077507,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 348
+    },
+    {
+      "epoch": 0.13943268078306031,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1624,
+      "step": 349
+    },
+    {
+      "epoch": 0.13983220135836996,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1525,
+      "step": 350
+    },
+    {
+      "epoch": 0.14023172193367958,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1568,
+      "step": 351
+    },
+    {
+      "epoch": 0.1406312425089892,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1618,
+      "step": 352
+    },
+    {
+      "epoch": 0.14103076308429885,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 353
+    },
+    {
+      "epoch": 0.14143028365960847,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1542,
+      "step": 354
+    },
+    {
+      "epoch": 0.1418298042349181,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1482,
+      "step": 355
+    },
+    {
+      "epoch": 0.14222932481022774,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1621,
+      "step": 356
+    },
+    {
+      "epoch": 0.14262884538553736,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 357
+    },
+    {
+      "epoch": 0.14302836596084698,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1633,
+      "step": 358
+    },
+    {
+      "epoch": 0.14342788653615662,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 359
+    },
+    {
+      "epoch": 0.14382740711146624,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1562,
+      "step": 360
+    },
+    {
+      "epoch": 0.14422692768677586,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.156,
+      "step": 361
+    },
+    {
+      "epoch": 0.1446264482620855,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1581,
+      "step": 362
+    },
+    {
+      "epoch": 0.14502596883739513,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1488,
+      "step": 363
+    },
+    {
+      "epoch": 0.14542548941270475,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1453,
+      "step": 364
+    },
+    {
+      "epoch": 0.14582500998801437,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 365
+    },
+    {
+      "epoch": 0.14622453056332402,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 366
+    },
+    {
+      "epoch": 0.14662405113863364,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1548,
+      "step": 367
+    },
+    {
+      "epoch": 0.14702357171394326,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1584,
+      "step": 368
+    },
+    {
+      "epoch": 0.1474230922892529,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 369
+    },
+    {
+      "epoch": 0.14782261286456252,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1503,
+      "step": 370
+    },
+    {
+      "epoch": 0.14822213343987214,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 371
+    },
+    {
+      "epoch": 0.1486216540151818,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 372
+    },
+    {
+      "epoch": 0.1490211745904914,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 373
+    },
+    {
+      "epoch": 0.14942069516580103,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1533,
+      "step": 374
+    },
+    {
+      "epoch": 0.14982021574111068,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1571,
+      "step": 375
+    },
+    {
+      "epoch": 0.1502197363164203,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1485,
+      "step": 376
+    },
+    {
+      "epoch": 0.15061925689172992,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1492,
+      "step": 377
+    },
+    {
+      "epoch": 0.15101877746703954,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 378
+    },
+    {
+      "epoch": 0.1514182980423492,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 379
+    },
+    {
+      "epoch": 0.1518178186176588,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1573,
+      "step": 380
+    },
+    {
+      "epoch": 0.15221733919296843,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1579,
+      "step": 381
+    },
+    {
+      "epoch": 0.15261685976827807,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.156,
+      "step": 382
+    },
+    {
+      "epoch": 0.1530163803435877,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1511,
+      "step": 383
+    },
+    {
+      "epoch": 0.1534159009188973,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1443,
+      "step": 384
+    },
+    {
+      "epoch": 0.15381542149420696,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1468,
+      "step": 385
+    },
+    {
+      "epoch": 0.15421494206951658,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 386
+    },
+    {
+      "epoch": 0.1546144626448262,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.1898,
+      "step": 387
+    },
+    {
+      "epoch": 0.15501398322013585,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.2046,
+      "step": 388
+    },
+    {
+      "epoch": 0.15541350379544547,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.1693,
+      "step": 389
+    },
+    {
+      "epoch": 0.1558130243707551,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1749,
+      "step": 390
+    },
+    {
+      "epoch": 0.15621254494606474,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.177,
+      "step": 391
+    },
+    {
+      "epoch": 0.15661206552137436,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1668,
+      "step": 392
+    },
+    {
+      "epoch": 0.15701158609668397,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1896,
+      "step": 393
+    },
+    {
+      "epoch": 0.1574111066719936,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1686,
+      "step": 394
+    },
+    {
+      "epoch": 0.15781062724730324,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1647,
+      "step": 395
+    },
+    {
+      "epoch": 0.15821014782261286,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1523,
+      "step": 396
+    },
+    {
+      "epoch": 0.15860966839792248,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1616,
+      "step": 397
+    },
+    {
+      "epoch": 0.15900918897323213,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 398
+    },
+    {
+      "epoch": 0.15940870954854175,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1722,
+      "step": 399
+    },
+    {
+      "epoch": 0.15980823012385137,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1697,
+      "step": 400
+    },
+    {
+      "epoch": 0.16020775069916102,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1583,
+      "step": 401
+    },
+    {
+      "epoch": 0.16060727127447064,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1689,
+      "step": 402
+    },
+    {
+      "epoch": 0.16100679184978026,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 403
+    },
+    {
+      "epoch": 0.1614063124250899,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 404
+    },
+    {
+      "epoch": 0.16180583300039952,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1612,
+      "step": 405
+    },
+    {
+      "epoch": 0.16220535357570914,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.145,
+      "step": 406
+    },
+    {
+      "epoch": 0.1626048741510188,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 407
+    },
+    {
+      "epoch": 0.1630043947263284,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 408
+    },
+    {
+      "epoch": 0.16340391530163803,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1601,
+      "step": 409
+    },
+    {
+      "epoch": 0.16380343587694765,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1604,
+      "step": 410
+    },
+    {
+      "epoch": 0.1642029564522573,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1564,
+      "step": 411
+    },
+    {
+      "epoch": 0.16460247702756692,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1513,
+      "step": 412
+    },
+    {
+      "epoch": 0.16500199760287654,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1528,
+      "step": 413
+    },
+    {
+      "epoch": 0.16540151817818619,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 414
+    },
+    {
+      "epoch": 0.1658010387534958,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1512,
+      "step": 415
+    },
+    {
+      "epoch": 0.16620055932880543,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 416
+    },
+    {
+      "epoch": 0.16660007990411507,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1574,
+      "step": 417
+    },
+    {
+      "epoch": 0.1669996004794247,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1474,
+      "step": 418
+    },
+    {
+      "epoch": 0.1673991210547343,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1436,
+      "step": 419
+    },
+    {
+      "epoch": 0.16779864163004396,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1517,
+      "step": 420
+    },
+    {
+      "epoch": 0.16819816220535358,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1496,
+      "step": 421
+    },
+    {
+      "epoch": 0.1685976827806632,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 422
+    },
+    {
+      "epoch": 0.16899720335597282,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 423
+    },
+    {
+      "epoch": 0.16939672393128247,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 424
+    },
+    {
+      "epoch": 0.1697962445065921,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1494,
+      "step": 425
+    },
+    {
+      "epoch": 0.1701957650819017,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1435,
+      "step": 426
+    },
+    {
+      "epoch": 0.17059528565721135,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 427
+    },
+    {
+      "epoch": 0.17099480623252097,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1506,
+      "step": 428
+    },
+    {
+      "epoch": 0.1713943268078306,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1417,
+      "step": 429
+    },
+    {
+      "epoch": 0.17179384738314024,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1499,
+      "step": 430
+    },
+    {
+      "epoch": 0.17219336795844986,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1516,
+      "step": 431
+    },
+    {
+      "epoch": 0.17259288853375948,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1509,
+      "step": 432
+    },
+    {
+      "epoch": 0.17299240910906913,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1441,
+      "step": 433
+    },
+    {
+      "epoch": 0.17339192968437875,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 434
+    },
+    {
+      "epoch": 0.17379145025968837,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 435
+    },
+    {
+      "epoch": 0.17419097083499802,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1448,
+      "step": 436
+    },
+    {
+      "epoch": 0.17459049141030764,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1343,
+      "step": 437
+    },
+    {
+      "epoch": 0.17499001198561726,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 438
+    },
+    {
+      "epoch": 0.17538953256092688,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1444,
+      "step": 439
+    },
+    {
+      "epoch": 0.17578905313623652,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1317,
+      "step": 440
+    },
+    {
+      "epoch": 0.17618857371154614,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1473,
+      "step": 441
+    },
+    {
+      "epoch": 0.17658809428685576,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1418,
+      "step": 442
+    },
+    {
+      "epoch": 0.1769876148621654,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 443
+    },
+    {
+      "epoch": 0.17738713543747503,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1554,
+      "step": 444
+    },
+    {
+      "epoch": 0.17778665601278465,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 445
+    },
+    {
+      "epoch": 0.1781861765880943,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1531,
+      "step": 446
+    },
+    {
+      "epoch": 0.17858569716340392,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1341,
+      "step": 447
+    },
+    {
+      "epoch": 0.17898521773871354,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1401,
+      "step": 448
+    },
+    {
+      "epoch": 0.17938473831402318,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1591,
+      "step": 449
+    },
+    {
+      "epoch": 0.1797842588893328,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1412,
+      "step": 450
+    },
+    {
+      "epoch": 0.18018377946464242,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 451
+    },
+    {
+      "epoch": 0.18058330003995204,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1486,
+      "step": 452
+    },
+    {
+      "epoch": 0.1809828206152617,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1508,
+      "step": 453
+    },
+    {
+      "epoch": 0.1813823411905713,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.14,
+      "step": 454
+    },
+    {
+      "epoch": 0.18178186176588093,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 455
+    },
+    {
+      "epoch": 0.18218138234119058,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 456
+    },
+    {
+      "epoch": 0.1825809029165002,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 457
+    },
+    {
+      "epoch": 0.18298042349180982,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 458
+    },
+    {
+      "epoch": 0.18337994406711947,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1456,
+      "step": 459
+    },
+    {
+      "epoch": 0.18377946464242909,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 460
+    },
+    {
+      "epoch": 0.1841789852177387,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.1426,
+      "step": 461
+    },
+    {
+      "epoch": 0.18457850579304835,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1609,
+      "step": 462
+    },
+    {
+      "epoch": 0.18497802636835797,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1669,
+      "step": 463
+    },
+    {
+      "epoch": 0.1853775469436676,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1558,
+      "step": 464
+    },
+    {
+      "epoch": 0.18577706751897724,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 465
+    },
+    {
+      "epoch": 0.18617658809428686,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 466
+    },
+    {
+      "epoch": 0.18657610866959648,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1687,
+      "step": 467
+    },
+    {
+      "epoch": 0.1869756292449061,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1547,
+      "step": 468
+    },
+    {
+      "epoch": 0.18737514982021575,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1379,
+      "step": 469
+    },
+    {
+      "epoch": 0.18777467039552537,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 470
+    },
+    {
+      "epoch": 0.188174190970835,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.1585,
+      "step": 471
+    },
+    {
+      "epoch": 0.18857371154614463,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.153,
+      "step": 472
+    },
+    {
+      "epoch": 0.18897323212145425,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 473
+    },
+    {
+      "epoch": 0.18937275269676387,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1651,
+      "step": 474
+    },
+    {
+      "epoch": 0.18977227327207352,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1518,
+      "step": 475
+    },
+    {
+      "epoch": 0.19017179384738314,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 476
+    },
+    {
+      "epoch": 0.19057131442269276,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1641,
+      "step": 477
+    },
+    {
+      "epoch": 0.1909708349980024,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 478
+    },
+    {
+      "epoch": 0.19137035557331203,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 479
+    },
+    {
+      "epoch": 0.19176987614862165,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.15,
+      "step": 480
+    },
+    {
+      "epoch": 0.1921693967239313,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.125,
+      "step": 481
+    },
+    {
+      "epoch": 0.19256891729924092,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1514,
+      "step": 482
+    },
+    {
+      "epoch": 0.19296843787455054,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.151,
+      "step": 483
+    },
+    {
+      "epoch": 0.19336795844986016,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 484
+    },
+    {
+      "epoch": 0.1937674790251698,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1395,
+      "step": 485
+    },
+    {
+      "epoch": 0.19416699960047942,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1447,
+      "step": 486
+    },
+    {
+      "epoch": 0.19456652017578904,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 487
+    },
+    {
+      "epoch": 0.1949660407510987,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 488
+    },
+    {
+      "epoch": 0.1953655613264083,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1455,
+      "step": 489
+    },
+    {
+      "epoch": 0.19576508190171793,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 490
+    },
+    {
+      "epoch": 0.19616460247702758,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 491
+    },
+    {
+      "epoch": 0.1965641230523372,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 492
+    },
+    {
+      "epoch": 0.19696364362764682,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 493
+    },
+    {
+      "epoch": 0.19736316420295646,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 494
+    },
+    {
+      "epoch": 0.19776268477826608,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1524,
+      "step": 495
+    },
+    {
+      "epoch": 0.1981622053535757,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1425,
+      "step": 496
+    },
+    {
+      "epoch": 0.19856172592888532,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 497
+    },
+    {
+      "epoch": 0.19896124650419497,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 498
+    },
+    {
+      "epoch": 0.1993607670795046,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1432,
+      "step": 499
+    },
+    {
+      "epoch": 0.1997602876548142,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1328,
+      "step": 500
+    },
+    {
+      "epoch": 0.20015980823012386,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1392,
+      "step": 501
+    },
+    {
+      "epoch": 0.20055932880543348,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 502
+    },
+    {
+      "epoch": 0.2009588493807431,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1478,
+      "step": 503
+    },
+    {
+      "epoch": 0.20135836995605275,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1399,
+      "step": 504
+    },
+    {
+      "epoch": 0.20175789053136237,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 505
+    },
+    {
+      "epoch": 0.20215741110667199,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1413,
+      "step": 506
+    },
+    {
+      "epoch": 0.20255693168198163,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 507
+    },
+    {
+      "epoch": 0.20295645225729125,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1337,
+      "step": 508
+    },
+    {
+      "epoch": 0.20335597283260087,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1325,
+      "step": 509
+    },
+    {
+      "epoch": 0.20375549340791052,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 510
+    },
+    {
+      "epoch": 0.20415501398322014,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1469,
+      "step": 511
+    },
+    {
+      "epoch": 0.20455453455852976,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 512
+    },
+    {
+      "epoch": 0.20495405513383938,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1358,
+      "step": 513
+    },
+    {
+      "epoch": 0.20535357570914903,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1423,
+      "step": 514
+    },
+    {
+      "epoch": 0.20575309628445865,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1375,
+      "step": 515
+    },
+    {
+      "epoch": 0.20615261685976827,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 516
+    },
+    {
+      "epoch": 0.20655213743507791,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 517
+    },
+    {
+      "epoch": 0.20695165801038753,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 518
+    },
+    {
+      "epoch": 0.20735117858569715,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 519
+    },
+    {
+      "epoch": 0.2077506991610068,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 520
+    },
+    {
+      "epoch": 0.20815021973631642,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 521
+    },
+    {
+      "epoch": 0.20854974031162604,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1218,
+      "step": 522
+    },
+    {
+      "epoch": 0.2089492608869357,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1369,
+      "step": 523
+    },
+    {
+      "epoch": 0.2093487814622453,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 524
+    },
+    {
+      "epoch": 0.20974830203755493,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 525
+    },
+    {
+      "epoch": 0.21014782261286455,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 526
+    },
+    {
+      "epoch": 0.2105473431881742,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 527
+    },
+    {
+      "epoch": 0.21094686376348382,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1422,
+      "step": 528
+    },
+    {
+      "epoch": 0.21134638433879344,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 529
+    },
+    {
+      "epoch": 0.21174590491410308,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1324,
+      "step": 530
+    },
+    {
+      "epoch": 0.2121454254894127,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1373,
+      "step": 531
+    },
+    {
+      "epoch": 0.21254494606472232,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 532
+    },
+    {
+      "epoch": 0.21294446664003197,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1275,
+      "step": 533
+    },
+    {
+      "epoch": 0.2133439872153416,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 534
+    },
+    {
+      "epoch": 0.2137435077906512,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 535
+    },
+    {
+      "epoch": 0.21414302836596086,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1393,
+      "step": 536
+    },
+    {
+      "epoch": 0.21454254894127048,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 537
+    },
+    {
+      "epoch": 0.2149420695165801,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.144,
+      "step": 538
+    },
+    {
+      "epoch": 0.21534159009188975,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1174,
+      "step": 539
+    },
+    {
+      "epoch": 0.21574111066719937,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.1477,
+      "step": 540
+    },
+    {
+      "epoch": 0.21614063124250898,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 541
+    },
+    {
+      "epoch": 0.2165401518178186,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1227,
+      "step": 542
+    },
+    {
+      "epoch": 0.21693967239312825,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 543
+    },
+    {
+      "epoch": 0.21733919296843787,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1385,
+      "step": 544
+    },
+    {
+      "epoch": 0.2177387135437475,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 545
+    },
+    {
+      "epoch": 0.21813823411905714,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 546
+    },
+    {
+      "epoch": 0.21853775469436676,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1442,
+      "step": 547
+    },
+    {
+      "epoch": 0.21893727526967638,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 548
+    },
+    {
+      "epoch": 0.21933679584498603,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 549
+    },
+    {
+      "epoch": 0.21973631642029565,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 550
+    },
+    {
+      "epoch": 0.22013583699560527,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 551
+    },
+    {
+      "epoch": 0.2205353575709149,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1384,
+      "step": 552
+    },
+    {
+      "epoch": 0.22093487814622453,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 553
+    },
+    {
+      "epoch": 0.22133439872153415,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 554
+    },
+    {
+      "epoch": 0.2217339192968438,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1501,
+      "step": 555
+    },
+    {
+      "epoch": 0.22213343987215342,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1299,
+      "step": 556
+    },
+    {
+      "epoch": 0.22253296044746304,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 557
+    },
+    {
+      "epoch": 0.22293248102277266,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005,
+      "loss": 1.1489,
+      "step": 558
+    },
+    {
+      "epoch": 0.2233320015980823,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 559
+    },
+    {
+      "epoch": 0.22373152217339193,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 560
+    },
+    {
+      "epoch": 0.22413104274870155,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 561
+    },
+    {
+      "epoch": 0.2245305633240112,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1329,
+      "step": 562
+    },
+    {
+      "epoch": 0.22493008389932082,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1349,
+      "step": 563
+    },
+    {
+      "epoch": 0.22532960447463043,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1348,
+      "step": 564
+    },
+    {
+      "epoch": 0.22572912504994008,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1405,
+      "step": 565
+    },
+    {
+      "epoch": 0.2261286456252497,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 566
+    },
+    {
+      "epoch": 0.22652816620055932,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1309,
+      "step": 567
+    },
+    {
+      "epoch": 0.22692768677586897,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 568
+    },
+    {
+      "epoch": 0.2273272073511786,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 569
+    },
+    {
+      "epoch": 0.2277267279264882,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 570
+    },
+    {
+      "epoch": 0.22812624850179783,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 571
+    },
+    {
+      "epoch": 0.22852576907710748,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.1154,
+      "step": 572
+    },
+    {
+      "epoch": 0.2289252896524171,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 573
+    },
+    {
+      "epoch": 0.22932481022772672,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.135,
+      "step": 574
+    },
+    {
+      "epoch": 0.22972433080303636,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 575
+    },
+    {
+      "epoch": 0.23012385137834598,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1201,
+      "step": 576
+    },
+    {
+      "epoch": 0.2305233719536556,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 577
+    },
+    {
+      "epoch": 0.23092289252896525,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1216,
+      "step": 578
+    },
+    {
+      "epoch": 0.23132241310427487,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 579
+    },
+    {
+      "epoch": 0.2317219336795845,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1421,
+      "step": 580
+    },
+    {
+      "epoch": 0.23212145425489414,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1248,
+      "step": 581
+    },
+    {
+      "epoch": 0.23252097483020376,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 582
+    },
+    {
+      "epoch": 0.23292049540551338,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 583
+    },
+    {
+      "epoch": 0.23332001598082303,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.13,
+      "step": 584
+    },
+    {
+      "epoch": 0.23371953655613265,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1497,
+      "step": 585
+    },
+    {
+      "epoch": 0.23411905713144227,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 586
+    },
+    {
+      "epoch": 0.23451857770675189,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 587
+    },
+    {
+      "epoch": 0.23491809828206153,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 588
+    },
+    {
+      "epoch": 0.23531761885737115,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1211,
+      "step": 589
+    },
+    {
+      "epoch": 0.23571713943268077,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1391,
+      "step": 590
+    },
+    {
+      "epoch": 0.23611666000799042,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 591
+    },
+    {
+      "epoch": 0.23651618058330004,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1238,
+      "step": 592
+    },
+    {
+      "epoch": 0.23691570115860966,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 593
+    },
+    {
+      "epoch": 0.2373152217339193,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.123,
+      "step": 594
+    },
+    {
+      "epoch": 0.23771474230922893,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.141,
+      "step": 595
+    },
+    {
+      "epoch": 0.23811426288453855,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1366,
+      "step": 596
+    },
+    {
+      "epoch": 0.2385137834598482,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1364,
+      "step": 597
+    },
+    {
+      "epoch": 0.23891330403515781,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1281,
+      "step": 598
+    },
+    {
+      "epoch": 0.23931282461046743,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 599
+    },
+    {
+      "epoch": 0.23971234518577705,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 600
+    },
+    {
+      "epoch": 0.2401118657610867,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1346,
+      "step": 601
+    },
+    {
+      "epoch": 0.24051138633639632,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 602
+    },
+    {
+      "epoch": 0.24091090691170594,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 603
+    },
+    {
+      "epoch": 0.2413104274870156,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1287,
+      "step": 604
+    },
+    {
+      "epoch": 0.2417099480623252,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.131,
+      "step": 605
+    },
+    {
+      "epoch": 0.24210946863763483,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 606
+    },
+    {
+      "epoch": 0.24250898921294448,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 607
+    },
+    {
+      "epoch": 0.2429085097882541,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1347,
+      "step": 608
+    },
+    {
+      "epoch": 0.24330803036356372,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1315,
+      "step": 609
+    },
+    {
+      "epoch": 0.24370755093887336,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1335,
+      "step": 610
+    },
+    {
+      "epoch": 0.24410707151418298,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1253,
+      "step": 611
+    },
+    {
+      "epoch": 0.2445065920894926,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 612
+    },
+    {
+      "epoch": 0.24490611266480225,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 613
+    },
+    {
+      "epoch": 0.24530563324011187,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1334,
+      "step": 614
+    },
+    {
+      "epoch": 0.2457051538154215,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1246,
+      "step": 615
+    },
+    {
+      "epoch": 0.2461046743907311,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 616
+    },
+    {
+      "epoch": 0.24650419496604076,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1424,
+      "step": 617
+    },
+    {
+      "epoch": 0.24690371554135038,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1149,
+      "step": 618
+    },
+    {
+      "epoch": 0.24730323611666,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 619
+    },
+    {
+      "epoch": 0.24770275669196964,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 620
+    },
+    {
+      "epoch": 0.24810227726727926,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1306,
+      "step": 621
+    },
+    {
+      "epoch": 0.24850179784258888,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1264,
+      "step": 622
+    },
+    {
+      "epoch": 0.24890131841789853,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1187,
+      "step": 623
+    },
+    {
+      "epoch": 0.24930083899320815,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1114,
+      "step": 624
+    },
+    {
+      "epoch": 0.24970035956851777,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 625
+    },
+    {
+      "epoch": 0.2500998801438274,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1336,
+      "step": 626
+    },
+    {
+      "epoch": 0.250499400719137,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.1356,
+      "step": 627
+    },
+    {
+      "epoch": 0.2508989212944467,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 628
+    },
+    {
+      "epoch": 0.2512984418697563,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1232,
+      "step": 629
+    },
+    {
+      "epoch": 0.2516979624450659,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1225,
+      "step": 630
+    },
+    {
+      "epoch": 0.25209748302037555,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 631
+    },
+    {
+      "epoch": 0.25249700359568517,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 632
+    },
+    {
+      "epoch": 0.2528965241709948,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1255,
+      "step": 633
+    },
+    {
+      "epoch": 0.25329604474630446,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1247,
+      "step": 634
+    },
+    {
+      "epoch": 0.2536955653216141,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1361,
+      "step": 635
+    },
+    {
+      "epoch": 0.2540950858969237,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1327,
+      "step": 636
+    },
+    {
+      "epoch": 0.2544946064722333,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.134,
+      "step": 637
+    },
+    {
+      "epoch": 0.25489412704754294,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 638
+    },
+    {
+      "epoch": 0.25529364762285256,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 639
+    },
+    {
+      "epoch": 0.2556931681981622,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1386,
+      "step": 640
+    },
+    {
+      "epoch": 0.25609268877347185,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1331,
+      "step": 641
+    },
+    {
+      "epoch": 0.2564922093487815,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1305,
+      "step": 642
+    },
+    {
+      "epoch": 0.2568917299240911,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 643
+    },
+    {
+      "epoch": 0.2572912504994007,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1218,
+      "step": 644
+    },
+    {
+      "epoch": 0.25769077107471033,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1406,
+      "step": 645
+    },
+    {
+      "epoch": 0.25809029165001995,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1359,
+      "step": 646
+    },
+    {
+      "epoch": 0.25848981222532963,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 647
+    },
+    {
+      "epoch": 0.25888933280063925,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1351,
+      "step": 648
+    },
+    {
+      "epoch": 0.25928885337594887,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.132,
+      "step": 649
+    },
+    {
+      "epoch": 0.2596883739512585,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1172,
+      "step": 650
+    },
+    {
+      "epoch": 0.2600878945265681,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1223,
+      "step": 651
+    },
+    {
+      "epoch": 0.26048741510187773,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1121,
+      "step": 652
+    },
+    {
+      "epoch": 0.26088693567718735,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1363,
+      "step": 653
+    },
+    {
+      "epoch": 0.261286456252497,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.127,
+      "step": 654
+    },
+    {
+      "epoch": 0.26168597682780664,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1282,
+      "step": 655
+    },
+    {
+      "epoch": 0.26208549740311626,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1261,
+      "step": 656
+    },
+    {
+      "epoch": 0.2624850179784259,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1165,
+      "step": 657
+    },
+    {
+      "epoch": 0.2628845385537355,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1372,
+      "step": 658
+    },
+    {
+      "epoch": 0.2632840591290451,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 659
+    },
+    {
+      "epoch": 0.2636835797043548,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 660
+    },
+    {
+      "epoch": 0.2640831002796644,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 661
+    },
+    {
+      "epoch": 0.26448262085497404,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1355,
+      "step": 662
+    },
+    {
+      "epoch": 0.26488214143028366,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.124,
+      "step": 663
+    },
+    {
+      "epoch": 0.2652816620055933,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.115,
+      "step": 664
+    },
+    {
+      "epoch": 0.2656811825809029,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 665
+    },
+    {
+      "epoch": 0.2660807031562126,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1194,
+      "step": 666
+    },
+    {
+      "epoch": 0.2664802237315222,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1218,
+      "step": 667
+    },
+    {
+      "epoch": 0.2668797443068318,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1148,
+      "step": 668
+    },
+    {
+      "epoch": 0.26727926488214143,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 669
+    },
+    {
+      "epoch": 0.26767878545745105,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 670
+    },
+    {
+      "epoch": 0.26807830603276067,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1332,
+      "step": 671
+    },
+    {
+      "epoch": 0.2684778266080703,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1139,
+      "step": 672
+    },
+    {
+      "epoch": 0.26887734718337997,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1394,
+      "step": 673
+    },
+    {
+      "epoch": 0.2692768677586896,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1202,
+      "step": 674
+    },
+    {
+      "epoch": 0.2696763883339992,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1184,
+      "step": 675
+    },
+    {
+      "epoch": 0.2700759089093088,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 676
+    },
+    {
+      "epoch": 0.27047542948461845,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 677
+    },
+    {
+      "epoch": 0.27087495005992807,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1219,
+      "step": 678
+    },
+    {
+      "epoch": 0.27127447063523774,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1241,
+      "step": 679
+    },
+    {
+      "epoch": 0.27167399121054736,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1102,
+      "step": 680
+    },
+    {
+      "epoch": 0.272073511785857,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1313,
+      "step": 681
+    },
+    {
+      "epoch": 0.2724730323611666,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1362,
+      "step": 682
+    },
+    {
+      "epoch": 0.2728725529364762,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1326,
+      "step": 683
+    },
+    {
+      "epoch": 0.27327207351178584,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 684
+    },
+    {
+      "epoch": 0.27367159408709546,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 685
+    },
+    {
+      "epoch": 0.27407111466240514,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1239,
+      "step": 686
+    },
+    {
+      "epoch": 0.27447063523771476,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1283,
+      "step": 687
+    },
+    {
+      "epoch": 0.2748701558130244,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1168,
+      "step": 688
+    },
+    {
+      "epoch": 0.275269676388334,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 689
+    },
+    {
+      "epoch": 0.2756691969636436,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1296,
+      "step": 690
+    },
+    {
+      "epoch": 0.27606871753895323,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1217,
+      "step": 691
+    },
+    {
+      "epoch": 0.2764682381142629,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1094,
+      "step": 692
+    },
+    {
+      "epoch": 0.27686775868957253,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1169,
+      "step": 693
+    },
+    {
+      "epoch": 0.27726727926488215,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 694
+    },
+    {
+      "epoch": 0.27766679984019177,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1377,
+      "step": 695
+    },
+    {
+      "epoch": 0.2780663204155014,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1182,
+      "step": 696
+    },
+    {
+      "epoch": 0.278465840990811,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1103,
+      "step": 697
+    },
+    {
+      "epoch": 0.27886536156612063,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 698
+    },
+    {
+      "epoch": 0.2792648821414303,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 699
+    },
+    {
+      "epoch": 0.2796644027167399,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 700
+    },
+    {
+      "epoch": 0.28006392329204954,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1206,
+      "step": 701
+    },
+    {
+      "epoch": 0.28046344386735916,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 702
+    },
+    {
+      "epoch": 0.2808629644426688,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1101,
+      "step": 703
+    },
+    {
+      "epoch": 0.2812624850179784,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 704
+    },
+    {
+      "epoch": 0.2816620055932881,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1226,
+      "step": 705
+    },
+    {
+      "epoch": 0.2820615261685977,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1168,
+      "step": 706
+    },
+    {
+      "epoch": 0.2824610467439073,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1158,
+      "step": 707
+    },
+    {
+      "epoch": 0.28286056731921694,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1119,
+      "step": 708
+    },
+    {
+      "epoch": 0.28326008789452656,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 709
+    },
+    {
+      "epoch": 0.2836596084698362,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 710
+    },
+    {
+      "epoch": 0.2840591290451458,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1153,
+      "step": 711
+    },
+    {
+      "epoch": 0.2844586496204555,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1101,
+      "step": 712
+    },
+    {
+      "epoch": 0.2848581701957651,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 713
+    },
+    {
+      "epoch": 0.2852576907710747,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 714
+    },
+    {
+      "epoch": 0.28565721134638433,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1138,
+      "step": 715
+    },
+    {
+      "epoch": 0.28605673192169395,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1177,
+      "step": 716
+    },
+    {
+      "epoch": 0.28645625249700357,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1249,
+      "step": 717
+    },
+    {
+      "epoch": 0.28685577307231325,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1251,
+      "step": 718
+    },
+    {
+      "epoch": 0.28725529364762287,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1244,
+      "step": 719
+    },
+    {
+      "epoch": 0.2876548142229325,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1146,
+      "step": 720
+    },
+    {
+      "epoch": 0.2880543347982421,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1307,
+      "step": 721
+    },
+    {
+      "epoch": 0.2884538553735517,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1092,
+      "step": 722
+    },
+    {
+      "epoch": 0.28885337594886135,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1134,
+      "step": 723
+    },
+    {
+      "epoch": 0.289252896524171,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1236,
+      "step": 724
+    },
+    {
+      "epoch": 0.28965241709948064,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1193,
+      "step": 725
+    },
+    {
+      "epoch": 0.29005193767479026,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 726
+    },
+    {
+      "epoch": 0.2904514582500999,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 727
+    },
+    {
+      "epoch": 0.2908509788254095,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1203,
+      "step": 728
+    },
+    {
+      "epoch": 0.2912504994007191,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1312,
+      "step": 729
+    },
+    {
+      "epoch": 0.29165001997602874,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1167,
+      "step": 730
+    },
+    {
+      "epoch": 0.2920495405513384,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1268,
+      "step": 731
+    },
+    {
+      "epoch": 0.29244906112664804,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 732
+    },
+    {
+      "epoch": 0.29284858170195766,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1234,
+      "step": 733
+    },
+    {
+      "epoch": 0.2932481022772673,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1292,
+      "step": 734
+    },
+    {
+      "epoch": 0.2936476228525769,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1626,
+      "step": 735
+    },
+    {
+      "epoch": 0.2940471434278865,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1371,
+      "step": 736
+    },
+    {
+      "epoch": 0.2944466640031962,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.138,
+      "step": 737
+    },
+    {
+      "epoch": 0.2948461845785058,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1269,
+      "step": 738
+    },
+    {
+      "epoch": 0.29524570515381543,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1353,
+      "step": 739
+    },
+    {
+      "epoch": 0.29564522572912505,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1402,
+      "step": 740
+    },
+    {
+      "epoch": 0.29604474630443467,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1354,
+      "step": 741
+    },
+    {
+      "epoch": 0.2964442668797443,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1138,
+      "step": 742
+    },
+    {
+      "epoch": 0.2968437874550539,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1408,
+      "step": 743
+    },
+    {
+      "epoch": 0.2972433080303636,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1339,
+      "step": 744
+    },
+    {
+      "epoch": 0.2976428286056732,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 745
+    },
+    {
+      "epoch": 0.2980423491809828,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1109,
+      "step": 746
+    },
+    {
+      "epoch": 0.29844186975629244,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 747
+    },
+    {
+      "epoch": 0.29884139033160206,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1201,
+      "step": 748
+    },
+    {
+      "epoch": 0.2992409109069117,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1236,
+      "step": 749
+    },
+    {
+      "epoch": 0.29964043148222136,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 750
+    },
+    {
+      "epoch": 0.300039952057531,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1265,
+      "step": 751
+    },
+    {
+      "epoch": 0.3004394726328406,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1172,
+      "step": 752
+    },
+    {
+      "epoch": 0.3008389932081502,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1404,
+      "step": 753
+    },
+    {
+      "epoch": 0.30123851378345984,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1316,
+      "step": 754
+    },
+    {
+      "epoch": 0.30163803435876946,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1273,
+      "step": 755
+    },
+    {
+      "epoch": 0.3020375549340791,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1202,
+      "step": 756
+    },
+    {
+      "epoch": 0.30243707550938875,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1228,
+      "step": 757
+    },
+    {
+      "epoch": 0.3028365960846984,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.119,
+      "step": 758
+    },
+    {
+      "epoch": 0.303236116660008,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.119,
+      "step": 759
+    },
+    {
+      "epoch": 0.3036356372353176,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1221,
+      "step": 760
+    },
+    {
+      "epoch": 0.30403515781062723,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1136,
+      "step": 761
+    },
+    {
+      "epoch": 0.30443467838593685,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1258,
+      "step": 762
+    },
+    {
+      "epoch": 0.3048341989612465,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1278,
+      "step": 763
+    },
+    {
+      "epoch": 0.30523371953655615,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1177,
+      "step": 764
+    },
+    {
+      "epoch": 0.30563324011186577,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1233,
+      "step": 765
+    },
+    {
+      "epoch": 0.3060327606871754,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1077,
+      "step": 766
+    },
+    {
+      "epoch": 0.306432281262485,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1229,
+      "step": 767
+    },
+    {
+      "epoch": 0.3068318018377946,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1112,
+      "step": 768
+    },
+    {
+      "epoch": 0.3072313224131043,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1127,
+      "step": 769
+    },
+    {
+      "epoch": 0.3076308429884139,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1207,
+      "step": 770
+    },
+    {
+      "epoch": 0.30803036356372354,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1061,
+      "step": 771
+    },
+    {
+      "epoch": 0.30842988413903316,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1155,
+      "step": 772
+    },
+    {
+      "epoch": 0.3088294047143428,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1171,
+      "step": 773
+    },
+    {
+      "epoch": 0.3092289252896524,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1338,
+      "step": 774
+    },
+    {
+      "epoch": 0.309628445864962,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1294,
+      "step": 775
+    },
+    {
+      "epoch": 0.3100279664402717,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1079,
+      "step": 776
+    },
+    {
+      "epoch": 0.3104274870155813,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 777
+    },
+    {
+      "epoch": 0.31082700759089094,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1129,
+      "step": 778
+    },
+    {
+      "epoch": 0.31122652816620056,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1111,
+      "step": 779
+    },
+    {
+      "epoch": 0.3116260487415102,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1093,
+      "step": 780
+    },
+    {
+      "epoch": 0.3120255693168198,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1203,
+      "step": 781
+    },
+    {
+      "epoch": 0.31242508989212947,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.112,
+      "step": 782
+    },
+    {
+      "epoch": 0.3128246104674391,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1165,
+      "step": 783
+    },
+    {
+      "epoch": 0.3132241310427487,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1284,
+      "step": 784
+    },
+    {
+      "epoch": 0.31362365161805833,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1089,
+      "step": 785
+    },
+    {
+      "epoch": 0.31402317219336795,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1198,
+      "step": 786
+    },
+    {
+      "epoch": 0.31442269276867757,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.128,
+      "step": 787
+    },
+    {
+      "epoch": 0.3148222133439872,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 788
+    },
+    {
+      "epoch": 0.31522173391929686,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1155,
+      "step": 789
+    },
+    {
+      "epoch": 0.3156212544946065,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1202,
+      "step": 790
+    },
+    {
+      "epoch": 0.3160207750699161,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1068,
+      "step": 791
+    },
+    {
+      "epoch": 0.3164202956452257,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1252,
+      "step": 792
+    },
+    {
+      "epoch": 0.31681981622053534,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1064,
+      "step": 793
+    },
+    {
+      "epoch": 0.31721933679584496,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1271,
+      "step": 794
+    },
+    {
+      "epoch": 0.31761885737115464,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.113,
+      "step": 795
+    },
+    {
+      "epoch": 0.31801837794646426,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1096,
+      "step": 796
+    },
+    {
+      "epoch": 0.3184178985217739,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1106,
+      "step": 797
+    },
+    {
+      "epoch": 0.3188174190970835,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1102,
+      "step": 798
+    },
+    {
+      "epoch": 0.3192169396723931,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1239,
+      "step": 799
+    },
+    {
+      "epoch": 0.31961646024770274,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 800
+    },
+    {
+      "epoch": 0.32001598082301236,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1133,
+      "step": 801
+    },
+    {
+      "epoch": 0.32041550139832203,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 802
+    },
+    {
+      "epoch": 0.32081502197363165,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1266,
+      "step": 803
+    },
+    {
+      "epoch": 0.3212145425489413,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1072,
+      "step": 804
+    },
+    {
+      "epoch": 0.3216140631242509,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.113,
+      "step": 805
+    },
+    {
+      "epoch": 0.3220135836995605,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1033,
+      "step": 806
+    },
+    {
+      "epoch": 0.32241310427487013,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1226,
+      "step": 807
+    },
+    {
+      "epoch": 0.3228126248501798,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1169,
+      "step": 808
+    },
+    {
+      "epoch": 0.3232121454254894,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1078,
+      "step": 809
+    },
+    {
+      "epoch": 0.32361166600079905,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1135,
+      "step": 810
+    },
+    {
+      "epoch": 0.32401118657610867,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1145,
+      "step": 811
+    },
+    {
+      "epoch": 0.3244107071514183,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 812
+    },
+    {
+      "epoch": 0.3248102277267279,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.121,
+      "step": 813
+    },
+    {
+      "epoch": 0.3252097483020376,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 814
+    },
+    {
+      "epoch": 0.3256092688773472,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1085,
+      "step": 815
+    },
+    {
+      "epoch": 0.3260087894526568,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1207,
+      "step": 816
+    },
+    {
+      "epoch": 0.32640831002796644,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1218,
+      "step": 817
+    },
+    {
+      "epoch": 0.32680783060327606,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1157,
+      "step": 818
+    },
+    {
+      "epoch": 0.3272073511785857,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1155,
+      "step": 819
+    },
+    {
+      "epoch": 0.3276068717538953,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.108,
+      "step": 820
+    },
+    {
+      "epoch": 0.328006392329205,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1131,
+      "step": 821
+    },
+    {
+      "epoch": 0.3284059129045146,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1157,
+      "step": 822
+    },
+    {
+      "epoch": 0.3288054334798242,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1193,
+      "step": 823
+    },
+    {
+      "epoch": 0.32920495405513384,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.114,
+      "step": 824
+    },
+    {
+      "epoch": 0.32960447463044346,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1106,
+      "step": 825
+    },
+    {
+      "epoch": 0.3300039952057531,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1167,
+      "step": 826
+    },
+    {
+      "epoch": 0.33040351578106275,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1092,
+      "step": 827
+    },
+    {
+      "epoch": 0.33080303635637237,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1154,
+      "step": 828
+    },
+    {
+      "epoch": 0.331202556931682,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1141,
+      "step": 829
+    },
+    {
+      "epoch": 0.3316020775069916,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1112,
+      "step": 830
+    },
+    {
+      "epoch": 0.33200159808230123,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1048,
+      "step": 831
+    },
+    {
+      "epoch": 0.33240111865761085,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1215,
+      "step": 832
+    },
+    {
+      "epoch": 0.33280063923292047,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.122,
+      "step": 833
+    },
+    {
+      "epoch": 0.33320015980823015,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1039,
+      "step": 834
+    },
+    {
+      "epoch": 0.33359968038353977,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1194,
+      "step": 835
+    },
+    {
+      "epoch": 0.3339992009588494,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 836
+    },
+    {
+      "epoch": 0.334398721534159,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1022,
+      "step": 837
+    },
+    {
+      "epoch": 0.3347982421094686,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1134,
+      "step": 838
+    },
+    {
+      "epoch": 0.33519776268477824,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 839
+    },
+    {
+      "epoch": 0.3355972832600879,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0934,
+      "step": 840
+    },
+    {
+      "epoch": 0.33599680383539754,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.115,
+      "step": 841
+    },
+    {
+      "epoch": 0.33639632441070716,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 842
+    },
+    {
+      "epoch": 0.3367958449860168,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1192,
+      "step": 843
+    },
+    {
+      "epoch": 0.3371953655613264,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1116,
+      "step": 844
+    },
+    {
+      "epoch": 0.337594886136636,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0995,
+      "step": 845
+    },
+    {
+      "epoch": 0.33799440671194564,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1177,
+      "step": 846
+    },
+    {
+      "epoch": 0.3383939272872553,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1161,
+      "step": 847
+    },
+    {
+      "epoch": 0.33879344786256493,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1124,
+      "step": 848
+    },
+    {
+      "epoch": 0.33919296843787455,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1089,
+      "step": 849
+    },
+    {
+      "epoch": 0.3395924890131842,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1224,
+      "step": 850
+    },
+    {
+      "epoch": 0.3399920095884938,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1175,
+      "step": 851
+    },
+    {
+      "epoch": 0.3403915301638034,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1012,
+      "step": 852
+    },
+    {
+      "epoch": 0.3407910507391131,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1187,
+      "step": 853
+    },
+    {
+      "epoch": 0.3411905713144227,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1145,
+      "step": 854
+    },
+    {
+      "epoch": 0.34159009188973233,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 855
+    },
+    {
+      "epoch": 0.34198961246504195,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 856
+    },
+    {
+      "epoch": 0.34238913304035157,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1092,
+      "step": 857
+    },
+    {
+      "epoch": 0.3427886536156612,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 858
+    },
+    {
+      "epoch": 0.34318817419097086,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1124,
+      "step": 859
+    },
+    {
+      "epoch": 0.3435876947662805,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1176,
+      "step": 860
+    },
+    {
+      "epoch": 0.3439872153415901,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.108,
+      "step": 861
+    },
+    {
+      "epoch": 0.3443867359168997,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1174,
+      "step": 862
+    },
+    {
+      "epoch": 0.34478625649220934,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 863
+    },
+    {
+      "epoch": 0.34518577706751896,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1434,
+      "step": 864
+    },
+    {
+      "epoch": 0.3455852976428286,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1176,
+      "step": 865
+    },
+    {
+      "epoch": 0.34598481821813826,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1277,
+      "step": 866
+    },
+    {
+      "epoch": 0.3463843387934479,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1293,
+      "step": 867
+    },
+    {
+      "epoch": 0.3467838593687575,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1049,
+      "step": 868
+    },
+    {
+      "epoch": 0.3471833799440671,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.136,
+      "step": 869
+    },
+    {
+      "epoch": 0.34758290051937674,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1131,
+      "step": 870
+    },
+    {
+      "epoch": 0.34798242109468636,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1289,
+      "step": 871
+    },
+    {
+      "epoch": 0.34838194166999603,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1295,
+      "step": 872
+    },
+    {
+      "epoch": 0.34878146224530565,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.117,
+      "step": 873
+    },
+    {
+      "epoch": 0.34918098282061527,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.114,
+      "step": 874
+    },
+    {
+      "epoch": 0.3495805033959249,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1288,
+      "step": 875
+    },
+    {
+      "epoch": 0.3499800239712345,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.115,
+      "step": 876
+    },
+    {
+      "epoch": 0.35037954454654413,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1176,
+      "step": 877
+    },
+    {
+      "epoch": 0.35077906512185375,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1187,
+      "step": 878
+    },
+    {
+      "epoch": 0.3511785856971634,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1207,
+      "step": 879
+    },
+    {
+      "epoch": 0.35157810627247305,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1152,
+      "step": 880
+    },
+    {
+      "epoch": 0.35197762684778267,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.116,
+      "step": 881
+    },
+    {
+      "epoch": 0.3523771474230923,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1172,
+      "step": 882
+    },
+    {
+      "epoch": 0.3527766679984019,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1058,
+      "step": 883
+    },
+    {
+      "epoch": 0.3531761885737115,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1245,
+      "step": 884
+    },
+    {
+      "epoch": 0.3535757091490212,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1213,
+      "step": 885
+    },
+    {
+      "epoch": 0.3539752297243308,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1153,
+      "step": 886
+    },
+    {
+      "epoch": 0.35437475029964044,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 887
+    },
+    {
+      "epoch": 0.35477427087495006,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1298,
+      "step": 888
+    },
+    {
+      "epoch": 0.3551737914502597,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1091,
+      "step": 889
+    },
+    {
+      "epoch": 0.3555733120255693,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1128,
+      "step": 890
+    },
+    {
+      "epoch": 0.3559728326008789,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.104,
+      "step": 891
+    },
+    {
+      "epoch": 0.3563723531761886,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1159,
+      "step": 892
+    },
+    {
+      "epoch": 0.3567718737514982,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1067,
+      "step": 893
+    },
+    {
+      "epoch": 0.35717139432680783,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1072,
+      "step": 894
+    },
+    {
+      "epoch": 0.35757091490211745,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1018,
+      "step": 895
+    },
+    {
+      "epoch": 0.3579704354774271,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1174,
+      "step": 896
+    },
+    {
+      "epoch": 0.3583699560527367,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1152,
+      "step": 897
+    },
+    {
+      "epoch": 0.35876947662804637,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.11,
+      "step": 898
+    },
+    {
+      "epoch": 0.359168997203356,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.1139,
+      "step": 899
+    },
+    {
+      "epoch": 0.3595685177786656,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1054,
+      "step": 900
+    },
+    {
+      "epoch": 0.35996803835397523,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1211,
+      "step": 901
+    },
+    {
+      "epoch": 0.36036755892928485,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1226,
+      "step": 902
+    },
+    {
+      "epoch": 0.36076707950459447,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1124,
+      "step": 903
+    },
+    {
+      "epoch": 0.3611666000799041,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1111,
+      "step": 904
+    },
+    {
+      "epoch": 0.36156612065521376,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1186,
+      "step": 905
+    },
+    {
+      "epoch": 0.3619656412305234,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1333,
+      "step": 906
+    },
+    {
+      "epoch": 0.362365161805833,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1039,
+      "step": 907
+    },
+    {
+      "epoch": 0.3627646823811426,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.111,
+      "step": 908
+    },
+    {
+      "epoch": 0.36316420295645224,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 909
+    },
+    {
+      "epoch": 0.36356372353176186,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.12,
+      "step": 910
+    },
+    {
+      "epoch": 0.36396324410707154,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1181,
+      "step": 911
+    },
+    {
+      "epoch": 0.36436276468238116,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1208,
+      "step": 912
+    },
+    {
+      "epoch": 0.3647622852576908,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1053,
+      "step": 913
+    },
+    {
+      "epoch": 0.3651618058330004,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1272,
+      "step": 914
+    },
+    {
+      "epoch": 0.36556132640831,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 915
+    },
+    {
+      "epoch": 0.36596084698361964,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1091,
+      "step": 916
+    },
+    {
+      "epoch": 0.3663603675589293,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1176,
+      "step": 917
+    },
+    {
+      "epoch": 0.36675988813423893,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1093,
+      "step": 918
+    },
+    {
+      "epoch": 0.36715940870954855,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1085,
+      "step": 919
+    },
+    {
+      "epoch": 0.36755892928485817,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1154,
+      "step": 920
+    },
+    {
+      "epoch": 0.3679584498601678,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1095,
+      "step": 921
+    },
+    {
+      "epoch": 0.3683579704354774,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1023,
+      "step": 922
+    },
+    {
+      "epoch": 0.36875749101078703,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.117,
+      "step": 923
+    },
+    {
+      "epoch": 0.3691570115860967,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1042,
+      "step": 924
+    },
+    {
+      "epoch": 0.3695565321614063,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1231,
+      "step": 925
+    },
+    {
+      "epoch": 0.36995605273671595,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1149,
+      "step": 926
+    },
+    {
+      "epoch": 0.37035557331202557,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.109,
+      "step": 927
+    },
+    {
+      "epoch": 0.3707550938873352,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1165,
+      "step": 928
+    },
+    {
+      "epoch": 0.3711546144626448,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.126,
+      "step": 929
+    },
+    {
+      "epoch": 0.3715541350379545,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1044,
+      "step": 930
+    },
+    {
+      "epoch": 0.3719536556132641,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.112,
+      "step": 931
+    },
+    {
+      "epoch": 0.3723531761885737,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1212,
+      "step": 932
+    },
+    {
+      "epoch": 0.37275269676388334,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1152,
+      "step": 933
+    },
+    {
+      "epoch": 0.37315221733919296,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1004,
+      "step": 934
+    },
+    {
+      "epoch": 0.3735517379145026,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.115,
+      "step": 935
+    },
+    {
+      "epoch": 0.3739512584898122,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1087,
+      "step": 936
+    },
+    {
+      "epoch": 0.3743507790651219,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1125,
+      "step": 937
+    },
+    {
+      "epoch": 0.3747502996404315,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.106,
+      "step": 938
+    },
+    {
+      "epoch": 0.3751498202157411,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.105,
+      "step": 939
+    },
+    {
+      "epoch": 0.37554934079105073,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1088,
+      "step": 940
+    },
+    {
+      "epoch": 0.37594886136636035,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1079,
+      "step": 941
+    },
+    {
+      "epoch": 0.37634838194167,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1259,
+      "step": 942
+    },
+    {
+      "epoch": 0.37674790251697965,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1178,
+      "step": 943
+    },
+    {
+      "epoch": 0.37714742309228927,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1171,
+      "step": 944
+    },
+    {
+      "epoch": 0.3775469436675989,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1141,
+      "step": 945
+    },
+    {
+      "epoch": 0.3779464642429085,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1031,
+      "step": 946
+    },
+    {
+      "epoch": 0.37834598481821813,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1032,
+      "step": 947
+    },
+    {
+      "epoch": 0.37874550539352775,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1065,
+      "step": 948
+    },
+    {
+      "epoch": 0.37914502596883737,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 949
+    },
+    {
+      "epoch": 0.37954454654414704,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1033,
+      "step": 950
+    },
+    {
+      "epoch": 0.37994406711945666,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1065,
+      "step": 951
+    },
+    {
+      "epoch": 0.3803435876947663,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.0993,
+      "step": 952
+    },
+    {
+      "epoch": 0.3807431082700759,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1102,
+      "step": 953
+    },
+    {
+      "epoch": 0.3811426288453855,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1112,
+      "step": 954
+    },
+    {
+      "epoch": 0.38154214942069514,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1267,
+      "step": 955
+    },
+    {
+      "epoch": 0.3819416699960048,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1136,
+      "step": 956
+    },
+    {
+      "epoch": 0.38234119057131444,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1153,
+      "step": 957
+    },
+    {
+      "epoch": 0.38274071114662406,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1176,
+      "step": 958
+    },
+    {
+      "epoch": 0.3831402317219337,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1032,
+      "step": 959
+    },
+    {
+      "epoch": 0.3835397522972433,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.098,
+      "step": 960
+    },
+    {
+      "epoch": 0.3839392728725529,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1079,
+      "step": 961
+    },
+    {
+      "epoch": 0.3843387934478626,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1118,
+      "step": 962
+    },
+    {
+      "epoch": 0.3847383140231722,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.11,
+      "step": 963
+    },
+    {
+      "epoch": 0.38513783459848183,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.114,
+      "step": 964
+    },
+    {
+      "epoch": 0.38553735517379145,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 965
+    },
+    {
+      "epoch": 0.38593687574910107,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1117,
+      "step": 966
+    },
+    {
+      "epoch": 0.3863363963244107,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1025,
+      "step": 967
+    },
+    {
+      "epoch": 0.3867359168997203,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.105,
+      "step": 968
+    },
+    {
+      "epoch": 0.38713543747503,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0999,
+      "step": 969
+    },
+    {
+      "epoch": 0.3875349580503396,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1003,
+      "step": 970
+    },
+    {
+      "epoch": 0.3879344786256492,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1154,
+      "step": 971
+    },
+    {
+      "epoch": 0.38833399920095885,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1136,
+      "step": 972
+    },
+    {
+      "epoch": 0.38873351977626847,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1006,
+      "step": 973
+    },
+    {
+      "epoch": 0.3891330403515781,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 974
+    },
+    {
+      "epoch": 0.38953256092688776,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1163,
+      "step": 975
+    },
+    {
+      "epoch": 0.3899320815021974,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1066,
+      "step": 976
+    },
+    {
+      "epoch": 0.390331602077507,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 977
+    },
+    {
+      "epoch": 0.3907311226528166,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1129,
+      "step": 978
+    },
+    {
+      "epoch": 0.39113064322812624,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0989,
+      "step": 979
+    },
+    {
+      "epoch": 0.39153016380343586,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 980
+    },
+    {
+      "epoch": 0.3919296843787455,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1098,
+      "step": 981
+    },
+    {
+      "epoch": 0.39232920495405516,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1177,
+      "step": 982
+    },
+    {
+      "epoch": 0.3927287255293648,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1183,
+      "step": 983
+    },
+    {
+      "epoch": 0.3931282461046744,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 984
+    },
+    {
+      "epoch": 0.393527766679984,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1094,
+      "step": 985
+    },
+    {
+      "epoch": 0.39392728725529363,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1243,
+      "step": 986
+    },
+    {
+      "epoch": 0.39432680783060325,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1172,
+      "step": 987
+    },
+    {
+      "epoch": 0.39472632840591293,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1021,
+      "step": 988
+    },
+    {
+      "epoch": 0.39512584898122255,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1192,
+      "step": 989
+    },
+    {
+      "epoch": 0.39552536955653217,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.11,
+      "step": 990
+    },
+    {
+      "epoch": 0.3959248901318418,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1117,
+      "step": 991
+    },
+    {
+      "epoch": 0.3963244107071514,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1064,
+      "step": 992
+    },
+    {
+      "epoch": 0.39672393128246103,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 993
+    },
+    {
+      "epoch": 0.39712345185777065,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 994
+    },
+    {
+      "epoch": 0.3975229724330803,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1062,
+      "step": 995
+    },
+    {
+      "epoch": 0.39792249300838994,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1056,
+      "step": 996
+    },
+    {
+      "epoch": 0.39832201358369956,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1188,
+      "step": 997
+    },
+    {
+      "epoch": 0.3987215341590092,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1159,
+      "step": 998
+    },
+    {
+      "epoch": 0.3991210547343188,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.117,
+      "step": 999
+    },
+    {
+      "epoch": 0.3995205753096284,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1096,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3999200958849381,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0965,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4003196164602477,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1007,
+      "step": 1002
+    },
+    {
+      "epoch": 0.40071913703555734,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1099,
+      "step": 1003
+    },
+    {
+      "epoch": 0.40111865761086696,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.11,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4015181781861766,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1132,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4019176987614862,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1055,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4023172193367959,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1181,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4027167399121055,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1068,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4031162604874151,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.114,
+      "step": 1009
+    },
+    {
+      "epoch": 0.40351578106272473,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1061,
+      "step": 1010
+    },
+    {
+      "epoch": 0.40391530163803435,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1111,
+      "step": 1011
+    },
+    {
+      "epoch": 0.40431482221334397,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1056,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4047143427886536,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1122,
+      "step": 1013
+    },
+    {
+      "epoch": 0.40511386336396327,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.103,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4055133839392729,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4059129045145825,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0958,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4063124250898921,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.113,
+      "step": 1017
+    },
+    {
+      "epoch": 0.40671194566520175,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1144,
+      "step": 1018
+    },
+    {
+      "epoch": 0.40711146624051137,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 1019
+    },
+    {
+      "epoch": 0.40751098681582104,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1020
+    },
+    {
+      "epoch": 0.40791050739113066,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1152,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4083100279664403,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1046,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4087095485417499,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1062,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4091090691170595,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1174,
+      "step": 1024
+    },
+    {
+      "epoch": 0.40950858969236914,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0985,
+      "step": 1025
+    },
+    {
+      "epoch": 0.40990811026767876,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.107,
+      "step": 1026
+    },
+    {
+      "epoch": 0.41030763084298844,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1027
+    },
+    {
+      "epoch": 0.41070715141829806,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1137,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4111066719936077,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1065,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4115061925689173,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1051,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4119057131442269,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0995,
+      "step": 1031
+    },
+    {
+      "epoch": 0.41230523371953653,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1028,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4127047542948462,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1069,
+      "step": 1033
+    },
+    {
+      "epoch": 0.41310427487015583,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1155,
+      "step": 1034
+    },
+    {
+      "epoch": 0.41350379544546545,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1184,
+      "step": 1035
+    },
+    {
+      "epoch": 0.41390331602077507,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1085,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4143028365960847,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1053,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4147023571713943,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.111,
+      "step": 1038
+    },
+    {
+      "epoch": 0.41510187774670393,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1093,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4155013983220136,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0993,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4159009188973232,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1076,
+      "step": 1041
+    },
+    {
+      "epoch": 0.41630043947263284,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.105,
+      "step": 1042
+    },
+    {
+      "epoch": 0.41669996004794246,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1077,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4170994806232521,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1102,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4174990011985617,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4178985217738714,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.102,
+      "step": 1046
+    },
+    {
+      "epoch": 0.418298042349181,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1051,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4186975629244906,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0991,
+      "step": 1048
+    },
+    {
+      "epoch": 0.41909708349980024,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1085,
+      "step": 1049
+    },
+    {
+      "epoch": 0.41949660407510986,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.107,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4198961246504195,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.1038,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4202956452257291,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1047,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4206951658010388,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4210946863763484,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1119,
+      "step": 1054
+    },
+    {
+      "epoch": 0.421494206951658,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.106,
+      "step": 1055
+    },
+    {
+      "epoch": 0.42189372752696763,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1006,
+      "step": 1056
+    },
+    {
+      "epoch": 0.42229324810227725,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.108,
+      "step": 1057
+    },
+    {
+      "epoch": 0.42269276867758687,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0981,
+      "step": 1058
+    },
+    {
+      "epoch": 0.42309228925289655,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1095,
+      "step": 1059
+    },
+    {
+      "epoch": 0.42349180982820617,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1015,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4238913304035158,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1184,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4242908509788254,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0952,
+      "step": 1062
+    },
+    {
+      "epoch": 0.424690371554135,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1005,
+      "step": 1063
+    },
+    {
+      "epoch": 0.42508989212944465,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1064,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4254894127047543,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.106,
+      "step": 1065
+    },
+    {
+      "epoch": 0.42588893328006394,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0977,
+      "step": 1066
+    },
+    {
+      "epoch": 0.42628845385537356,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1119,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4266879744306832,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1161,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4270874950059928,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0958,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4274870155813024,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 1070
+    },
+    {
+      "epoch": 0.42788653615661204,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1184,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4282860567319217,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1183,
+      "step": 1072
+    },
+    {
+      "epoch": 0.42868557730723134,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1076,
+      "step": 1073
+    },
+    {
+      "epoch": 0.42908509788254096,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.1098,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4294846184578506,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1199,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4298841390331602,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1058,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4302836596084698,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1095,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4306831801837795,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1184,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4310827007590891,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0886,
+      "step": 1079
+    },
+    {
+      "epoch": 0.43148222133439873,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.111,
+      "step": 1080
+    },
+    {
+      "epoch": 0.43188174190970835,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0993,
+      "step": 1081
+    },
+    {
+      "epoch": 0.43228126248501797,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1068,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4326807830603276,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1311,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4330803036356372,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0983,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4334798242109469,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1168,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4338793447862565,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1093,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4342788653615661,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1235,
+      "step": 1087
+    },
+    {
+      "epoch": 0.43467838593687574,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1045,
+      "step": 1088
+    },
+    {
+      "epoch": 0.43507790651218536,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1013,
+      "step": 1089
+    },
+    {
+      "epoch": 0.435477427087495,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1054,
+      "step": 1090
+    },
+    {
+      "epoch": 0.43587694766280466,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 1091
+    },
+    {
+      "epoch": 0.4362764682381143,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1086,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4366759888134239,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1044,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4370755093887335,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1061,
+      "step": 1094
+    },
+    {
+      "epoch": 0.43747502996404314,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1059,
+      "step": 1095
+    },
+    {
+      "epoch": 0.43787455053935276,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.109,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4382740711146624,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1059,
+      "step": 1097
+    },
+    {
+      "epoch": 0.43867359168997205,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1094,
+      "step": 1098
+    },
+    {
+      "epoch": 0.4390731122652817,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0991,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4394726328405913,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1029,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4398721534159009,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1135,
+      "step": 1101
+    },
+    {
+      "epoch": 0.44027167399121053,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 1102
+    },
+    {
+      "epoch": 0.44067119456652015,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1105,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4410707151418298,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0954,
+      "step": 1104
+    },
+    {
+      "epoch": 0.44147023571713945,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0878,
+      "step": 1105
+    },
+    {
+      "epoch": 0.44186975629244907,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0994,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4422692768677587,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0992,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4426687974430683,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1084,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4430683180183779,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0918,
+      "step": 1109
+    },
+    {
+      "epoch": 0.4434678385936876,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1117,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4438673591689972,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1111
+    },
+    {
+      "epoch": 0.44426687974430684,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0944,
+      "step": 1112
+    },
+    {
+      "epoch": 0.44466640031961646,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1118,
+      "step": 1113
+    },
+    {
+      "epoch": 0.4450659208949261,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1046,
+      "step": 1114
+    },
+    {
+      "epoch": 0.4454654414702357,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.107,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4458649620455453,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0954,
+      "step": 1116
+    },
+    {
+      "epoch": 0.446264482620855,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1131,
+      "step": 1117
+    },
+    {
+      "epoch": 0.4466640031961646,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0977,
+      "step": 1118
+    },
+    {
+      "epoch": 0.44706352377147424,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.094,
+      "step": 1119
+    },
+    {
+      "epoch": 0.44746304434678386,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4478625649220935,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1049,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4482620854974031,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1015,
+      "step": 1122
+    },
+    {
+      "epoch": 0.44866160607271277,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1017,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4490611266480224,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1021,
+      "step": 1124
+    },
+    {
+      "epoch": 0.449460647223332,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 1125
+    },
+    {
+      "epoch": 0.44986016779864163,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1063,
+      "step": 1126
+    },
+    {
+      "epoch": 0.45025968837395125,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1025,
+      "step": 1127
+    },
+    {
+      "epoch": 0.45065920894926087,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1174,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4510587295245705,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1129
+    },
+    {
+      "epoch": 0.45145825009988017,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4518577706751898,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1016,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4522572912504994,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0972,
+      "step": 1132
+    },
+    {
+      "epoch": 0.452656811825809,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1077,
+      "step": 1133
+    },
+    {
+      "epoch": 0.45305633240111864,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1007,
+      "step": 1134
+    },
+    {
+      "epoch": 0.45345585297642826,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1082,
+      "step": 1135
+    },
+    {
+      "epoch": 0.45385537355173794,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 1136
+    },
+    {
+      "epoch": 0.45425489412704756,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4546544147023572,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0933,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4550539352776668,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1135,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4554534558529764,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1041,
+      "step": 1140
+    },
+    {
+      "epoch": 0.45585297642828604,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1054,
+      "step": 1141
+    },
+    {
+      "epoch": 0.45625249700359566,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1036,
+      "step": 1142
+    },
+    {
+      "epoch": 0.45665201757890533,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1019,
+      "step": 1143
+    },
+    {
+      "epoch": 0.45705153815421495,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1012,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4574510587295246,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.096,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4578505793048342,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4582500998801438,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0976,
+      "step": 1147
+    },
+    {
+      "epoch": 0.45864962045545343,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1063,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4590491410307631,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1149
+    },
+    {
+      "epoch": 0.45944866160607273,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0933,
+      "step": 1150
+    },
+    {
+      "epoch": 0.45984818218138235,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1062,
+      "step": 1151
+    },
+    {
+      "epoch": 0.46024770275669197,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1137,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4606472233320016,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1079,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4610467439073112,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4614462644826209,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1008,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4618457850579305,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1076,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4622453056332401,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1033,
+      "step": 1157
+    },
+    {
+      "epoch": 0.46264482620854974,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1003,
+      "step": 1158
+    },
+    {
+      "epoch": 0.46304434678385936,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1013,
+      "step": 1159
+    },
+    {
+      "epoch": 0.463443867359169,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4638433879344786,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0962,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4642429085097883,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0988,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4646424290850979,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0962,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4650419496604075,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1197,
+      "step": 1164
+    },
+    {
+      "epoch": 0.46544147023571714,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1057,
+      "step": 1165
+    },
+    {
+      "epoch": 0.46584099081102676,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.114,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4662405113863364,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1031,
+      "step": 1167
+    },
+    {
+      "epoch": 0.46664003196164605,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0964,
+      "step": 1168
+    },
+    {
+      "epoch": 0.46703955253695567,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1088,
+      "step": 1169
+    },
+    {
+      "epoch": 0.4674390731122653,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4678385936875749,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1086,
+      "step": 1171
+    },
+    {
+      "epoch": 0.46823811426288453,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0996,
+      "step": 1172
+    },
+    {
+      "epoch": 0.46863763483819415,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0955,
+      "step": 1173
+    },
+    {
+      "epoch": 0.46903715541350377,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1042,
+      "step": 1174
+    },
+    {
+      "epoch": 0.46943667598881345,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1077,
+      "step": 1175
+    },
+    {
+      "epoch": 0.46983619656412307,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4702357171394327,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1061,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4706352377147423,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1025,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4710347582900519,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1008,
+      "step": 1179
+    },
+    {
+      "epoch": 0.47143427886536154,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0991,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4718337994406712,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1065,
+      "step": 1181
+    },
+    {
+      "epoch": 0.47223332001598084,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 1182
+    },
+    {
+      "epoch": 0.47263284059129046,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1035,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4730323611666001,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1053,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4734318817419097,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1088,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4738314023172193,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0987,
+      "step": 1186
+    },
+    {
+      "epoch": 0.47423092289252894,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4746304434678386,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 1188
+    },
+    {
+      "epoch": 0.47502996404314823,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1189
+    },
+    {
+      "epoch": 0.47542948461845785,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1048,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4758290051937675,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1037,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4762285257690771,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1056,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4766280463443867,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0974,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4770275669196964,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0984,
+      "step": 1194
+    },
+    {
+      "epoch": 0.477427087495006,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 1195
+    },
+    {
+      "epoch": 0.47782660807031563,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.106,
+      "step": 1196
+    },
+    {
+      "epoch": 0.47822612864562525,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1106,
+      "step": 1197
+    },
+    {
+      "epoch": 0.47862564922093487,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1059,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4790251697962445,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0934,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4794246903715541,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1034,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4798242109468638,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1015,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4802237315221734,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1113,
+      "step": 1202
+    },
+    {
+      "epoch": 0.480623252097483,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1088,
+      "step": 1203
+    },
+    {
+      "epoch": 0.48102277267279264,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1086,
+      "step": 1204
+    },
+    {
+      "epoch": 0.48142229324810226,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1095,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4818218138234119,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0936,
+      "step": 1206
+    },
+    {
+      "epoch": 0.48222133439872156,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0981,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4826208549740312,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4830203755493408,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1047,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4834198961246504,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1044,
+      "step": 1210
+    },
+    {
+      "epoch": 0.48381941669996004,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1211
+    },
+    {
+      "epoch": 0.48421893727526966,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1052,
+      "step": 1212
+    },
+    {
+      "epoch": 0.48461845785057933,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.093,
+      "step": 1213
+    },
+    {
+      "epoch": 0.48501797842588895,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0988,
+      "step": 1214
+    },
+    {
+      "epoch": 0.48541749900119857,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1106,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4858170195765082,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0976,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4862165401518178,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1056,
+      "step": 1217
+    },
+    {
+      "epoch": 0.48661606072712743,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1132,
+      "step": 1218
+    },
+    {
+      "epoch": 0.48701558130243705,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0918,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4874151018777467,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1038,
+      "step": 1220
+    },
+    {
+      "epoch": 0.48781462245305635,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0982,
+      "step": 1221
+    },
+    {
+      "epoch": 0.48821414302836597,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1068,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4886136636036756,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1056,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4890131841789852,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1067,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4894127047542948,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1066,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4898122253296045,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0924,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4902117459049141,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 1227
+    },
+    {
+      "epoch": 0.49061126648022374,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0965,
+      "step": 1228
+    },
+    {
+      "epoch": 0.49101078705553336,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.098,
+      "step": 1229
+    },
+    {
+      "epoch": 0.491410307630843,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1095,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4918098282061526,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1062,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4922093487814622,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1001,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4926088693567719,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4930083899320815,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1053,
+      "step": 1234
+    },
+    {
+      "epoch": 0.49340791050739113,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0979,
+      "step": 1235
+    },
+    {
+      "epoch": 0.49380743108270075,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 1236
+    },
+    {
+      "epoch": 0.4942069516580104,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1113,
+      "step": 1237
+    },
+    {
+      "epoch": 0.49460647223332,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 1238
+    },
+    {
+      "epoch": 0.49500599280862967,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4954055133839393,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0871,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4958050339592489,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1037,
+      "step": 1241
+    },
+    {
+      "epoch": 0.49620455453455853,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0938,
+      "step": 1242
+    },
+    {
+      "epoch": 0.49660407510986815,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0929,
+      "step": 1243
+    },
+    {
+      "epoch": 0.49700359568517777,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0903,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4974031162604874,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1018,
+      "step": 1245
+    },
+    {
+      "epoch": 0.49780263683579706,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1063,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4982021574111067,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1045,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4986016779864163,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0994,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4990011985617259,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0974,
+      "step": 1249
+    },
+    {
+      "epoch": 0.49940071913703554,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0989,
+      "step": 1250
+    },
+    {
+      "epoch": 0.49980023971234516,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1052,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5001997602876548,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1056,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5005992808629645,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0985,
+      "step": 1253
+    },
+    {
+      "epoch": 0.500998801438274,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5013983220135837,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1023,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5017978425888934,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0969,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5021973631642029,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5025968837395126,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0974,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5029964043148222,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5033959248901319,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1099,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5037954454654414,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1097,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5041949660407511,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1036,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5045944866160608,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1126,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5049940071913703,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1264
+    },
+    {
+      "epoch": 0.50539352776668,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0883,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5057930483419896,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0981,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5061925689172992,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1096,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5065920894926089,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1091,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5069916100679185,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0957,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5073911306432282,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1023,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5077906512185377,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1057,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5081901717938474,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0922,
+      "step": 1272
+    },
+    {
+      "epoch": 0.508589692369157,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5089892129444666,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0994,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5093887335197763,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0919,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5097882540950859,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5101877746703956,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.107,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5105872952457051,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1076,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5109868158210148,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5113863363963244,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0959,
+      "step": 1280
+    },
+    {
+      "epoch": 0.511785856971634,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1068,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5121853775469437,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0945,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5125848981222533,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0891,
+      "step": 1283
+    },
+    {
+      "epoch": 0.512984418697563,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1016,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5133839392728725,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1033,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5137834598481822,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1044,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5141829804234919,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1069,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5145825009988014,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0928,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5149820215741111,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0955,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5153815421494207,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0982,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5157810627247303,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.0972,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5161805833000399,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1148,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5165801038753496,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5169796244506593,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0958,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5173791450259688,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0981,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5177786656012785,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0979,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5181781861765881,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0915,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5185777067518977,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0964,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5189772273272073,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0994,
+      "step": 1299
+    },
+    {
+      "epoch": 0.519376747902517,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1063,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5197762684778267,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5201757890531362,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.11,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5205753096284459,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0948,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5209748302037555,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0973,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5213743507790651,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1017,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5217738713543747,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5221733919296844,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.089,
+      "step": 1307
+    },
+    {
+      "epoch": 0.522572912504994,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5229724330803036,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.105,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5233719536556133,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5237714742309229,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5241709948062325,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0932,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5245705153815422,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.1035,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5249700359568518,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5253695565321614,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0998,
+      "step": 1315
+    },
+    {
+      "epoch": 0.525769077107471,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0954,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5261685976827807,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1071,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5265681182580902,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0974,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5269676388333999,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1193,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5273671594087096,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5277666799840192,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1009,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5281662005593288,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5285657211346384,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1031,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5289652417099481,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1172,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5293647622852576,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1024,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5297642828605673,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0851,
+      "step": 1326
+    },
+    {
+      "epoch": 0.530163803435877,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5305633240111866,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0996,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5309628445864962,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1007,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5313623651618058,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5317618857371155,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5321614063124251,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5325609268877347,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0971,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5329604474630444,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 1334
+    },
+    {
+      "epoch": 0.533359968038354,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0992,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5337594886136636,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5341590091889732,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0944,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5345585297642829,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5349580503395925,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0881,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5353575709149021,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0985,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5357570914902118,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1024,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5361566120655213,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 1342
+    },
+    {
+      "epoch": 0.536556132640831,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0936,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5369556532161406,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1062,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5373551737914503,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5377546943667599,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.093,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5381542149420695,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0933,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5385537355173792,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0921,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5389532560926887,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0962,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5393527766679984,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0991,
+      "step": 1350
+    },
+    {
+      "epoch": 0.539752297243308,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0964,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5401518178186177,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0975,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5405513383939273,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.1128,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5409508589692369,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.114,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5413503795445466,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1042,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5417499001198561,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5421494206951658,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0948,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5425489412704755,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1038,
+      "step": 1358
+    },
+    {
+      "epoch": 0.542948461845785,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1042,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5433479824210947,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1009,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5437475029964043,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1107,
+      "step": 1361
+    },
+    {
+      "epoch": 0.544147023571714,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0923,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5445465441470235,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1038,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5449460647223332,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1006,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5453455852976429,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5457451058729524,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.099,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5461446264482621,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1045,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5465441470235717,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0982,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5469436675988814,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1018,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5473431881741909,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1028,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5477427087495006,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0954,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5481422293248103,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0924,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5485417499001198,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0987,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5489412704754295,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5493407910507391,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5497403116260487,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1041,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5501398322013584,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0957,
+      "step": 1377
+    },
+    {
+      "epoch": 0.550539352776668,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0926,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5509388733519777,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1007,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5513383939272872,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5517379145025969,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.1093,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5521374350779065,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1018,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5525369556532161,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0938,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5529364762285258,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0876,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5533359968038354,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0922,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5537355173791451,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5541350379544546,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1078,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5545345585297643,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0975,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5549340791050739,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0985,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5553335996803835,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0978,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5557331202556932,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0855,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5561326408310028,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5565321614063125,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1086,
+      "step": 1393
+    },
+    {
+      "epoch": 0.556931681981622,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1041,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5573312025569317,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5577307231322413,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5581302437075509,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0983,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5585297642828606,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.1132,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5589292848581702,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5593288054334798,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5597283260087894,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5601278465840991,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0894,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5605273671594088,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0944,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5609268877347183,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1404
+    },
+    {
+      "epoch": 0.561326408310028,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5617259288853376,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.104,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5621254494606472,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5625249700359568,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.108,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5629244906112665,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0967,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5633240111865762,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5637235317618857,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5641230523371954,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1134,
+      "step": 1412
+    },
+    {
+      "epoch": 0.564522572912505,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5649220934878146,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5653216140631242,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.1022,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5657211346384339,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0896,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5661206552137436,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5665201757890531,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5669196963643628,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1034,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5673192169396724,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0945,
+      "step": 1420
+    },
+    {
+      "epoch": 0.567718737514982,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0945,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5681182580902916,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1028,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5685177786656013,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 1423
+    },
+    {
+      "epoch": 0.568917299240911,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5693168198162205,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0933,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5697163403915302,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1003,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5701158609668397,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0943,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5705153815421494,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5709149021174591,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0999,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5713144226927687,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1085,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5717139432680783,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5721134638433879,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0943,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5725129844186976,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5729125049940071,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5733120255693168,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1115,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5737115461446265,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1065,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5741110667199361,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0973,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5745105872952457,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0921,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5749101078705553,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1002,
+      "step": 1439
+    },
+    {
+      "epoch": 0.575309628445865,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5757091490211745,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0985,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5761086695964842,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5765081901717939,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.115,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5769077107471035,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0922,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5773072313224131,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5777067518977227,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5781062724730324,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 1447
+    },
+    {
+      "epoch": 0.578505793048342,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0933,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5789053136236516,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5793048341989613,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5797043547742708,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5801038753495805,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.098,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5805033959248901,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0975,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5809029165001998,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1069,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5813024370755094,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1455
+    },
+    {
+      "epoch": 0.581701957650819,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0881,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5821014782261287,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5825009988014382,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5829005193767479,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0902,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5833000399520575,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5836995605273672,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1032,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5840990811026768,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1075,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5844986016779864,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.1002,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5848981222532961,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0987,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5852976428286056,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5856971634039153,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0942,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5860966839792249,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5864962045545346,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0974,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5868957251298442,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5872952457051538,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5876947662804635,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0861,
+      "step": 1471
+    },
+    {
+      "epoch": 0.588094286855773,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.101,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5884938074310827,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5888933280063924,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 1474
+    },
+    {
+      "epoch": 0.589292848581702,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0918,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5896923691570116,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5900918897323212,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5904914103076309,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0954,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5908909308829404,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5912904514582501,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5916899720335598,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0929,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5920894926088693,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.102,
+      "step": 1482
+    },
+    {
+      "epoch": 0.592489013184179,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0926,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5928885337594886,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1009,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5932880543347983,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0855,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5936875749101078,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5940870954854175,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5944866160607272,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1002,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5948861366360367,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5952856572113464,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0955,
+      "step": 1490
+    },
+    {
+      "epoch": 0.595685177786656,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5960846983619656,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0999,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5964842189372753,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5968837395125849,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5972832600878946,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5976827806632041,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0903,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5980823012385138,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0855,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5984818218138234,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 1498
+    },
+    {
+      "epoch": 0.598881342389133,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0907,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5992808629644427,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0945,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5996803835397523,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 1501
+    },
+    {
+      "epoch": 0.600079904115062,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6004794246903715,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0927,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6008789452656812,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0922,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6012784658409908,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0924,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6016779864163004,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.1013,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6020775069916101,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6024770275669197,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.1005,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6028765481422294,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0924,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6032760687175389,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1078,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6036755892928486,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6040751098681582,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.0932,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6044746304434678,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6048741510187775,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6052736715940871,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0995,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6056731921693967,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6060727127447063,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1517
+    },
+    {
+      "epoch": 0.606472233320016,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0861,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6068717538953257,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0991,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6072712744706352,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0966,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6076707950459449,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0896,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6080703156212545,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6084698361965641,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6088693567718737,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1046,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6092688773471834,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 1525
+    },
+    {
+      "epoch": 0.609668397922493,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1002,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6100679184978026,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0978,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6104674390731123,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.1081,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6108669596484219,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1032,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6112664802237315,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0934,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6116660007990411,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0982,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6120655213743508,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6124650419496604,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.094,
+      "step": 1533
+    },
+    {
+      "epoch": 0.61286456252497,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6132640831002797,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6136636036755893,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6140631242508989,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.101,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6144626448262086,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6148621654015182,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6152616859768278,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0948,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6156612065521374,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1022,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6160607271274471,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0955,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6164602477027566,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1014,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6168597682780663,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0896,
+      "step": 1544
+    },
+    {
+      "epoch": 0.617259288853376,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6176588094286856,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0932,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6180583300039952,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0876,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6184578505793048,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0966,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6188573711546145,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1061,
+      "step": 1549
+    },
+    {
+      "epoch": 0.619256891729924,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0967,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6196564123052337,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0879,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6200559328805434,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1004,
+      "step": 1552
+    },
+    {
+      "epoch": 0.620455453455853,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0982,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6208549740311626,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6212544946064722,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0915,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6216540151817819,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.098,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6220535357570914,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0954,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6224530563324011,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6228525769077108,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1049,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6232520974830204,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 1560
+    },
+    {
+      "epoch": 0.62365161805833,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0907,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6240511386336396,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6244506592089493,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.1028,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6248501797842589,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6252497003595685,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6256492209348782,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6260487415101877,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0999,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6264482620854974,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.1026,
+      "step": 1568
+    },
+    {
+      "epoch": 0.626847782660807,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6272473032361167,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1023,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6276468238114263,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0978,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6280463443867359,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6284458649620456,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0953,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6288453855373551,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0885,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6292449061126648,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.1003,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6296444266879744,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1037,
+      "step": 1576
+    },
+    {
+      "epoch": 0.630043947263284,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0883,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6304434678385937,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0907,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6308429884139033,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 1579
+    },
+    {
+      "epoch": 0.631242508989213,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.091,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6316420295645225,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6320415501398322,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6324410707151419,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 1583
+    },
+    {
+      "epoch": 0.6328405912904514,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0942,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6332401118657611,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 1585
+    },
+    {
+      "epoch": 0.6336396324410707,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6340391530163804,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6344386735916899,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0981,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6348381941669996,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0895,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6352377147423093,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6356372353176188,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0893,
+      "step": 1591
+    },
+    {
+      "epoch": 0.6360367558929285,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0983,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6364362764682381,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6368357970435478,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6372353176188573,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1595
+    },
+    {
+      "epoch": 0.637634838194167,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6380343587694767,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0965,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6384338793447862,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6388333999200959,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0847,
+      "step": 1599
+    },
+    {
+      "epoch": 0.6392329204954055,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0959,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6396324410707152,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1055,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6400319616460247,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0854,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6404314822213344,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1603
+    },
+    {
+      "epoch": 0.6408310027966441,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6412305233719536,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6416300439472633,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6420295645225729,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6424290850978825,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6428286056731922,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0854,
+      "step": 1609
+    },
+    {
+      "epoch": 0.6432281262485018,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6436276468238115,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0946,
+      "step": 1611
+    },
+    {
+      "epoch": 0.644027167399121,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6444266879744307,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0992,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6448262085497403,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0935,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6452257291250499,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0971,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6456252497003596,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0844,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6460247702756692,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6464242908509789,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0931,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6468238114262884,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0966,
+      "step": 1619
+    },
+    {
+      "epoch": 0.6472233320015981,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1025,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6476228525769077,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0005,
+      "loss": 1.107,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6480223731522173,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0943,
+      "step": 1622
+    },
+    {
+      "epoch": 0.648421893727527,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6488214143028366,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.1045,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6492209348781462,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0917,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6496204554534558,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0948,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6500199760287655,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0929,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6504194966040752,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.1023,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6508190171793847,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0932,
+      "step": 1629
+    },
+    {
+      "epoch": 0.6512185377546944,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 1630
+    },
+    {
+      "epoch": 0.651618058330004,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0943,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6520175789053136,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0926,
+      "step": 1632
+    },
+    {
+      "epoch": 0.6524170994806232,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0971,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6528166200559329,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0893,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6532161406312426,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.09,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6536156612065521,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6540151817818618,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0932,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6544147023571714,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0918,
+      "step": 1638
+    },
+    {
+      "epoch": 0.654814222932481,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0916,
+      "step": 1639
+    },
+    {
+      "epoch": 0.6552137435077906,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6556132640831003,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0899,
+      "step": 1641
+    },
+    {
+      "epoch": 0.65601278465841,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6564123052337195,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6568118258090292,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6572113463843388,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0797,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6576108669596484,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0993,
+      "step": 1646
+    },
+    {
+      "epoch": 0.658010387534958,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0876,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6584099081102677,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0903,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6588094286855773,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0934,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6592089492608869,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6596084698361966,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6600079904115062,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.105,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6604075109868158,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0938,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6608070315621255,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0922,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6612065521374351,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6616060727127447,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0893,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6620055932880543,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1657
+    },
+    {
+      "epoch": 0.662405113863364,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0895,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6628046344386735,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6632041550139832,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1028,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6636036755892929,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0941,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6640031961646025,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6644027167399121,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.096,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6648022373152217,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0946,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6652017578905314,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0992,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6656012784658409,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.101,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6660007990411506,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0934,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6664003196164603,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6667998401917699,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0896,
+      "step": 1669
+    },
+    {
+      "epoch": 0.6671993607670795,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0941,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6675988813423891,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6679984019176988,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6683979224930084,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.1034,
+      "step": 1673
+    },
+    {
+      "epoch": 0.668797443068318,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6691969636436277,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.1017,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6695964842189372,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6699960047942469,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0828,
+      "step": 1677
+    },
+    {
+      "epoch": 0.6703955253695565,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6707950459448662,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0928,
+      "step": 1679
+    },
+    {
+      "epoch": 0.6711945665201758,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0943,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6715940870954854,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0919,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6719936076707951,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0931,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6723931282461046,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6727926488214143,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0833,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6731921693967239,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6735916899720336,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0916,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6739912105473432,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0883,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6743907311226528,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.095,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6747902516979625,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 1689
+    },
+    {
+      "epoch": 0.675189772273272,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0913,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6755892928485817,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0855,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6759888134238913,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 1692
+    },
+    {
+      "epoch": 0.676388333999201,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6767878545745106,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6771873751498202,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0876,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6775868957251299,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0928,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6779864163004394,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6783859368757491,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6787854574510588,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 1699
+    },
+    {
+      "epoch": 0.6791849780263683,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 1700
+    },
+    {
+      "epoch": 0.679584498601678,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1041,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6799840191769876,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6803835397522973,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6807830603276068,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0989,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6811825809029165,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6815821014782262,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6819816220535357,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6823811426288454,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1708
+    },
+    {
+      "epoch": 0.682780663204155,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.094,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6831801837794647,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1028,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6835797043547742,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0878,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6839792249300839,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0833,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6843787455053936,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0889,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6847782660807031,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0858,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6851777866560128,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6855773072313224,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0927,
+      "step": 1716
+    },
+    {
+      "epoch": 0.685976827806632,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6863763483819417,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0878,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6867758689572513,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.1032,
+      "step": 1719
+    },
+    {
+      "epoch": 0.687175389532561,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6875749101078705,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6879744306831802,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6883739512584898,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0921,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6887734718337994,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6891729924091091,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6895725129844187,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6899720335597284,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6903715541350379,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6907710747103476,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6911705952856572,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6915701158609668,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0844,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6919696364362765,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0978,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6923691570115861,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0859,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6927686775868958,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1022,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6931681981622053,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0963,
+      "step": 1735
+    },
+    {
+      "epoch": 0.693567718737515,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.097,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6939672393128246,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.09,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6943667598881342,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0886,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6947662804634439,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6951658010387535,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6955653216140631,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6959648421893727,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6963643627646824,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0878,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6967638833399921,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6971634039153016,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0858,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6975629244906113,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6979624450659209,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6983619656412305,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.091,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6987614862165401,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0931,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6991610067918498,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6995605273671595,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 1751
+    },
+    {
+      "epoch": 0.699960047942469,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0929,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7003595685177787,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0931,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7007590890930883,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7011586096683979,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7015581302437075,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7019576508190172,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7023571713943269,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7027566919696364,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0873,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7031562125449461,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.103,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7035557331202557,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7039552536955653,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0853,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7043547742708749,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.1039,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7047542948461846,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1098,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7051538154214942,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7055533359968038,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.1023,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7059528565721135,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0966,
+      "step": 1767
+    },
+    {
+      "epoch": 0.706352377147423,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7067518977227327,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0986,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7071514182980424,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 1770
+    },
+    {
+      "epoch": 0.707550938873352,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7079504594486616,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.095,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7083499800239712,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.092,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7087495005992809,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0965,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7091490211745904,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0936,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7095485417499001,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7099480623252098,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7103475829005194,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0899,
+      "step": 1778
+    },
+    {
+      "epoch": 0.710747103475829,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7111466240511386,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0931,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7115461446264483,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7119456652017578,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7123451857770675,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7127447063523772,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0913,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7131442269276868,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0944,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7135437475029964,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0859,
+      "step": 1786
+    },
+    {
+      "epoch": 0.713943268078306,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 1787
+    },
+    {
+      "epoch": 0.7143427886536157,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7147423092289253,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.093,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7151418298042349,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1006,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7155413503795446,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7159408709548541,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7163403915301638,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7167399121054734,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0902,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7171394326807831,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0855,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7175389532560927,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0902,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7179384738314023,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 1797
+    },
+    {
+      "epoch": 0.718337994406712,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7187375149820215,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0934,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7191370355573312,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0861,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7195365561326408,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7199360767079505,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7203355972832601,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7207351178585697,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7211346384338794,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7215341590091889,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7219336795844986,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7223332001598082,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0893,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7227327207351179,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7231322413104275,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0972,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7235317618857371,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7239312824610468,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7243308030363563,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0913,
+      "step": 1813
+    },
+    {
+      "epoch": 0.724730323611666,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7251298441869757,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0935,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7255293647622852,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0947,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7259288853375949,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0919,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7263284059129045,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1062,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7267279264882142,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0973,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7271274470635237,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7275269676388334,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0989,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7279264882141431,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7283260087894526,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0996,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7287255293647623,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7291250499400719,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7295245705153816,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7299240910906911,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7303236116660008,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7307231322413105,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0938,
+      "step": 1829
+    },
+    {
+      "epoch": 0.73112265281662,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7315221733919297,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7319216939672393,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.1091,
+      "step": 1832
+    },
+    {
+      "epoch": 0.732321214542549,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7327207351178586,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7331202556931682,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7335197762684779,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7339192968437874,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7343188174190971,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.09,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7347183379944067,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7351178585697163,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0965,
+      "step": 1840
+    },
+    {
+      "epoch": 0.735517379145026,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7359168997203356,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.101,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7363164202956453,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7367159408709548,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7371154614462645,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7375149820215741,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7379145025968837,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7383140231721934,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0873,
+      "step": 1848
+    },
+    {
+      "epoch": 0.738713543747503,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 1849
+    },
+    {
+      "epoch": 0.7391130643228127,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7395125848981222,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7399121054734319,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7403116260487415,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0854,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7407111466240511,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7411106671993608,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7415101877746704,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0927,
+      "step": 1856
+    },
+    {
+      "epoch": 0.74190970834998,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0963,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7423092289252896,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0929,
+      "step": 1858
+    },
+    {
+      "epoch": 0.7427087495005993,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 1859
+    },
+    {
+      "epoch": 0.743108270075909,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.1038,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7435077906512185,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7439073112265282,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7443068318018378,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 1863
+    },
+    {
+      "epoch": 0.7447063523771474,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 1864
+    },
+    {
+      "epoch": 0.745105872952457,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7455053935277667,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0903,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7459049141030764,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 1867
+    },
+    {
+      "epoch": 0.7463044346783859,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0997,
+      "step": 1868
+    },
+    {
+      "epoch": 0.7467039552536956,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 1869
+    },
+    {
+      "epoch": 0.7471034758290052,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7475029964043148,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 1871
+    },
+    {
+      "epoch": 0.7479025169796244,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0917,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7483020375549341,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0927,
+      "step": 1873
+    },
+    {
+      "epoch": 0.7487015581302437,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.1003,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7491010787055533,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 1875
+    },
+    {
+      "epoch": 0.749500599280863,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7499001198561726,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0861,
+      "step": 1877
+    },
+    {
+      "epoch": 0.7502996404314822,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 1878
+    },
+    {
+      "epoch": 0.7506991610067919,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0886,
+      "step": 1879
+    },
+    {
+      "epoch": 0.7510986815821015,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0917,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7514982021574111,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0819,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7518977227327207,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7522972433080304,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0913,
+      "step": 1883
+    },
+    {
+      "epoch": 0.75269676388334,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0928,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7530962844586496,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0871,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7534958050339593,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7538953256092689,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7542948461845785,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7546943667598881,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.1018,
+      "step": 1889
+    },
+    {
+      "epoch": 0.7550938873351978,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7554934079105073,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1891
+    },
+    {
+      "epoch": 0.755892928485817,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0915,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7562924490611267,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0903,
+      "step": 1893
+    },
+    {
+      "epoch": 0.7566919696364363,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7570914902117459,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.089,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7574910107870555,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7578905313623652,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 1897
+    },
+    {
+      "epoch": 0.7582900519376747,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 1898
+    },
+    {
+      "epoch": 0.7586895725129844,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 1899
+    },
+    {
+      "epoch": 0.7590890930882941,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7594886136636037,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0975,
+      "step": 1901
+    },
+    {
+      "epoch": 0.7598881342389133,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7602876548142229,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 1903
+    },
+    {
+      "epoch": 0.7606871753895326,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7610866959648422,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 1905
+    },
+    {
+      "epoch": 0.7614862165401518,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7618857371154615,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.1093,
+      "step": 1907
+    },
+    {
+      "epoch": 0.762285257690771,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 1908
+    },
+    {
+      "epoch": 0.7626847782660807,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0915,
+      "step": 1909
+    },
+    {
+      "epoch": 0.7630842988413903,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7634838194167,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7638833399920096,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7642828605673192,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0844,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7646823811426289,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7650819017179384,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0955,
+      "step": 1915
+    },
+    {
+      "epoch": 0.7654814222932481,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7658809428685577,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7662804634438674,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 1918
+    },
+    {
+      "epoch": 0.766679984019177,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 1919
+    },
+    {
+      "epoch": 0.7670795045944866,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0943,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7674790251697963,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0941,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7678785457451058,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 1922
+    },
+    {
+      "epoch": 0.7682780663204155,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0941,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7686775868957252,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0855,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7690771074710347,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7694766280463444,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 1926
+    },
+    {
+      "epoch": 0.769876148621654,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7702756691969637,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.1004,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7706751897722732,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 1929
+    },
+    {
+      "epoch": 0.7710747103475829,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0984,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7714742309228926,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7718737514982021,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7722732720735118,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7726727926488214,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0965,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7730723132241311,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0912,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7734718337994406,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0977,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7738713543747503,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 1937
+    },
+    {
+      "epoch": 0.77427087495006,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1095,
+      "step": 1938
+    },
+    {
+      "epoch": 0.7746703955253695,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7750699161006792,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7754694366759888,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0992,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7758689572512985,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 1942
+    },
+    {
+      "epoch": 0.776268477826608,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7766679984019177,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7770675189772274,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7774670395525369,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7778665601278466,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0932,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7782660807031562,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0831,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7786656012784658,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0836,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7790651218537755,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7794646424290851,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7798641630043948,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7802636835797043,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 1953
+    },
+    {
+      "epoch": 0.780663204155014,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7810627247303236,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7814622453056332,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7818617658809429,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7822612864562525,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0851,
+      "step": 1958
+    },
+    {
+      "epoch": 0.7826608070315622,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7830603276068717,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7834598481821814,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 1961
+    },
+    {
+      "epoch": 0.783859368757491,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7842588893328006,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0996,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7846584099081103,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0886,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7850579304834199,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7854574510587295,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0944,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7858569716340391,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7862564922093488,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7866560127846585,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 1969
+    },
+    {
+      "epoch": 0.787055533359968,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7874550539352777,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0923,
+      "step": 1971
+    },
+    {
+      "epoch": 0.7878545745105873,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7882540950858969,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0941,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7886536156612065,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7890531362365162,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7894526568118259,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7898521773871354,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7902516979624451,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7906512185377547,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7910507391130643,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7914502596883739,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7918497802636836,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7922493008389933,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.0871,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7926488214143028,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7930483419896125,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7934478625649221,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0899,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7938473831402317,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0005,
+      "loss": 1.095,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7942469037155413,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1058,
+      "step": 1988
+    },
+    {
+      "epoch": 0.794646424290851,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7950459448661606,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0005,
+      "loss": 1.0984,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7954454654414702,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0907,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7958449860167799,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0977,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7962445065920895,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7966440271673991,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0917,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7970435477427088,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7974430683180184,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0841,
+      "step": 1996
+    },
+    {
+      "epoch": 0.797842588893328,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0902,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7982421094686376,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0896,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7986416300439473,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.095,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7990411506192568,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0851,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7994406711945665,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7998401917698762,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0939,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8002397123451858,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8006392329204954,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 2004
+    },
+    {
+      "epoch": 0.801038753495805,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.1013,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8014382740711147,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8018377946464242,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8022373152217339,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8026368357970436,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8030363563723532,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0952,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8034358769476628,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0879,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8038353975229724,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0883,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8042349180982821,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0831,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8046344386735917,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8050339592489013,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0976,
+      "step": 2015
+    },
+    {
+      "epoch": 0.805433479824211,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0836,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8058330003995205,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8062325209748302,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8066320415501398,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8070315621254495,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8074310827007591,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8078306032760687,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8082301238513784,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8086296444266879,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8090291650019976,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8094286855773072,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8098282061526169,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8102277267279265,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8106272473032361,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8110267678785458,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8114262884538553,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0876,
+      "step": 2031
+    },
+    {
+      "epoch": 0.811825809029165,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8122253296044746,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8126248501797843,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0878,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8130243707550939,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 2035
+    },
+    {
+      "epoch": 0.8134238913304035,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8138234119057132,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 2037
+    },
+    {
+      "epoch": 0.8142229324810227,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8146224530563324,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 2039
+    },
+    {
+      "epoch": 0.8150219736316421,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8154214942069516,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2041
+    },
+    {
+      "epoch": 0.8158210147822613,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8162205353575709,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8166200559328806,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8170195765081901,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8174190970834998,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8178186176588095,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 2047
+    },
+    {
+      "epoch": 0.818218138234119,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8186176588094287,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 2049
+    },
+    {
+      "epoch": 0.8190171793847383,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0968,
+      "step": 2050
+    },
+    {
+      "epoch": 0.819416699960048,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8198162205353575,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8202157411106672,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 2053
+    },
+    {
+      "epoch": 0.8206152616859769,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0886,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8210147822612864,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8214143028365961,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8218138234119057,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8222133439872154,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8226128645625249,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8230123851378346,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8234119057131443,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8238114262884538,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8242109468637635,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8246104674390731,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0919,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8250099880143827,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 2065
+    },
+    {
+      "epoch": 0.8254095085896924,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0915,
+      "step": 2066
+    },
+    {
+      "epoch": 0.825809029165002,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0881,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8262085497403117,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8266080703156212,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8270075908909309,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8274071114662405,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8278066320415501,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0834,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8282061526168598,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8286056731921694,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0958,
+      "step": 2074
+    },
+    {
+      "epoch": 0.829005193767479,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.09,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8294047143427886,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8298042349180983,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8302037554934079,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8306032760687175,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8310027966440272,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8314023172193368,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8318018377946464,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 2082
+    },
+    {
+      "epoch": 0.832201358369956,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8326008789452657,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8330003995205754,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8333999200958849,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8337994406711946,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8341989612465042,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8345984818218138,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0907,
+      "step": 2089
+    },
+    {
+      "epoch": 0.8349980023971234,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0955,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8353975229724331,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8357970435477428,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8361965641230523,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 2093
+    },
+    {
+      "epoch": 0.836596084698362,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0895,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8369956052736716,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8373951258489812,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8377946464242908,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0974,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8381941669996005,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8385936875749102,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8389932081502197,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8393927287255294,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 2101
+    },
+    {
+      "epoch": 0.839792249300839,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0836,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8401917698761486,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8405912904514582,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8409908110267679,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0924,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8413903316020775,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8417898521773871,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.092,
+      "step": 2107
+    },
+    {
+      "epoch": 0.8421893727526968,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8425888933280063,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 2109
+    },
+    {
+      "epoch": 0.842988413903316,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8433879344786257,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 2111
+    },
+    {
+      "epoch": 0.8437874550539353,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0887,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8441869756292449,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8445864962045545,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8449860167798642,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8453855373551737,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8457850579304834,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0891,
+      "step": 2117
+    },
+    {
+      "epoch": 0.8461845785057931,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0833,
+      "step": 2118
+    },
+    {
+      "epoch": 0.8465840990811027,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 2119
+    },
+    {
+      "epoch": 0.8469836196564123,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8473831402317219,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0919,
+      "step": 2121
+    },
+    {
+      "epoch": 0.8477826608070316,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8481821813823411,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0871,
+      "step": 2123
+    },
+    {
+      "epoch": 0.8485817019576508,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 2124
+    },
+    {
+      "epoch": 0.8489812225329605,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 2125
+    },
+    {
+      "epoch": 0.84938074310827,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8497802636835797,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 2127
+    },
+    {
+      "epoch": 0.8501797842588893,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2128
+    },
+    {
+      "epoch": 0.850579304834199,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0861,
+      "step": 2129
+    },
+    {
+      "epoch": 0.8509788254095086,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8513783459848182,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 2131
+    },
+    {
+      "epoch": 0.8517778665601279,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8521773871354374,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2133
+    },
+    {
+      "epoch": 0.8525769077107471,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 2134
+    },
+    {
+      "epoch": 0.8529764282860567,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 2135
+    },
+    {
+      "epoch": 0.8533759488613664,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2136
+    },
+    {
+      "epoch": 0.853775469436676,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2137
+    },
+    {
+      "epoch": 0.8541749900119856,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2138
+    },
+    {
+      "epoch": 0.8545745105872953,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2139
+    },
+    {
+      "epoch": 0.8549740311626048,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8553735517379145,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.093,
+      "step": 2141
+    },
+    {
+      "epoch": 0.8557730723132241,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0831,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8561725928885338,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.1007,
+      "step": 2143
+    },
+    {
+      "epoch": 0.8565721134638434,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2144
+    },
+    {
+      "epoch": 0.856971634039153,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 2145
+    },
+    {
+      "epoch": 0.8573711546144627,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0921,
+      "step": 2146
+    },
+    {
+      "epoch": 0.8577706751897722,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2147
+    },
+    {
+      "epoch": 0.8581701957650819,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0977,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8585697163403915,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0913,
+      "step": 2149
+    },
+    {
+      "epoch": 0.8589692369157012,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8593687574910108,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 2151
+    },
+    {
+      "epoch": 0.8597682780663204,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8601677986416301,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.1041,
+      "step": 2153
+    },
+    {
+      "epoch": 0.8605673192169396,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0961,
+      "step": 2154
+    },
+    {
+      "epoch": 0.8609668397922493,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2155
+    },
+    {
+      "epoch": 0.861366360367559,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8617658809428685,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 2157
+    },
+    {
+      "epoch": 0.8621654015181782,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8625649220934878,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 2159
+    },
+    {
+      "epoch": 0.8629644426687975,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 2160
+    },
+    {
+      "epoch": 0.863363963244107,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2161
+    },
+    {
+      "epoch": 0.8637634838194167,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8641630043947264,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2163
+    },
+    {
+      "epoch": 0.8645625249700359,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8649620455453456,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 2165
+    },
+    {
+      "epoch": 0.8653615661206552,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2166
+    },
+    {
+      "epoch": 0.8657610866959649,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2167
+    },
+    {
+      "epoch": 0.8661606072712744,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8665601278465841,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 2169
+    },
+    {
+      "epoch": 0.8669596484218938,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8673591689972033,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 2171
+    },
+    {
+      "epoch": 0.867758689572513,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8681582101478226,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2173
+    },
+    {
+      "epoch": 0.8685577307231322,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8689572512984419,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 2175
+    },
+    {
+      "epoch": 0.8693567718737515,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8697562924490612,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 2177
+    },
+    {
+      "epoch": 0.8701558130243707,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2178
+    },
+    {
+      "epoch": 0.8705553335996804,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 2179
+    },
+    {
+      "epoch": 0.87095485417499,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8713543747502996,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 2181
+    },
+    {
+      "epoch": 0.8717538953256093,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8721534159009189,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 2183
+    },
+    {
+      "epoch": 0.8725529364762286,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8729524570515381,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8733519776268478,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8737514982021574,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 2187
+    },
+    {
+      "epoch": 0.874151018777467,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8745505393527767,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 2189
+    },
+    {
+      "epoch": 0.8749500599280863,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2190
+    },
+    {
+      "epoch": 0.875349580503396,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8757491010787055,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8761486216540152,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8765481422293248,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8769476628046344,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8773471833799441,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0834,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8777467039552537,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8781462245305633,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0975,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8785457451058729,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 2199
+    },
+    {
+      "epoch": 0.8789452656811826,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8793447862564923,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8797443068318018,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0891,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8801438274071115,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 2203
+    },
+    {
+      "epoch": 0.8805433479824211,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8809428685577307,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8813423891330403,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0925,
+      "step": 2206
+    },
+    {
+      "epoch": 0.88174190970835,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8821414302836597,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8825409508589692,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8829404714342789,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8833399920095885,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0905,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8837395125848981,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8841390331602077,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8845385537355174,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 2214
+    },
+    {
+      "epoch": 0.884938074310827,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0819,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8853375948861366,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8857371154614463,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 2217
+    },
+    {
+      "epoch": 0.8861366360367559,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8865361566120655,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8869356771873752,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.092,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8873351977626848,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8877347183379944,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 2222
+    },
+    {
+      "epoch": 0.888134238913304,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8885337594886137,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8889332800639232,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0889,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8893328006392329,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8897323212145426,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8901318417898522,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8905313623651618,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8909308829404714,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8913304035157811,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8917299240910906,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8921294446664003,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2233
+    },
+    {
+      "epoch": 0.89252896524171,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0899,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8929284858170196,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0895,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8933280063923292,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0885,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8937275269676388,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0829,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8941270475429485,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 2238
+    },
+    {
+      "epoch": 0.894526568118258,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8949260886935677,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8953256092688774,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.1001,
+      "step": 2241
+    },
+    {
+      "epoch": 0.895725129844187,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8961246504194966,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8965241709948062,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 2244
+    },
+    {
+      "epoch": 0.8969236915701159,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8973232121454255,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8977227327207351,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8981222532960448,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8985217738713543,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2249
+    },
+    {
+      "epoch": 0.898921294446664,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8993208150219736,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8997203355972833,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9001198561725929,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9005193767479025,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9009188973232122,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9013184178985217,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9017179384738314,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 2257
+    },
+    {
+      "epoch": 0.902117459049141,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.1016,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9025169796244507,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9029165001997603,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9033160207750699,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9037155413503796,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9041150619256891,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 2263
+    },
+    {
+      "epoch": 0.9045145825009988,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0917,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9049141030763085,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 2265
+    },
+    {
+      "epoch": 0.905313623651618,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9057131442269277,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9061126648022373,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 2268
+    },
+    {
+      "epoch": 0.906512185377547,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0951,
+      "step": 2269
+    },
+    {
+      "epoch": 0.9069117059528565,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0763,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9073112265281662,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9077107471034759,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 2272
+    },
+    {
+      "epoch": 0.9081102676787854,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9085097882540951,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9089093088294047,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9093088294047144,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9097083499800239,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9101078705553336,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9105073911306433,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9109069117059528,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.092,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9113064322812625,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0829,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9117059528565721,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0853,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9121054734318818,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9125049940071913,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0847,
+      "step": 2284
+    },
+    {
+      "epoch": 0.912904514582501,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0924,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9133040351578107,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0895,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9137035557331202,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9141030763084299,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9145025968837395,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0778,
+      "step": 2289
+    },
+    {
+      "epoch": 0.9149021174590491,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9153016380343588,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9157011586096684,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0893,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9161006791849781,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0829,
+      "step": 2293
+    },
+    {
+      "epoch": 0.9165001997602876,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0919,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9168997203355973,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 2295
+    },
+    {
+      "epoch": 0.9172992409109069,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9176987614862165,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 2297
+    },
+    {
+      "epoch": 0.9180982820615262,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9184978026368358,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2299
+    },
+    {
+      "epoch": 0.9188973232121455,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 2300
+    },
+    {
+      "epoch": 0.919296843787455,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.093,
+      "step": 2301
+    },
+    {
+      "epoch": 0.9196963643627647,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9200958849380743,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0897,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9204954055133839,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0885,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9208949260886936,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 2305
+    },
+    {
+      "epoch": 0.9212944466640032,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9216939672393129,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9220934878146224,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9224930083899321,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 2309
+    },
+    {
+      "epoch": 0.9228925289652418,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0828,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9232920495405513,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2311
+    },
+    {
+      "epoch": 0.923691570115861,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9240910906911706,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 2313
+    },
+    {
+      "epoch": 0.9244906112664802,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9248901318417898,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2315
+    },
+    {
+      "epoch": 0.9252896524170995,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0992,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9256891729924092,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9260886935677187,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9264882141430284,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2319
+    },
+    {
+      "epoch": 0.926887734718338,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0879,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9272872552936476,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9276867758689572,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0931,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9280862964442669,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.089,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9284858170195766,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0841,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9288853375948861,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0916,
+      "step": 2325
+    },
+    {
+      "epoch": 0.9292848581701958,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9296843787455054,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 2327
+    },
+    {
+      "epoch": 0.930083899320815,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9304834198961246,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0797,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9308829404714343,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2330
+    },
+    {
+      "epoch": 0.931282461046744,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9316819816220535,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9320815021973632,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9324810227726728,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9328805433479824,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9332800639232921,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9336795844986017,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9340791050739113,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9344786256492209,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9348781462245306,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.089,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9352776667998401,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2341
+    },
+    {
+      "epoch": 0.9356771873751498,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9360767079504595,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2343
+    },
+    {
+      "epoch": 0.9364762285257691,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0885,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9368757491010787,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 2345
+    },
+    {
+      "epoch": 0.9372752696763883,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 2346
+    },
+    {
+      "epoch": 0.937674790251698,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9380743108270075,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9384738314023172,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0853,
+      "step": 2349
+    },
+    {
+      "epoch": 0.9388733519776269,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0889,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9392728725529365,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0935,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9396723931282461,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9400719137035557,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9404714342788654,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9408709548541749,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 2355
+    },
+    {
+      "epoch": 0.9412704754294846,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0828,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9416699960047943,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9420695165801038,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9424690371554135,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0929,
+      "step": 2359
+    },
+    {
+      "epoch": 0.9428685577307231,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9432680783060328,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 2361
+    },
+    {
+      "epoch": 0.9436675988813424,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2362
+    },
+    {
+      "epoch": 0.944067119456652,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9444666400319617,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9448661606072712,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.094,
+      "step": 2365
+    },
+    {
+      "epoch": 0.9452656811825809,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9456652017578905,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 2367
+    },
+    {
+      "epoch": 0.9460647223332002,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9464642429085098,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0853,
+      "step": 2369
+    },
+    {
+      "epoch": 0.9468637634838194,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9472632840591291,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 2371
+    },
+    {
+      "epoch": 0.9476628046344386,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9480623252097483,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2373
+    },
+    {
+      "epoch": 0.9484618457850579,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2374
+    },
+    {
+      "epoch": 0.9488613663603676,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2375
+    },
+    {
+      "epoch": 0.9492608869356772,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9496604075109868,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2377
+    },
+    {
+      "epoch": 0.9500599280862965,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2378
+    },
+    {
+      "epoch": 0.950459448661606,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 2379
+    },
+    {
+      "epoch": 0.9508589692369157,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9512584898122254,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 2381
+    },
+    {
+      "epoch": 0.951658010387535,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2382
+    },
+    {
+      "epoch": 0.9520575309628446,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2383
+    },
+    {
+      "epoch": 0.9524570515381542,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 2384
+    },
+    {
+      "epoch": 0.9528565721134639,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 2385
+    },
+    {
+      "epoch": 0.9532560926887734,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 2386
+    },
+    {
+      "epoch": 0.9536556132640831,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 2387
+    },
+    {
+      "epoch": 0.9540551338393928,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 2388
+    },
+    {
+      "epoch": 0.9544546544147023,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 2389
+    },
+    {
+      "epoch": 0.954854174990012,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0949,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9552536955653216,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2391
+    },
+    {
+      "epoch": 0.9556532161406313,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9560527367159408,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2393
+    },
+    {
+      "epoch": 0.9564522572912505,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 2394
+    },
+    {
+      "epoch": 0.9568517778665602,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 2395
+    },
+    {
+      "epoch": 0.9572512984418697,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2396
+    },
+    {
+      "epoch": 0.9576508190171794,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2397
+    },
+    {
+      "epoch": 0.958050339592489,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0975,
+      "step": 2398
+    },
+    {
+      "epoch": 0.9584498601677987,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 2399
+    },
+    {
+      "epoch": 0.9588493807431082,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9592489013184179,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 2401
+    },
+    {
+      "epoch": 0.9596484218937276,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9600479424690371,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 2403
+    },
+    {
+      "epoch": 0.9604474630443468,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0873,
+      "step": 2404
+    },
+    {
+      "epoch": 0.9608469836196564,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 2405
+    },
+    {
+      "epoch": 0.961246504194966,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9616460247702757,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 2407
+    },
+    {
+      "epoch": 0.9620455453455853,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 2408
+    },
+    {
+      "epoch": 0.962445065920895,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2409
+    },
+    {
+      "epoch": 0.9628445864962045,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9632441070715142,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 2411
+    },
+    {
+      "epoch": 0.9636436276468238,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9640431482221334,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0921,
+      "step": 2413
+    },
+    {
+      "epoch": 0.9644426687974431,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 2414
+    },
+    {
+      "epoch": 0.9648421893727527,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 2415
+    },
+    {
+      "epoch": 0.9652417099480624,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9656412305233719,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2417
+    },
+    {
+      "epoch": 0.9660407510986816,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 2418
+    },
+    {
+      "epoch": 0.9664402716739912,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2419
+    },
+    {
+      "epoch": 0.9668397922493008,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9672393128246105,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0831,
+      "step": 2421
+    },
+    {
+      "epoch": 0.9676388333999201,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9680383539752297,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2423
+    },
+    {
+      "epoch": 0.9684378745505393,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 2424
+    },
+    {
+      "epoch": 0.968837395125849,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.1012,
+      "step": 2425
+    },
+    {
+      "epoch": 0.9692369157011587,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9696364362764682,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 2427
+    },
+    {
+      "epoch": 0.9700359568517779,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9704354774270875,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2429
+    },
+    {
+      "epoch": 0.9708349980023971,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9712345185777067,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 2431
+    },
+    {
+      "epoch": 0.9716340391530164,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9720335597283261,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2433
+    },
+    {
+      "epoch": 0.9724330803036356,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9728326008789453,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 2435
+    },
+    {
+      "epoch": 0.9732321214542549,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0868,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9736316420295645,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 2437
+    },
+    {
+      "epoch": 0.9740311626048741,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9744306831801838,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 2439
+    },
+    {
+      "epoch": 0.9748302037554935,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0881,
+      "step": 2440
+    },
+    {
+      "epoch": 0.975229724330803,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 2441
+    },
+    {
+      "epoch": 0.9756292449061127,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0892,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9760287654814223,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 2443
+    },
+    {
+      "epoch": 0.9764282860567319,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2444
+    },
+    {
+      "epoch": 0.9768278066320415,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0847,
+      "step": 2445
+    },
+    {
+      "epoch": 0.9772273272073512,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9776268477826608,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 2447
+    },
+    {
+      "epoch": 0.9780263683579704,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2448
+    },
+    {
+      "epoch": 0.9784258889332801,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 2449
+    },
+    {
+      "epoch": 0.9788254095085897,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9792249300838993,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2451
+    },
+    {
+      "epoch": 0.979624450659209,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9800239712345186,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 2453
+    },
+    {
+      "epoch": 0.9804234918098282,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9808230123851378,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2455
+    },
+    {
+      "epoch": 0.9812225329604475,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 2456
+    },
+    {
+      "epoch": 0.981622053535757,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9820215741110667,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9824210946863764,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0762,
+      "step": 2459
+    },
+    {
+      "epoch": 0.982820615261686,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9832201358369956,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9836196564123052,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9840191769876149,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0938,
+      "step": 2463
+    },
+    {
+      "epoch": 0.9844186975629244,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9848182181382341,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 2465
+    },
+    {
+      "epoch": 0.9852177387135438,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9856172592888534,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 2467
+    },
+    {
+      "epoch": 0.986016779864163,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.09,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9864163004394726,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2469
+    },
+    {
+      "epoch": 0.9868158210147823,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9872153415900919,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 2471
+    },
+    {
+      "epoch": 0.9876148621654015,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9880143827407112,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 2473
+    },
+    {
+      "epoch": 0.9884139033160207,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.091,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9888134238913304,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 2475
+    },
+    {
+      "epoch": 0.98921294446664,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9896124650419497,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9900119856172593,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9904115061925689,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2479
+    },
+    {
+      "epoch": 0.9908110267678786,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0894,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9912105473431881,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0801,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9916100679184978,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9920095884938074,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9924091090691171,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9928086296444267,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 2485
+    },
+    {
+      "epoch": 0.9932081502197363,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0948,
+      "step": 2486
+    },
+    {
+      "epoch": 0.993607670795046,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 2487
+    },
+    {
+      "epoch": 0.9940071913703555,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9944067119456652,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 2489
+    },
+    {
+      "epoch": 0.9948062325209748,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9952057530962845,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0937,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9956052736715941,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9960047942469037,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 2493
+    },
+    {
+      "epoch": 0.9964043148222134,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9968038353975229,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9972033559728326,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9976028765481423,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9980023971234518,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9984019176987615,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9988014382740711,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9992009588493808,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9996004794246903,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 2502
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2503
+    },
+    {
+      "epoch": 1.0003995205753096,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0007990411506194,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2505
+    },
+    {
+      "epoch": 1.001198561725929,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0861,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0015980823012385,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0797,
+      "step": 2507
+    },
+    {
+      "epoch": 1.001997602876548,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 2508
+    },
+    {
+      "epoch": 1.0023971234518578,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 2509
+    },
+    {
+      "epoch": 1.0027966440271674,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2510
+    },
+    {
+      "epoch": 1.003196164602477,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0874,
+      "step": 2511
+    },
+    {
+      "epoch": 1.0035956851777867,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0039952057530963,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 2513
+    },
+    {
+      "epoch": 1.0043947263284059,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0047942469037154,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0845,
+      "step": 2515
+    },
+    {
+      "epoch": 1.0051937674790252,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0055932880543348,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 2517
+    },
+    {
+      "epoch": 1.0059928086296444,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0063923292049541,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0889,
+      "step": 2519
+    },
+    {
+      "epoch": 1.0067918497802637,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0071913703555733,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 2521
+    },
+    {
+      "epoch": 1.0075908909308828,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.089,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0079904115061926,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 2523
+    },
+    {
+      "epoch": 1.0083899320815022,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0087894526568117,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 2525
+    },
+    {
+      "epoch": 1.0091889732321215,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 2526
+    },
+    {
+      "epoch": 1.009588493807431,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 2527
+    },
+    {
+      "epoch": 1.0099880143827407,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 2528
+    },
+    {
+      "epoch": 1.0103875349580504,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 2529
+    },
+    {
+      "epoch": 1.01078705553336,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0111865761086696,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 2531
+    },
+    {
+      "epoch": 1.0115860966839791,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 2532
+    },
+    {
+      "epoch": 1.011985617259289,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 2533
+    },
+    {
+      "epoch": 1.0123851378345985,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.1004,
+      "step": 2534
+    },
+    {
+      "epoch": 1.012784658409908,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 2535
+    },
+    {
+      "epoch": 1.0131841789852178,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0879,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0135836995605274,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 2537
+    },
+    {
+      "epoch": 1.013983220135837,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0143827407111465,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 2539
+    },
+    {
+      "epoch": 1.0147822612864563,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0889,
+      "step": 2540
+    },
+    {
+      "epoch": 1.0151817818617659,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 2541
+    },
+    {
+      "epoch": 1.0155813024370755,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0159808230123852,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 2543
+    },
+    {
+      "epoch": 1.0163803435876948,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0167798641630044,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 2545
+    },
+    {
+      "epoch": 1.017179384738314,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 2546
+    },
+    {
+      "epoch": 1.0175789053136237,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 2547
+    },
+    {
+      "epoch": 1.0179784258889333,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 2548
+    },
+    {
+      "epoch": 1.0183779464642428,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 2549
+    },
+    {
+      "epoch": 1.0187774670395526,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 2550
+    },
+    {
+      "epoch": 1.0191769876148622,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 2551
+    },
+    {
+      "epoch": 1.0195765081901718,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 2552
+    },
+    {
+      "epoch": 1.0199760287654813,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 2553
+    },
+    {
+      "epoch": 1.020375549340791,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0901,
+      "step": 2554
+    },
+    {
+      "epoch": 1.0207750699161007,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 2555
+    },
+    {
+      "epoch": 1.0211745904914102,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 2556
+    },
+    {
+      "epoch": 1.02157411106672,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 2557
+    },
+    {
+      "epoch": 1.0219736316420296,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 2558
+    },
+    {
+      "epoch": 1.0223731522173392,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 2559
+    },
+    {
+      "epoch": 1.0227726727926487,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 2560
+    },
+    {
+      "epoch": 1.0231721933679585,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 2561
+    },
+    {
+      "epoch": 1.023571713943268,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 2562
+    },
+    {
+      "epoch": 1.0239712345185776,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0911,
+      "step": 2563
+    },
+    {
+      "epoch": 1.0243707550938874,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 2564
+    },
+    {
+      "epoch": 1.024770275669197,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 2565
+    },
+    {
+      "epoch": 1.0251697962445065,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 2566
+    },
+    {
+      "epoch": 1.0255693168198161,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 2567
+    },
+    {
+      "epoch": 1.025968837395126,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 2568
+    },
+    {
+      "epoch": 1.0263683579704355,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 2569
+    },
+    {
+      "epoch": 1.026767878545745,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2570
+    },
+    {
+      "epoch": 1.0271673991210548,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 2571
+    },
+    {
+      "epoch": 1.0275669196963644,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2572
+    },
+    {
+      "epoch": 1.027966440271674,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 2573
+    },
+    {
+      "epoch": 1.0283659608469837,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 2574
+    },
+    {
+      "epoch": 1.0287654814222933,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 2575
+    },
+    {
+      "epoch": 1.0291650019976029,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 2576
+    },
+    {
+      "epoch": 1.0295645225729124,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0834,
+      "step": 2577
+    },
+    {
+      "epoch": 1.0299640431482222,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 2578
+    },
+    {
+      "epoch": 1.0303635637235318,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 2579
+    },
+    {
+      "epoch": 1.0307630842988413,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 2580
+    },
+    {
+      "epoch": 1.0311626048741511,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 2581
+    },
+    {
+      "epoch": 1.0315621254494607,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 2582
+    },
+    {
+      "epoch": 1.0319616460247703,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2583
+    },
+    {
+      "epoch": 1.0323611666000798,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2584
+    },
+    {
+      "epoch": 1.0327606871753896,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0797,
+      "step": 2585
+    },
+    {
+      "epoch": 1.0331602077506992,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 2586
+    },
+    {
+      "epoch": 1.0335597283260087,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 2587
+    },
+    {
+      "epoch": 1.0339592489013185,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 2588
+    },
+    {
+      "epoch": 1.034358769476628,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2589
+    },
+    {
+      "epoch": 1.0347582900519376,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0889,
+      "step": 2590
+    },
+    {
+      "epoch": 1.0351578106272472,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 2591
+    },
+    {
+      "epoch": 1.035557331202557,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 2592
+    },
+    {
+      "epoch": 1.0359568517778666,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 2593
+    },
+    {
+      "epoch": 1.0363563723531761,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 2594
+    },
+    {
+      "epoch": 1.036755892928486,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 2595
+    },
+    {
+      "epoch": 1.0371554135037955,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 2596
+    },
+    {
+      "epoch": 1.037554934079105,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2597
+    },
+    {
+      "epoch": 1.0379544546544146,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 2598
+    },
+    {
+      "epoch": 1.0383539752297244,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0883,
+      "step": 2599
+    },
+    {
+      "epoch": 1.038753495805034,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2600
+    },
+    {
+      "epoch": 1.0391530163803435,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 2601
+    },
+    {
+      "epoch": 1.0395525369556533,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 2602
+    },
+    {
+      "epoch": 1.0399520575309629,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2603
+    },
+    {
+      "epoch": 1.0403515781062724,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 2604
+    },
+    {
+      "epoch": 1.040751098681582,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 2605
+    },
+    {
+      "epoch": 1.0411506192568918,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 2606
+    },
+    {
+      "epoch": 1.0415501398322013,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 2607
+    },
+    {
+      "epoch": 1.041949660407511,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 2608
+    },
+    {
+      "epoch": 1.0423491809828207,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2609
+    },
+    {
+      "epoch": 1.0427487015581303,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2610
+    },
+    {
+      "epoch": 1.0431482221334398,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 2611
+    },
+    {
+      "epoch": 1.0435477427087494,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 2612
+    },
+    {
+      "epoch": 1.0439472632840592,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 2613
+    },
+    {
+      "epoch": 1.0443467838593687,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 2614
+    },
+    {
+      "epoch": 1.0447463044346783,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 2615
+    },
+    {
+      "epoch": 1.045145825009988,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0888,
+      "step": 2616
+    },
+    {
+      "epoch": 1.0455453455852977,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 2617
+    },
+    {
+      "epoch": 1.0459448661606072,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 2618
+    },
+    {
+      "epoch": 1.046344386735917,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 2619
+    },
+    {
+      "epoch": 1.0467439073112266,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 2620
+    },
+    {
+      "epoch": 1.0471434278865361,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 2621
+    },
+    {
+      "epoch": 1.0475429484618457,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 2622
+    },
+    {
+      "epoch": 1.0479424690371555,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 2623
+    },
+    {
+      "epoch": 1.048341989612465,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2624
+    },
+    {
+      "epoch": 1.0487415101877746,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 2625
+    },
+    {
+      "epoch": 1.0491410307630844,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 2626
+    },
+    {
+      "epoch": 1.049540551338394,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2627
+    },
+    {
+      "epoch": 1.0499400719137035,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 2628
+    },
+    {
+      "epoch": 1.050339592489013,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2629
+    },
+    {
+      "epoch": 1.0507391130643229,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0859,
+      "step": 2630
+    },
+    {
+      "epoch": 1.0511386336396324,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 2631
+    },
+    {
+      "epoch": 1.051538154214942,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 2632
+    },
+    {
+      "epoch": 1.0519376747902518,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0862,
+      "step": 2633
+    },
+    {
+      "epoch": 1.0523371953655614,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2634
+    },
+    {
+      "epoch": 1.052736715940871,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 2635
+    },
+    {
+      "epoch": 1.0531362365161805,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 2636
+    },
+    {
+      "epoch": 1.0535357570914903,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 2637
+    },
+    {
+      "epoch": 1.0539352776667998,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 2638
+    },
+    {
+      "epoch": 1.0543347982421094,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0829,
+      "step": 2639
+    },
+    {
+      "epoch": 1.0547343188174192,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 2640
+    },
+    {
+      "epoch": 1.0551338393927288,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 2641
+    },
+    {
+      "epoch": 1.0555333599680383,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 2642
+    },
+    {
+      "epoch": 1.0559328805433479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2643
+    },
+    {
+      "epoch": 1.0563324011186577,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 2644
+    },
+    {
+      "epoch": 1.0567319216939672,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 2645
+    },
+    {
+      "epoch": 1.0571314422692768,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 2646
+    },
+    {
+      "epoch": 1.0575309628445866,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 2647
+    },
+    {
+      "epoch": 1.0579304834198962,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 2648
+    },
+    {
+      "epoch": 1.0583300039952057,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 2649
+    },
+    {
+      "epoch": 1.0587295245705153,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 2650
+    },
+    {
+      "epoch": 1.059129045145825,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2651
+    },
+    {
+      "epoch": 1.0595285657211346,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2652
+    },
+    {
+      "epoch": 1.0599280862964442,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2653
+    },
+    {
+      "epoch": 1.060327606871754,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2654
+    },
+    {
+      "epoch": 1.0607271274470635,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 2655
+    },
+    {
+      "epoch": 1.061126648022373,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 2656
+    },
+    {
+      "epoch": 1.0615261685976827,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0859,
+      "step": 2657
+    },
+    {
+      "epoch": 1.0619256891729925,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2658
+    },
+    {
+      "epoch": 1.062325209748302,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2659
+    },
+    {
+      "epoch": 1.0627247303236116,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0631242508989214,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 2661
+    },
+    {
+      "epoch": 1.063523771474231,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2662
+    },
+    {
+      "epoch": 1.0639232920495405,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 2663
+    },
+    {
+      "epoch": 1.0643228126248503,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 2664
+    },
+    {
+      "epoch": 1.0647223332001599,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2665
+    },
+    {
+      "epoch": 1.0651218537754694,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 2666
+    },
+    {
+      "epoch": 1.065521374350779,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 2667
+    },
+    {
+      "epoch": 1.0659208949260888,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 2668
+    },
+    {
+      "epoch": 1.0663204155013983,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 2669
+    },
+    {
+      "epoch": 1.066719936076708,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0671194566520175,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 2671
+    },
+    {
+      "epoch": 1.0675189772273272,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 2672
+    },
+    {
+      "epoch": 1.0679184978026368,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 2673
+    },
+    {
+      "epoch": 1.0683180183779464,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 2674
+    },
+    {
+      "epoch": 1.0687175389532562,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 2675
+    },
+    {
+      "epoch": 1.0691170595285657,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 2676
+    },
+    {
+      "epoch": 1.0695165801038753,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 2677
+    },
+    {
+      "epoch": 1.069916100679185,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 2678
+    },
+    {
+      "epoch": 1.0703156212544946,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 2679
+    },
+    {
+      "epoch": 1.0707151418298042,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0711146624051138,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 2681
+    },
+    {
+      "epoch": 1.0715141829804236,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 2682
+    },
+    {
+      "epoch": 1.0719137035557331,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 2683
+    },
+    {
+      "epoch": 1.0723132241310427,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 2684
+    },
+    {
+      "epoch": 1.0727127447063525,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 2685
+    },
+    {
+      "epoch": 1.073112265281662,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 2686
+    },
+    {
+      "epoch": 1.0735117858569716,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2687
+    },
+    {
+      "epoch": 1.0739113064322812,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 2688
+    },
+    {
+      "epoch": 1.074310827007591,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 2689
+    },
+    {
+      "epoch": 1.0747103475829005,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2690
+    },
+    {
+      "epoch": 1.07510986815821,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 2691
+    },
+    {
+      "epoch": 1.0755093887335199,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 2692
+    },
+    {
+      "epoch": 1.0759089093088294,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 2693
+    },
+    {
+      "epoch": 1.076308429884139,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2694
+    },
+    {
+      "epoch": 1.0767079504594486,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2695
+    },
+    {
+      "epoch": 1.0771074710347583,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2696
+    },
+    {
+      "epoch": 1.077506991610068,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 2697
+    },
+    {
+      "epoch": 1.0779065121853775,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0836,
+      "step": 2698
+    },
+    {
+      "epoch": 1.0783060327606873,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 2699
+    },
+    {
+      "epoch": 1.0787055533359968,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0872,
+      "step": 2700
+    },
+    {
+      "epoch": 1.0791050739113064,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 2701
+    },
+    {
+      "epoch": 1.079504594486616,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 2702
+    },
+    {
+      "epoch": 1.0799041150619257,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2703
+    },
+    {
+      "epoch": 1.0803036356372353,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2704
+    },
+    {
+      "epoch": 1.0807031562125449,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2705
+    },
+    {
+      "epoch": 1.0811026767878547,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 2706
+    },
+    {
+      "epoch": 1.0815021973631642,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 2707
+    },
+    {
+      "epoch": 1.0819017179384738,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 2708
+    },
+    {
+      "epoch": 1.0823012385137836,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 2709
+    },
+    {
+      "epoch": 1.0827007590890931,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 2710
+    },
+    {
+      "epoch": 1.0831002796644027,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 2711
+    },
+    {
+      "epoch": 1.0834998002397123,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 2712
+    },
+    {
+      "epoch": 1.083899320815022,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 2713
+    },
+    {
+      "epoch": 1.0842988413903316,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 2714
+    },
+    {
+      "epoch": 1.0846983619656412,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0869,
+      "step": 2715
+    },
+    {
+      "epoch": 1.0850978825409507,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 2716
+    },
+    {
+      "epoch": 1.0854974031162605,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 2717
+    },
+    {
+      "epoch": 1.08589692369157,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 2718
+    },
+    {
+      "epoch": 1.0862964442668797,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 2719
+    },
+    {
+      "epoch": 1.0866959648421894,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0885,
+      "step": 2720
+    },
+    {
+      "epoch": 1.087095485417499,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2721
+    },
+    {
+      "epoch": 1.0874950059928086,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 2722
+    },
+    {
+      "epoch": 1.0878945265681184,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 2723
+    },
+    {
+      "epoch": 1.088294047143428,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 2724
+    },
+    {
+      "epoch": 1.0886935677187375,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 2725
+    },
+    {
+      "epoch": 1.089093088294047,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 2726
+    },
+    {
+      "epoch": 1.0894926088693568,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 2727
+    },
+    {
+      "epoch": 1.0898921294446664,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 2728
+    },
+    {
+      "epoch": 1.090291650019976,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2729
+    },
+    {
+      "epoch": 1.0906911705952858,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2730
+    },
+    {
+      "epoch": 1.0910906911705953,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2731
+    },
+    {
+      "epoch": 1.0914902117459049,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 2732
+    },
+    {
+      "epoch": 1.0918897323212144,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 2733
+    },
+    {
+      "epoch": 1.0922892528965242,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 2734
+    },
+    {
+      "epoch": 1.0926887734718338,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 2735
+    },
+    {
+      "epoch": 1.0930882940471434,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.09,
+      "step": 2736
+    },
+    {
+      "epoch": 1.0934878146224531,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 2737
+    },
+    {
+      "epoch": 1.0938873351977627,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2738
+    },
+    {
+      "epoch": 1.0942868557730723,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2739
+    },
+    {
+      "epoch": 1.0946863763483818,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 2740
+    },
+    {
+      "epoch": 1.0950858969236916,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 2741
+    },
+    {
+      "epoch": 1.0954854174990012,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 2742
+    },
+    {
+      "epoch": 1.0958849380743108,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0909,
+      "step": 2743
+    },
+    {
+      "epoch": 1.0962844586496205,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 2744
+    },
+    {
+      "epoch": 1.09668397922493,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 2745
+    },
+    {
+      "epoch": 1.0970834998002397,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 2746
+    },
+    {
+      "epoch": 1.0974830203755492,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 2747
+    },
+    {
+      "epoch": 1.097882540950859,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 2748
+    },
+    {
+      "epoch": 1.0982820615261686,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0828,
+      "step": 2749
+    },
+    {
+      "epoch": 1.0986815821014781,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 2750
+    },
+    {
+      "epoch": 1.099081102676788,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 2751
+    },
+    {
+      "epoch": 1.0994806232520975,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 2752
+    },
+    {
+      "epoch": 1.099880143827407,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 2753
+    },
+    {
+      "epoch": 1.1002796644027169,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1006791849780264,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 2755
+    },
+    {
+      "epoch": 1.101078705553336,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 2756
+    },
+    {
+      "epoch": 1.1014782261286455,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 2757
+    },
+    {
+      "epoch": 1.1018777467039553,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2758
+    },
+    {
+      "epoch": 1.102277267279265,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2759
+    },
+    {
+      "epoch": 1.1026767878545745,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 2760
+    },
+    {
+      "epoch": 1.103076308429884,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 2761
+    },
+    {
+      "epoch": 1.1034758290051938,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0875,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1038753495805034,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 2763
+    },
+    {
+      "epoch": 1.104274870155813,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 2764
+    },
+    {
+      "epoch": 1.1046743907311227,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 2765
+    },
+    {
+      "epoch": 1.1050739113064323,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1054734318817419,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2767
+    },
+    {
+      "epoch": 1.1058729524570516,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1062724730323612,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 2769
+    },
+    {
+      "epoch": 1.1066719936076708,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0852,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1070715141829803,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 2771
+    },
+    {
+      "epoch": 1.1074710347582901,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1078705553335997,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 2773
+    },
+    {
+      "epoch": 1.1082700759089092,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 2774
+    },
+    {
+      "epoch": 1.108669596484219,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2775
+    },
+    {
+      "epoch": 1.1090691170595286,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1094686376348382,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 2777
+    },
+    {
+      "epoch": 1.1098681582101477,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 2778
+    },
+    {
+      "epoch": 1.1102676787854575,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2779
+    },
+    {
+      "epoch": 1.110667199360767,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 2780
+    },
+    {
+      "epoch": 1.1110667199360766,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 2781
+    },
+    {
+      "epoch": 1.1114662405113864,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 2782
+    },
+    {
+      "epoch": 1.111865761086696,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 2783
+    },
+    {
+      "epoch": 1.1122652816620056,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 2784
+    },
+    {
+      "epoch": 1.1126648022373151,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 2785
+    },
+    {
+      "epoch": 1.113064322812625,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 2786
+    },
+    {
+      "epoch": 1.1134638433879345,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 2787
+    },
+    {
+      "epoch": 1.113863363963244,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 2788
+    },
+    {
+      "epoch": 1.1142628845385538,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 2789
+    },
+    {
+      "epoch": 1.1146624051138634,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 2790
+    },
+    {
+      "epoch": 1.115061925689173,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0877,
+      "step": 2791
+    },
+    {
+      "epoch": 1.1154614462644825,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0841,
+      "step": 2792
+    },
+    {
+      "epoch": 1.1158609668397923,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 2793
+    },
+    {
+      "epoch": 1.1162604874151019,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2794
+    },
+    {
+      "epoch": 1.1166600079904114,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2795
+    },
+    {
+      "epoch": 1.1170595285657212,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 2796
+    },
+    {
+      "epoch": 1.1174590491410308,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 2797
+    },
+    {
+      "epoch": 1.1178585697163403,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 2798
+    },
+    {
+      "epoch": 1.1182580902916501,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 2799
+    },
+    {
+      "epoch": 1.1186576108669597,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 2800
+    },
+    {
+      "epoch": 1.1190571314422693,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 2801
+    },
+    {
+      "epoch": 1.1194566520175788,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0838,
+      "step": 2802
+    },
+    {
+      "epoch": 1.1198561725928886,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 2803
+    },
+    {
+      "epoch": 1.1202556931681982,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 2804
+    },
+    {
+      "epoch": 1.1206552137435077,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2805
+    },
+    {
+      "epoch": 1.1210547343188173,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 2806
+    },
+    {
+      "epoch": 1.121454254894127,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 2807
+    },
+    {
+      "epoch": 1.1218537754694367,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 2808
+    },
+    {
+      "epoch": 1.1222532960447462,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 2809
+    },
+    {
+      "epoch": 1.122652816620056,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 2810
+    },
+    {
+      "epoch": 1.1230523371953656,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 2811
+    },
+    {
+      "epoch": 1.1234518577706751,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2812
+    },
+    {
+      "epoch": 1.123851378345985,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 2813
+    },
+    {
+      "epoch": 1.1242508989212945,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 2814
+    },
+    {
+      "epoch": 1.124650419496604,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2815
+    },
+    {
+      "epoch": 1.1250499400719136,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 2816
+    },
+    {
+      "epoch": 1.1254494606472234,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 2817
+    },
+    {
+      "epoch": 1.125848981222533,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 2818
+    },
+    {
+      "epoch": 1.1262485017978425,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 2819
+    },
+    {
+      "epoch": 1.1266480223731523,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0831,
+      "step": 2820
+    },
+    {
+      "epoch": 1.1270475429484619,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 2821
+    },
+    {
+      "epoch": 1.1274470635237714,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 2822
+    },
+    {
+      "epoch": 1.127846584099081,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 2823
+    },
+    {
+      "epoch": 1.1282461046743908,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 2824
+    },
+    {
+      "epoch": 1.1286456252497004,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 2825
+    },
+    {
+      "epoch": 1.12904514582501,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0834,
+      "step": 2826
+    },
+    {
+      "epoch": 1.1294446664003197,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 2827
+    },
+    {
+      "epoch": 1.1298441869756293,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0829,
+      "step": 2828
+    },
+    {
+      "epoch": 1.1302437075509388,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0759,
+      "step": 2829
+    },
+    {
+      "epoch": 1.1306432281262486,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2830
+    },
+    {
+      "epoch": 1.1310427487015582,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 2831
+    },
+    {
+      "epoch": 1.1314422692768678,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2832
+    },
+    {
+      "epoch": 1.1318417898521773,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 2833
+    },
+    {
+      "epoch": 1.132241310427487,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 2834
+    },
+    {
+      "epoch": 1.1326408310027967,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 2835
+    },
+    {
+      "epoch": 1.1330403515781062,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 2836
+    },
+    {
+      "epoch": 1.1334398721534158,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 2837
+    },
+    {
+      "epoch": 1.1338393927287256,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0847,
+      "step": 2838
+    },
+    {
+      "epoch": 1.1342389133040351,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 2839
+    },
+    {
+      "epoch": 1.1346384338793447,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 2840
+    },
+    {
+      "epoch": 1.1350379544546545,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 2841
+    },
+    {
+      "epoch": 1.135437475029964,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 2842
+    },
+    {
+      "epoch": 1.1358369956052736,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 2843
+    },
+    {
+      "epoch": 1.1362365161805834,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0849,
+      "step": 2844
+    },
+    {
+      "epoch": 1.136636036755893,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2845
+    },
+    {
+      "epoch": 1.1370355573312025,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0848,
+      "step": 2846
+    },
+    {
+      "epoch": 1.137435077906512,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 2847
+    },
+    {
+      "epoch": 1.137834598481822,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.0828,
+      "step": 2848
+    },
+    {
+      "epoch": 1.1382341190571315,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 2849
+    },
+    {
+      "epoch": 1.138633639632441,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 2850
+    },
+    {
+      "epoch": 1.1390331602077506,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0882,
+      "step": 2851
+    },
+    {
+      "epoch": 1.1394326807830604,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 2852
+    },
+    {
+      "epoch": 1.13983220135837,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2853
+    },
+    {
+      "epoch": 1.1402317219336795,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 2854
+    },
+    {
+      "epoch": 1.1406312425089893,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 2855
+    },
+    {
+      "epoch": 1.1410307630842988,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2856
+    },
+    {
+      "epoch": 1.1414302836596084,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 2857
+    },
+    {
+      "epoch": 1.1418298042349182,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 2858
+    },
+    {
+      "epoch": 1.1422293248102278,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 2859
+    },
+    {
+      "epoch": 1.1426288453855373,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 2860
+    },
+    {
+      "epoch": 1.143028365960847,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 2861
+    },
+    {
+      "epoch": 1.1434278865361567,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 2862
+    },
+    {
+      "epoch": 1.1438274071114662,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 2863
+    },
+    {
+      "epoch": 1.1442269276867758,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 2864
+    },
+    {
+      "epoch": 1.1446264482620856,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2865
+    },
+    {
+      "epoch": 1.1450259688373952,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 2866
+    },
+    {
+      "epoch": 1.1454254894127047,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2867
+    },
+    {
+      "epoch": 1.1458250099880143,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 2868
+    },
+    {
+      "epoch": 1.146224530563324,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 2869
+    },
+    {
+      "epoch": 1.1466240511386336,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 2870
+    },
+    {
+      "epoch": 1.1470235717139432,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 2871
+    },
+    {
+      "epoch": 1.147423092289253,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 2872
+    },
+    {
+      "epoch": 1.1478226128645626,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 2873
+    },
+    {
+      "epoch": 1.1482221334398721,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 2874
+    },
+    {
+      "epoch": 1.148621654015182,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0876,
+      "step": 2875
+    },
+    {
+      "epoch": 1.1490211745904915,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 2876
+    },
+    {
+      "epoch": 1.149420695165801,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 2877
+    },
+    {
+      "epoch": 1.1498202157411106,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0956,
+      "step": 2878
+    },
+    {
+      "epoch": 1.1502197363164204,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 2879
+    },
+    {
+      "epoch": 1.15061925689173,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 2880
+    },
+    {
+      "epoch": 1.1510187774670395,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 2881
+    },
+    {
+      "epoch": 1.151418298042349,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 2882
+    },
+    {
+      "epoch": 1.1518178186176589,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 2883
+    },
+    {
+      "epoch": 1.1522173391929684,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 2884
+    },
+    {
+      "epoch": 1.152616859768278,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 2885
+    },
+    {
+      "epoch": 1.1530163803435878,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2886
+    },
+    {
+      "epoch": 1.1534159009188973,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 2887
+    },
+    {
+      "epoch": 1.153815421494207,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 2888
+    },
+    {
+      "epoch": 1.1542149420695167,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0801,
+      "step": 2889
+    },
+    {
+      "epoch": 1.1546144626448263,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0762,
+      "step": 2890
+    },
+    {
+      "epoch": 1.1550139832201358,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 2891
+    },
+    {
+      "epoch": 1.1554135037954454,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 2892
+    },
+    {
+      "epoch": 1.1558130243707552,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 2893
+    },
+    {
+      "epoch": 1.1562125449460647,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0856,
+      "step": 2894
+    },
+    {
+      "epoch": 1.1566120655213743,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 2895
+    },
+    {
+      "epoch": 1.1570115860966839,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 2896
+    },
+    {
+      "epoch": 1.1574111066719937,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 2897
+    },
+    {
+      "epoch": 1.1578106272473032,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 2898
+    },
+    {
+      "epoch": 1.1582101478226128,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 2899
+    },
+    {
+      "epoch": 1.1586096683979226,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 2900
+    },
+    {
+      "epoch": 1.1590091889732321,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 2901
+    },
+    {
+      "epoch": 1.1594087095485417,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 2902
+    },
+    {
+      "epoch": 1.1598082301238515,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 2903
+    },
+    {
+      "epoch": 1.160207750699161,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0778,
+      "step": 2904
+    },
+    {
+      "epoch": 1.1606072712744706,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 2905
+    },
+    {
+      "epoch": 1.1610067918497802,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 2906
+    },
+    {
+      "epoch": 1.16140631242509,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 2907
+    },
+    {
+      "epoch": 1.1618058330003995,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 2908
+    },
+    {
+      "epoch": 1.162205353575709,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 2909
+    },
+    {
+      "epoch": 1.1626048741510189,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 2910
+    },
+    {
+      "epoch": 1.1630043947263284,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 2911
+    },
+    {
+      "epoch": 1.163403915301638,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 2912
+    },
+    {
+      "epoch": 1.1638034358769476,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 2913
+    },
+    {
+      "epoch": 1.1642029564522574,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 2914
+    },
+    {
+      "epoch": 1.164602477027567,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 2915
+    },
+    {
+      "epoch": 1.1650019976028765,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 2916
+    },
+    {
+      "epoch": 1.1654015181781863,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 2917
+    },
+    {
+      "epoch": 1.1658010387534958,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 2918
+    },
+    {
+      "epoch": 1.1662005593288054,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0914,
+      "step": 2919
+    },
+    {
+      "epoch": 1.1666000799041152,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 2920
+    },
+    {
+      "epoch": 1.1669996004794247,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 2921
+    },
+    {
+      "epoch": 1.1673991210547343,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 2922
+    },
+    {
+      "epoch": 1.1677986416300439,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 2923
+    },
+    {
+      "epoch": 1.1681981622053537,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 2924
+    },
+    {
+      "epoch": 1.1685976827806632,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 2925
+    },
+    {
+      "epoch": 1.1689972033559728,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 2926
+    },
+    {
+      "epoch": 1.1693967239312824,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 2927
+    },
+    {
+      "epoch": 1.1697962445065921,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 2928
+    },
+    {
+      "epoch": 1.1701957650819017,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 2929
+    },
+    {
+      "epoch": 1.1705952856572113,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 2930
+    },
+    {
+      "epoch": 1.170994806232521,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 2931
+    },
+    {
+      "epoch": 1.1713943268078306,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 2932
+    },
+    {
+      "epoch": 1.1717938473831402,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 2933
+    },
+    {
+      "epoch": 1.17219336795845,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 2934
+    },
+    {
+      "epoch": 1.1725928885337595,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 2935
+    },
+    {
+      "epoch": 1.172992409109069,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 2936
+    },
+    {
+      "epoch": 1.1733919296843787,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 2937
+    },
+    {
+      "epoch": 1.1737914502596885,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0906,
+      "step": 2938
+    },
+    {
+      "epoch": 1.174190970834998,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2939
+    },
+    {
+      "epoch": 1.1745904914103076,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1749900119856171,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 2941
+    },
+    {
+      "epoch": 1.175389532560927,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 2942
+    },
+    {
+      "epoch": 1.1757890531362365,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 2943
+    },
+    {
+      "epoch": 1.176188573711546,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 2944
+    },
+    {
+      "epoch": 1.1765880942868558,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 2945
+    },
+    {
+      "epoch": 1.1769876148621654,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 2946
+    },
+    {
+      "epoch": 1.177387135437475,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 2947
+    },
+    {
+      "epoch": 1.1777866560127848,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 2948
+    },
+    {
+      "epoch": 1.1781861765880943,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 2949
+    },
+    {
+      "epoch": 1.178585697163404,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0836,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1789852177387135,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 2951
+    },
+    {
+      "epoch": 1.1793847383140232,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 2952
+    },
+    {
+      "epoch": 1.1797842588893328,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 2953
+    },
+    {
+      "epoch": 1.1801837794646424,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 2954
+    },
+    {
+      "epoch": 1.180583300039952,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 2955
+    },
+    {
+      "epoch": 1.1809828206152617,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 2956
+    },
+    {
+      "epoch": 1.1813823411905713,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 2957
+    },
+    {
+      "epoch": 1.1817818617658808,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 2958
+    },
+    {
+      "epoch": 1.1821813823411906,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0859,
+      "step": 2959
+    },
+    {
+      "epoch": 1.1825809029165002,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 2960
+    },
+    {
+      "epoch": 1.1829804234918098,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2961
+    },
+    {
+      "epoch": 1.1833799440671195,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 2962
+    },
+    {
+      "epoch": 1.1837794646424291,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 2963
+    },
+    {
+      "epoch": 1.1841789852177387,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 2964
+    },
+    {
+      "epoch": 1.1845785057930485,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 2965
+    },
+    {
+      "epoch": 1.184978026368358,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 2966
+    },
+    {
+      "epoch": 1.1853775469436676,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 2967
+    },
+    {
+      "epoch": 1.1857770675189772,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2968
+    },
+    {
+      "epoch": 1.186176588094287,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 2969
+    },
+    {
+      "epoch": 1.1865761086695965,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 2970
+    },
+    {
+      "epoch": 1.186975629244906,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 2971
+    },
+    {
+      "epoch": 1.1873751498202156,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 2972
+    },
+    {
+      "epoch": 1.1877746703955254,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 2973
+    },
+    {
+      "epoch": 1.188174190970835,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 2974
+    },
+    {
+      "epoch": 1.1885737115461446,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0846,
+      "step": 2975
+    },
+    {
+      "epoch": 1.1889732321214543,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 2976
+    },
+    {
+      "epoch": 1.189372752696764,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 2977
+    },
+    {
+      "epoch": 1.1897722732720735,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 2978
+    },
+    {
+      "epoch": 1.1901717938473833,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 2979
+    },
+    {
+      "epoch": 1.1905713144226928,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 2980
+    },
+    {
+      "epoch": 1.1909708349980024,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 2981
+    },
+    {
+      "epoch": 1.191370355573312,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 2982
+    },
+    {
+      "epoch": 1.1917698761486217,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0834,
+      "step": 2983
+    },
+    {
+      "epoch": 1.1921693967239313,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 2984
+    },
+    {
+      "epoch": 1.1925689172992409,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 2985
+    },
+    {
+      "epoch": 1.1929684378745504,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 2986
+    },
+    {
+      "epoch": 1.1933679584498602,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 2987
+    },
+    {
+      "epoch": 1.1937674790251698,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 2988
+    },
+    {
+      "epoch": 1.1941669996004793,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 2989
+    },
+    {
+      "epoch": 1.1945665201757891,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 2990
+    },
+    {
+      "epoch": 1.1949660407510987,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 2991
+    },
+    {
+      "epoch": 1.1953655613264083,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 2992
+    },
+    {
+      "epoch": 1.195765081901718,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 2993
+    },
+    {
+      "epoch": 1.1961646024770276,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 2994
+    },
+    {
+      "epoch": 1.1965641230523372,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 2995
+    },
+    {
+      "epoch": 1.1969636436276467,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 2996
+    },
+    {
+      "epoch": 1.1973631642029565,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 2997
+    },
+    {
+      "epoch": 1.197762684778266,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0858,
+      "step": 2998
+    },
+    {
+      "epoch": 1.1981622053535756,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 2999
+    },
+    {
+      "epoch": 1.1985617259288852,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 3000
+    },
+    {
+      "epoch": 1.198961246504195,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0763,
+      "step": 3001
+    },
+    {
+      "epoch": 1.1993607670795046,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 3002
+    },
+    {
+      "epoch": 1.1997602876548141,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 3003
+    },
+    {
+      "epoch": 1.200159808230124,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0884,
+      "step": 3004
+    },
+    {
+      "epoch": 1.2005593288054335,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3005
+    },
+    {
+      "epoch": 1.200958849380743,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3006
+    },
+    {
+      "epoch": 1.2013583699560528,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 3007
+    },
+    {
+      "epoch": 1.2017578905313624,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3008
+    },
+    {
+      "epoch": 1.202157411106672,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3009
+    },
+    {
+      "epoch": 1.2025569316819817,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2029564522572913,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 3011
+    },
+    {
+      "epoch": 1.2033559728326009,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 3012
+    },
+    {
+      "epoch": 1.2037554934079104,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 3013
+    },
+    {
+      "epoch": 1.2041550139832202,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 3014
+    },
+    {
+      "epoch": 1.2045545345585298,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3015
+    },
+    {
+      "epoch": 1.2049540551338394,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3016
+    },
+    {
+      "epoch": 1.205353575709149,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0853,
+      "step": 3017
+    },
+    {
+      "epoch": 1.2057530962844587,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 3018
+    },
+    {
+      "epoch": 1.2061526168597683,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3019
+    },
+    {
+      "epoch": 1.2065521374350778,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3020
+    },
+    {
+      "epoch": 1.2069516580103876,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 3021
+    },
+    {
+      "epoch": 1.2073511785856972,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 3022
+    },
+    {
+      "epoch": 1.2077506991610067,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 3023
+    },
+    {
+      "epoch": 1.2081502197363165,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 3024
+    },
+    {
+      "epoch": 1.208549740311626,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 3025
+    },
+    {
+      "epoch": 1.2089492608869357,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 3026
+    },
+    {
+      "epoch": 1.2093487814622452,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3027
+    },
+    {
+      "epoch": 1.209748302037555,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3028
+    },
+    {
+      "epoch": 1.2101478226128646,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.088,
+      "step": 3029
+    },
+    {
+      "epoch": 1.2105473431881741,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 3030
+    },
+    {
+      "epoch": 1.2109468637634837,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3031
+    },
+    {
+      "epoch": 1.2113463843387935,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0834,
+      "step": 3032
+    },
+    {
+      "epoch": 1.211745904914103,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 3033
+    },
+    {
+      "epoch": 1.2121454254894126,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 3034
+    },
+    {
+      "epoch": 1.2125449460647224,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 3035
+    },
+    {
+      "epoch": 1.212944466640032,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3036
+    },
+    {
+      "epoch": 1.2133439872153415,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3037
+    },
+    {
+      "epoch": 1.2137435077906513,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 3038
+    },
+    {
+      "epoch": 1.2141430283659609,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3039
+    },
+    {
+      "epoch": 1.2145425489412705,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 3040
+    },
+    {
+      "epoch": 1.21494206951658,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0825,
+      "step": 3041
+    },
+    {
+      "epoch": 1.2153415900918898,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3042
+    },
+    {
+      "epoch": 1.2157411106671994,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3043
+    },
+    {
+      "epoch": 1.216140631242509,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 3044
+    },
+    {
+      "epoch": 1.2165401518178185,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 3045
+    },
+    {
+      "epoch": 1.2169396723931283,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 3046
+    },
+    {
+      "epoch": 1.2173391929684378,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 3047
+    },
+    {
+      "epoch": 1.2177387135437474,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 3048
+    },
+    {
+      "epoch": 1.2181382341190572,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 3049
+    },
+    {
+      "epoch": 1.2185377546943668,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0759,
+      "step": 3050
+    },
+    {
+      "epoch": 1.2189372752696763,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0944,
+      "step": 3051
+    },
+    {
+      "epoch": 1.219336795844986,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3052
+    },
+    {
+      "epoch": 1.2197363164202957,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3053
+    },
+    {
+      "epoch": 1.2201358369956052,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3054
+    },
+    {
+      "epoch": 1.220535357570915,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 3055
+    },
+    {
+      "epoch": 1.2209348781462246,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0829,
+      "step": 3056
+    },
+    {
+      "epoch": 1.2213343987215342,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 3057
+    },
+    {
+      "epoch": 1.2217339192968437,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 3058
+    },
+    {
+      "epoch": 1.2221334398721535,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3059
+    },
+    {
+      "epoch": 1.222532960447463,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 3060
+    },
+    {
+      "epoch": 1.2229324810227726,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 3061
+    },
+    {
+      "epoch": 1.2233320015980822,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 3062
+    },
+    {
+      "epoch": 1.223731522173392,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3063
+    },
+    {
+      "epoch": 1.2241310427487015,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3064
+    },
+    {
+      "epoch": 1.2245305633240111,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3065
+    },
+    {
+      "epoch": 1.224930083899321,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3066
+    },
+    {
+      "epoch": 1.2253296044746305,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 3067
+    },
+    {
+      "epoch": 1.22572912504994,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3068
+    },
+    {
+      "epoch": 1.2261286456252498,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 3069
+    },
+    {
+      "epoch": 1.2265281662005594,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 3070
+    },
+    {
+      "epoch": 1.226927686775869,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0979,
+      "step": 3071
+    },
+    {
+      "epoch": 1.2273272073511785,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3072
+    },
+    {
+      "epoch": 1.2277267279264883,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 3073
+    },
+    {
+      "epoch": 1.2281262485017979,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 3074
+    },
+    {
+      "epoch": 1.2285257690771074,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3075
+    },
+    {
+      "epoch": 1.228925289652417,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 3076
+    },
+    {
+      "epoch": 1.2293248102277268,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 3077
+    },
+    {
+      "epoch": 1.2297243308030363,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3078
+    },
+    {
+      "epoch": 1.230123851378346,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3079
+    },
+    {
+      "epoch": 1.2305233719536557,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 3080
+    },
+    {
+      "epoch": 1.2309228925289653,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3081
+    },
+    {
+      "epoch": 1.2313224131042748,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 3082
+    },
+    {
+      "epoch": 1.2317219336795846,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 3083
+    },
+    {
+      "epoch": 1.2321214542548942,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3084
+    },
+    {
+      "epoch": 1.2325209748302037,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 3085
+    },
+    {
+      "epoch": 1.2329204954055133,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 3086
+    },
+    {
+      "epoch": 1.233320015980823,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 3087
+    },
+    {
+      "epoch": 1.2337195365561326,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3088
+    },
+    {
+      "epoch": 1.2341190571314422,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 3089
+    },
+    {
+      "epoch": 1.2345185777067518,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 3090
+    },
+    {
+      "epoch": 1.2349180982820616,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3091
+    },
+    {
+      "epoch": 1.2353176188573711,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3092
+    },
+    {
+      "epoch": 1.2357171394326807,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3093
+    },
+    {
+      "epoch": 1.2361166600079905,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3094
+    },
+    {
+      "epoch": 1.2365161805833,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 3095
+    },
+    {
+      "epoch": 1.2369157011586096,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 3096
+    },
+    {
+      "epoch": 1.2373152217339194,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 3097
+    },
+    {
+      "epoch": 1.237714742309229,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 3098
+    },
+    {
+      "epoch": 1.2381142628845385,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3099
+    },
+    {
+      "epoch": 1.2385137834598483,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 3100
+    },
+    {
+      "epoch": 1.2389133040351579,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 3101
+    },
+    {
+      "epoch": 1.2393128246104674,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 3102
+    },
+    {
+      "epoch": 1.239712345185777,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 3103
+    },
+    {
+      "epoch": 1.2401118657610868,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3104
+    },
+    {
+      "epoch": 1.2405113863363963,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 3105
+    },
+    {
+      "epoch": 1.240910906911706,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0853,
+      "step": 3106
+    },
+    {
+      "epoch": 1.2413104274870155,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3107
+    },
+    {
+      "epoch": 1.2417099480623253,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 3108
+    },
+    {
+      "epoch": 1.2421094686376348,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 3109
+    },
+    {
+      "epoch": 1.2425089892129444,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3110
+    },
+    {
+      "epoch": 1.2429085097882542,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3111
+    },
+    {
+      "epoch": 1.2433080303635637,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 3112
+    },
+    {
+      "epoch": 1.2437075509388733,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 3113
+    },
+    {
+      "epoch": 1.244107071514183,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 3114
+    },
+    {
+      "epoch": 1.2445065920894927,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3115
+    },
+    {
+      "epoch": 1.2449061126648022,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 3116
+    },
+    {
+      "epoch": 1.2453056332401118,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 3117
+    },
+    {
+      "epoch": 1.2457051538154216,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 3118
+    },
+    {
+      "epoch": 1.2461046743907311,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 3119
+    },
+    {
+      "epoch": 1.2465041949660407,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3120
+    },
+    {
+      "epoch": 1.2469037155413503,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3121
+    },
+    {
+      "epoch": 1.24730323611666,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3122
+    },
+    {
+      "epoch": 1.2477027566919696,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 3123
+    },
+    {
+      "epoch": 1.2481022772672792,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3124
+    },
+    {
+      "epoch": 1.248501797842589,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 3125
+    },
+    {
+      "epoch": 1.2489013184178985,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 3126
+    },
+    {
+      "epoch": 1.249300838993208,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 3127
+    },
+    {
+      "epoch": 1.2497003595685179,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 3128
+    },
+    {
+      "epoch": 1.2500998801438274,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 3129
+    },
+    {
+      "epoch": 1.250499400719137,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 3130
+    },
+    {
+      "epoch": 1.2508989212944468,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 3131
+    },
+    {
+      "epoch": 1.2512984418697564,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 3132
+    },
+    {
+      "epoch": 1.251697962445066,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 3133
+    },
+    {
+      "epoch": 1.2520974830203755,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 3134
+    },
+    {
+      "epoch": 1.252497003595685,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 3135
+    },
+    {
+      "epoch": 1.2528965241709948,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3136
+    },
+    {
+      "epoch": 1.2532960447463044,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 3137
+    },
+    {
+      "epoch": 1.253695565321614,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3138
+    },
+    {
+      "epoch": 1.2540950858969238,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 3139
+    },
+    {
+      "epoch": 1.2544946064722333,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3140
+    },
+    {
+      "epoch": 1.2548941270475429,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 3141
+    },
+    {
+      "epoch": 1.2552936476228527,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 3142
+    },
+    {
+      "epoch": 1.2556931681981622,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 3143
+    },
+    {
+      "epoch": 1.2560926887734718,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 3144
+    },
+    {
+      "epoch": 1.2564922093487816,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 3145
+    },
+    {
+      "epoch": 1.2568917299240912,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.085,
+      "step": 3146
+    },
+    {
+      "epoch": 1.2572912504994007,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 3147
+    },
+    {
+      "epoch": 1.2576907710747103,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 3148
+    },
+    {
+      "epoch": 1.2580902916500198,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 3149
+    },
+    {
+      "epoch": 1.2584898122253296,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 3150
+    },
+    {
+      "epoch": 1.2588893328006392,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 3151
+    },
+    {
+      "epoch": 1.2592888533759488,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 3152
+    },
+    {
+      "epoch": 1.2596883739512585,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3153
+    },
+    {
+      "epoch": 1.260087894526568,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 3154
+    },
+    {
+      "epoch": 1.2604874151018777,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 3155
+    },
+    {
+      "epoch": 1.2608869356771875,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 3156
+    },
+    {
+      "epoch": 1.261286456252497,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0819,
+      "step": 3157
+    },
+    {
+      "epoch": 1.2616859768278066,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3158
+    },
+    {
+      "epoch": 1.2620854974031164,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 3159
+    },
+    {
+      "epoch": 1.262485017978426,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 3160
+    },
+    {
+      "epoch": 1.2628845385537355,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 3161
+    },
+    {
+      "epoch": 1.263284059129045,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3162
+    },
+    {
+      "epoch": 1.2636835797043549,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 3163
+    },
+    {
+      "epoch": 1.2640831002796644,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 3164
+    },
+    {
+      "epoch": 1.264482620854974,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 3165
+    },
+    {
+      "epoch": 1.2648821414302835,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 3166
+    },
+    {
+      "epoch": 1.2652816620055933,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3167
+    },
+    {
+      "epoch": 1.265681182580903,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3168
+    },
+    {
+      "epoch": 1.2660807031562125,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 3169
+    },
+    {
+      "epoch": 1.2664802237315222,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3170
+    },
+    {
+      "epoch": 1.2668797443068318,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3171
+    },
+    {
+      "epoch": 1.2672792648821414,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 3172
+    },
+    {
+      "epoch": 1.2676787854574512,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0863,
+      "step": 3173
+    },
+    {
+      "epoch": 1.2680783060327607,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 3174
+    },
+    {
+      "epoch": 1.2684778266080703,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3175
+    },
+    {
+      "epoch": 1.26887734718338,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 3176
+    },
+    {
+      "epoch": 1.2692768677586896,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3177
+    },
+    {
+      "epoch": 1.2696763883339992,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 3178
+    },
+    {
+      "epoch": 1.2700759089093088,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 3179
+    },
+    {
+      "epoch": 1.2704754294846183,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 3180
+    },
+    {
+      "epoch": 1.2708749500599281,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 3181
+    },
+    {
+      "epoch": 1.2712744706352377,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3182
+    },
+    {
+      "epoch": 1.2716739912105472,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 3183
+    },
+    {
+      "epoch": 1.272073511785857,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 3184
+    },
+    {
+      "epoch": 1.2724730323611666,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0926,
+      "step": 3185
+    },
+    {
+      "epoch": 1.2728725529364762,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 3186
+    },
+    {
+      "epoch": 1.273272073511786,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 3187
+    },
+    {
+      "epoch": 1.2736715940870955,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 3188
+    },
+    {
+      "epoch": 1.274071114662405,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 3189
+    },
+    {
+      "epoch": 1.2744706352377149,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 3190
+    },
+    {
+      "epoch": 1.2748701558130244,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3191
+    },
+    {
+      "epoch": 1.275269676388334,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3192
+    },
+    {
+      "epoch": 1.2756691969636436,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 3193
+    },
+    {
+      "epoch": 1.2760687175389531,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3194
+    },
+    {
+      "epoch": 1.276468238114263,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 3195
+    },
+    {
+      "epoch": 1.2768677586895725,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 3196
+    },
+    {
+      "epoch": 1.277267279264882,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0885,
+      "step": 3197
+    },
+    {
+      "epoch": 1.2776667998401918,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3198
+    },
+    {
+      "epoch": 1.2780663204155014,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 3199
+    },
+    {
+      "epoch": 1.278465840990811,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2788653615661207,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3201
+    },
+    {
+      "epoch": 1.2792648821414303,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3202
+    },
+    {
+      "epoch": 1.2796644027167399,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 3203
+    },
+    {
+      "epoch": 1.2800639232920497,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 3204
+    },
+    {
+      "epoch": 1.2804634438673592,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 3205
+    },
+    {
+      "epoch": 1.2808629644426688,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0807,
+      "step": 3206
+    },
+    {
+      "epoch": 1.2812624850179783,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3207
+    },
+    {
+      "epoch": 1.2816620055932881,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 3208
+    },
+    {
+      "epoch": 1.2820615261685977,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 3209
+    },
+    {
+      "epoch": 1.2824610467439073,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2828605673192168,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 3211
+    },
+    {
+      "epoch": 1.2832600878945266,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 3212
+    },
+    {
+      "epoch": 1.2836596084698362,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 3213
+    },
+    {
+      "epoch": 1.2840591290451457,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 3214
+    },
+    {
+      "epoch": 1.2844586496204555,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3215
+    },
+    {
+      "epoch": 1.284858170195765,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3216
+    },
+    {
+      "epoch": 1.2852576907710747,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 3217
+    },
+    {
+      "epoch": 1.2856572113463844,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3218
+    },
+    {
+      "epoch": 1.286056731921694,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 3219
+    },
+    {
+      "epoch": 1.2864562524970036,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2868557730723134,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 3221
+    },
+    {
+      "epoch": 1.287255293647623,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 3222
+    },
+    {
+      "epoch": 1.2876548142229325,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3223
+    },
+    {
+      "epoch": 1.288054334798242,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 3224
+    },
+    {
+      "epoch": 1.2884538553735516,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 3225
+    },
+    {
+      "epoch": 1.2888533759488614,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0791,
+      "step": 3226
+    },
+    {
+      "epoch": 1.289252896524171,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3227
+    },
+    {
+      "epoch": 1.2896524170994805,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 3228
+    },
+    {
+      "epoch": 1.2900519376747903,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3229
+    },
+    {
+      "epoch": 1.2904514582500999,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3230
+    },
+    {
+      "epoch": 1.2908509788254094,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 3231
+    },
+    {
+      "epoch": 1.2912504994007192,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 3232
+    },
+    {
+      "epoch": 1.2916500199760288,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 3233
+    },
+    {
+      "epoch": 1.2920495405513384,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 3234
+    },
+    {
+      "epoch": 1.2924490611266481,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3235
+    },
+    {
+      "epoch": 1.2928485817019577,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3236
+    },
+    {
+      "epoch": 1.2932481022772673,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 3237
+    },
+    {
+      "epoch": 1.2936476228525768,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 3238
+    },
+    {
+      "epoch": 1.2940471434278864,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0867,
+      "step": 3239
+    },
+    {
+      "epoch": 1.2944466640031962,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 3240
+    },
+    {
+      "epoch": 1.2948461845785058,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 3241
+    },
+    {
+      "epoch": 1.2952457051538153,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0763,
+      "step": 3242
+    },
+    {
+      "epoch": 1.295645225729125,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 3243
+    },
+    {
+      "epoch": 1.2960447463044347,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.084,
+      "step": 3244
+    },
+    {
+      "epoch": 1.2964442668797442,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 3245
+    },
+    {
+      "epoch": 1.296843787455054,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 3246
+    },
+    {
+      "epoch": 1.2972433080303636,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 3247
+    },
+    {
+      "epoch": 1.2976428286056731,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 3248
+    },
+    {
+      "epoch": 1.298042349180983,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 3249
+    },
+    {
+      "epoch": 1.2984418697562925,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 3250
+    },
+    {
+      "epoch": 1.298841390331602,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 3251
+    },
+    {
+      "epoch": 1.2992409109069116,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3252
+    },
+    {
+      "epoch": 1.2996404314822214,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 3253
+    },
+    {
+      "epoch": 1.300039952057531,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 3254
+    },
+    {
+      "epoch": 1.3004394726328405,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3255
+    },
+    {
+      "epoch": 1.30083899320815,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 3256
+    },
+    {
+      "epoch": 1.30123851378346,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3257
+    },
+    {
+      "epoch": 1.3016380343587695,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 3258
+    },
+    {
+      "epoch": 1.302037554934079,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 3259
+    },
+    {
+      "epoch": 1.3024370755093888,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3260
+    },
+    {
+      "epoch": 1.3028365960846984,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 3261
+    },
+    {
+      "epoch": 1.303236116660008,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 3262
+    },
+    {
+      "epoch": 1.3036356372353177,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3263
+    },
+    {
+      "epoch": 1.3040351578106273,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 3264
+    },
+    {
+      "epoch": 1.3044346783859369,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0801,
+      "step": 3265
+    },
+    {
+      "epoch": 1.3048341989612466,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 3266
+    },
+    {
+      "epoch": 1.3052337195365562,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 3267
+    },
+    {
+      "epoch": 1.3056332401118658,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 3268
+    },
+    {
+      "epoch": 1.3060327606871753,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3269
+    },
+    {
+      "epoch": 1.306432281262485,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3270
+    },
+    {
+      "epoch": 1.3068318018377947,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 3271
+    },
+    {
+      "epoch": 1.3072313224131042,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 3272
+    },
+    {
+      "epoch": 1.3076308429884138,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 3273
+    },
+    {
+      "epoch": 1.3080303635637236,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 3274
+    },
+    {
+      "epoch": 1.3084298841390332,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3275
+    },
+    {
+      "epoch": 1.3088294047143427,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 3276
+    },
+    {
+      "epoch": 1.3092289252896525,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 3277
+    },
+    {
+      "epoch": 1.309628445864962,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 3278
+    },
+    {
+      "epoch": 1.3100279664402716,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3279
+    },
+    {
+      "epoch": 1.3104274870155814,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 3280
+    },
+    {
+      "epoch": 1.310827007590891,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 3281
+    },
+    {
+      "epoch": 1.3112265281662006,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 3282
+    },
+    {
+      "epoch": 1.3116260487415101,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 3283
+    },
+    {
+      "epoch": 1.3120255693168197,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 3284
+    },
+    {
+      "epoch": 1.3124250898921295,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 3285
+    },
+    {
+      "epoch": 1.312824610467439,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 3286
+    },
+    {
+      "epoch": 1.3132241310427486,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 3287
+    },
+    {
+      "epoch": 1.3136236516180584,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3288
+    },
+    {
+      "epoch": 1.314023172193368,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 3289
+    },
+    {
+      "epoch": 1.3144226927686775,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 3290
+    },
+    {
+      "epoch": 1.3148222133439873,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3291
+    },
+    {
+      "epoch": 1.3152217339192969,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 3292
+    },
+    {
+      "epoch": 1.3156212544946064,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 3293
+    },
+    {
+      "epoch": 1.3160207750699162,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0831,
+      "step": 3294
+    },
+    {
+      "epoch": 1.3164202956452258,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 3295
+    },
+    {
+      "epoch": 1.3168198162205353,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3296
+    },
+    {
+      "epoch": 1.317219336795845,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3297
+    },
+    {
+      "epoch": 1.3176188573711547,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3298
+    },
+    {
+      "epoch": 1.3180183779464643,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 3299
+    },
+    {
+      "epoch": 1.3184178985217738,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 3300
+    },
+    {
+      "epoch": 1.3188174190970834,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 3301
+    },
+    {
+      "epoch": 1.3192169396723932,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 3302
+    },
+    {
+      "epoch": 1.3196164602477027,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0816,
+      "step": 3303
+    },
+    {
+      "epoch": 1.3200159808230123,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 3304
+    },
+    {
+      "epoch": 1.320415501398322,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 3305
+    },
+    {
+      "epoch": 1.3208150219736317,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0765,
+      "step": 3306
+    },
+    {
+      "epoch": 1.3212145425489412,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 3307
+    },
+    {
+      "epoch": 1.321614063124251,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 3308
+    },
+    {
+      "epoch": 1.3220135836995606,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3309
+    },
+    {
+      "epoch": 1.3224131042748701,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 3310
+    },
+    {
+      "epoch": 1.32281262485018,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 3311
+    },
+    {
+      "epoch": 1.3232121454254895,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 3312
+    },
+    {
+      "epoch": 1.323611666000799,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 3313
+    },
+    {
+      "epoch": 1.3240111865761086,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3314
+    },
+    {
+      "epoch": 1.3244107071514182,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 3315
+    },
+    {
+      "epoch": 1.324810227726728,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3316
+    },
+    {
+      "epoch": 1.3252097483020375,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 3317
+    },
+    {
+      "epoch": 1.325609268877347,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3318
+    },
+    {
+      "epoch": 1.3260087894526569,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3319
+    },
+    {
+      "epoch": 1.3264083100279664,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 3320
+    },
+    {
+      "epoch": 1.326807830603276,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3321
+    },
+    {
+      "epoch": 1.3272073511785858,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 3322
+    },
+    {
+      "epoch": 1.3276068717538954,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 3323
+    },
+    {
+      "epoch": 1.328006392329205,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 3324
+    },
+    {
+      "epoch": 1.3284059129045147,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 3325
+    },
+    {
+      "epoch": 1.3288054334798243,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 3326
+    },
+    {
+      "epoch": 1.3292049540551338,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3327
+    },
+    {
+      "epoch": 1.3296044746304434,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 3328
+    },
+    {
+      "epoch": 1.330003995205753,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3329
+    },
+    {
+      "epoch": 1.3304035157810628,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3330
+    },
+    {
+      "epoch": 1.3308030363563723,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3331
+    },
+    {
+      "epoch": 1.3312025569316819,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3332
+    },
+    {
+      "epoch": 1.3316020775069917,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0878,
+      "step": 3333
+    },
+    {
+      "epoch": 1.3320015980823012,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 3334
+    },
+    {
+      "epoch": 1.3324011186576108,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0871,
+      "step": 3335
+    },
+    {
+      "epoch": 1.3328006392329206,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3336
+    },
+    {
+      "epoch": 1.3332001598082301,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 3337
+    },
+    {
+      "epoch": 1.3335996803835397,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 3338
+    },
+    {
+      "epoch": 1.3339992009588495,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0801,
+      "step": 3339
+    },
+    {
+      "epoch": 1.334398721534159,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 3340
+    },
+    {
+      "epoch": 1.3347982421094686,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 3341
+    },
+    {
+      "epoch": 1.3351977626847782,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0815,
+      "step": 3342
+    },
+    {
+      "epoch": 1.335597283260088,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 3343
+    },
+    {
+      "epoch": 1.3359968038353975,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 3344
+    },
+    {
+      "epoch": 1.336396324410707,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 3345
+    },
+    {
+      "epoch": 1.3367958449860167,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 3346
+    },
+    {
+      "epoch": 1.3371953655613265,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 3347
+    },
+    {
+      "epoch": 1.337594886136636,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3348
+    },
+    {
+      "epoch": 1.3379944067119456,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3349
+    },
+    {
+      "epoch": 1.3383939272872554,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3350
+    },
+    {
+      "epoch": 1.338793447862565,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 3351
+    },
+    {
+      "epoch": 1.3391929684378745,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 3352
+    },
+    {
+      "epoch": 1.3395924890131843,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 3353
+    },
+    {
+      "epoch": 1.3399920095884938,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3354
+    },
+    {
+      "epoch": 1.3403915301638034,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3355
+    },
+    {
+      "epoch": 1.3407910507391132,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3356
+    },
+    {
+      "epoch": 1.3411905713144228,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0762,
+      "step": 3357
+    },
+    {
+      "epoch": 1.3415900918897323,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3358
+    },
+    {
+      "epoch": 1.341989612465042,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3359
+    },
+    {
+      "epoch": 1.3423891330403515,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 3360
+    },
+    {
+      "epoch": 1.3427886536156612,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 3361
+    },
+    {
+      "epoch": 1.3431881741909708,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 3362
+    },
+    {
+      "epoch": 1.3435876947662804,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 3363
+    },
+    {
+      "epoch": 1.3439872153415902,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3364
+    },
+    {
+      "epoch": 1.3443867359168997,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3365
+    },
+    {
+      "epoch": 1.3447862564922093,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 3366
+    },
+    {
+      "epoch": 1.345185777067519,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3367
+    },
+    {
+      "epoch": 1.3455852976428286,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3368
+    },
+    {
+      "epoch": 1.3459848182181382,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0813,
+      "step": 3369
+    },
+    {
+      "epoch": 1.346384338793448,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 3370
+    },
+    {
+      "epoch": 1.3467838593687576,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 3371
+    },
+    {
+      "epoch": 1.3471833799440671,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 3372
+    },
+    {
+      "epoch": 1.3475829005193767,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 3373
+    },
+    {
+      "epoch": 1.3479824210946862,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 3374
+    },
+    {
+      "epoch": 1.348381941669996,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0828,
+      "step": 3375
+    },
+    {
+      "epoch": 1.3487814622453056,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 3376
+    },
+    {
+      "epoch": 1.3491809828206152,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 3377
+    },
+    {
+      "epoch": 1.349580503395925,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 3378
+    },
+    {
+      "epoch": 1.3499800239712345,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3379
+    },
+    {
+      "epoch": 1.350379544546544,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3380
+    },
+    {
+      "epoch": 1.3507790651218539,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 3381
+    },
+    {
+      "epoch": 1.3511785856971634,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 3382
+    },
+    {
+      "epoch": 1.351578106272473,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 3383
+    },
+    {
+      "epoch": 1.3519776268477828,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 3384
+    },
+    {
+      "epoch": 1.3523771474230923,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 3385
+    },
+    {
+      "epoch": 1.352776667998402,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 3386
+    },
+    {
+      "epoch": 1.3531761885737115,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3387
+    },
+    {
+      "epoch": 1.3535757091490213,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 3388
+    },
+    {
+      "epoch": 1.3539752297243308,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3389
+    },
+    {
+      "epoch": 1.3543747502996404,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3390
+    },
+    {
+      "epoch": 1.35477427087495,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 3391
+    },
+    {
+      "epoch": 1.3551737914502597,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0866,
+      "step": 3392
+    },
+    {
+      "epoch": 1.3555733120255693,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3393
+    },
+    {
+      "epoch": 1.3559728326008789,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 3394
+    },
+    {
+      "epoch": 1.3563723531761886,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 3395
+    },
+    {
+      "epoch": 1.3567718737514982,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 3396
+    },
+    {
+      "epoch": 1.3571713943268078,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 3397
+    },
+    {
+      "epoch": 1.3575709149021176,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 3398
+    },
+    {
+      "epoch": 1.3579704354774271,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 3399
+    },
+    {
+      "epoch": 1.3583699560527367,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 3400
+    },
+    {
+      "epoch": 1.3587694766280465,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3401
+    },
+    {
+      "epoch": 1.359168997203356,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 3402
+    },
+    {
+      "epoch": 1.3595685177786656,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 3403
+    },
+    {
+      "epoch": 1.3599680383539752,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 3404
+    },
+    {
+      "epoch": 1.3603675589292847,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 3405
+    },
+    {
+      "epoch": 1.3607670795045945,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0832,
+      "step": 3406
+    },
+    {
+      "epoch": 1.361166600079904,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 3407
+    },
+    {
+      "epoch": 1.3615661206552137,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 3408
+    },
+    {
+      "epoch": 1.3619656412305234,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 3409
+    },
+    {
+      "epoch": 1.362365161805833,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 3410
+    },
+    {
+      "epoch": 1.3627646823811426,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 3411
+    },
+    {
+      "epoch": 1.3631642029564524,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 3412
+    },
+    {
+      "epoch": 1.363563723531762,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 3413
+    },
+    {
+      "epoch": 1.3639632441070715,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3414
+    },
+    {
+      "epoch": 1.3643627646823813,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 3415
+    },
+    {
+      "epoch": 1.3647622852576908,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 3416
+    },
+    {
+      "epoch": 1.3651618058330004,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3417
+    },
+    {
+      "epoch": 1.36556132640831,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3418
+    },
+    {
+      "epoch": 1.3659608469836195,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3419
+    },
+    {
+      "epoch": 1.3663603675589293,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 3420
+    },
+    {
+      "epoch": 1.3667598881342389,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0898,
+      "step": 3421
+    },
+    {
+      "epoch": 1.3671594087095484,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 3422
+    },
+    {
+      "epoch": 1.3675589292848582,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3423
+    },
+    {
+      "epoch": 1.3679584498601678,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3424
+    },
+    {
+      "epoch": 1.3683579704354774,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 3425
+    },
+    {
+      "epoch": 1.3687574910107871,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 3426
+    },
+    {
+      "epoch": 1.3691570115860967,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3427
+    },
+    {
+      "epoch": 1.3695565321614063,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 3428
+    },
+    {
+      "epoch": 1.369956052736716,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 3429
+    },
+    {
+      "epoch": 1.3703555733120256,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 3430
+    },
+    {
+      "epoch": 1.3707550938873352,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3431
+    },
+    {
+      "epoch": 1.3711546144626447,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 3432
+    },
+    {
+      "epoch": 1.3715541350379545,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3433
+    },
+    {
+      "epoch": 1.371953655613264,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0904,
+      "step": 3434
+    },
+    {
+      "epoch": 1.3723531761885737,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 3435
+    },
+    {
+      "epoch": 1.3727526967638832,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 3436
+    },
+    {
+      "epoch": 1.373152217339193,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 3437
+    },
+    {
+      "epoch": 1.3735517379145026,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3438
+    },
+    {
+      "epoch": 1.3739512584898121,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0839,
+      "step": 3439
+    },
+    {
+      "epoch": 1.374350779065122,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 3440
+    },
+    {
+      "epoch": 1.3747502996404315,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 3441
+    },
+    {
+      "epoch": 1.375149820215741,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 3442
+    },
+    {
+      "epoch": 1.3755493407910508,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 3443
+    },
+    {
+      "epoch": 1.3759488613663604,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3444
+    },
+    {
+      "epoch": 1.37634838194167,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 3445
+    },
+    {
+      "epoch": 1.3767479025169798,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 3446
+    },
+    {
+      "epoch": 1.3771474230922893,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 3447
+    },
+    {
+      "epoch": 1.3775469436675989,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 3448
+    },
+    {
+      "epoch": 1.3779464642429085,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3449
+    },
+    {
+      "epoch": 1.378345984818218,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 3450
+    },
+    {
+      "epoch": 1.3787455053935278,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3451
+    },
+    {
+      "epoch": 1.3791450259688374,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 3452
+    },
+    {
+      "epoch": 1.379544546544147,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 3453
+    },
+    {
+      "epoch": 1.3799440671194567,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 3454
+    },
+    {
+      "epoch": 1.3803435876947663,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 3455
+    },
+    {
+      "epoch": 1.3807431082700758,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 3456
+    },
+    {
+      "epoch": 1.3811426288453856,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 3457
+    },
+    {
+      "epoch": 1.3815421494206952,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 3458
+    },
+    {
+      "epoch": 1.3819416699960048,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 3459
+    },
+    {
+      "epoch": 1.3823411905713145,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3460
+    },
+    {
+      "epoch": 1.3827407111466241,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 3461
+    },
+    {
+      "epoch": 1.3831402317219337,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3462
+    },
+    {
+      "epoch": 1.3835397522972432,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 3463
+    },
+    {
+      "epoch": 1.3839392728725528,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3464
+    },
+    {
+      "epoch": 1.3843387934478626,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3465
+    },
+    {
+      "epoch": 1.3847383140231722,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 3466
+    },
+    {
+      "epoch": 1.3851378345984817,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3467
+    },
+    {
+      "epoch": 1.3855373551737915,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 3468
+    },
+    {
+      "epoch": 1.385936875749101,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 3469
+    },
+    {
+      "epoch": 1.3863363963244106,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3470
+    },
+    {
+      "epoch": 1.3867359168997204,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 3471
+    },
+    {
+      "epoch": 1.38713543747503,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 3472
+    },
+    {
+      "epoch": 1.3875349580503396,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3473
+    },
+    {
+      "epoch": 1.3879344786256493,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 3474
+    },
+    {
+      "epoch": 1.388333999200959,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3475
+    },
+    {
+      "epoch": 1.3887335197762685,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3476
+    },
+    {
+      "epoch": 1.389133040351578,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 3477
+    },
+    {
+      "epoch": 1.3895325609268878,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0908,
+      "step": 3478
+    },
+    {
+      "epoch": 1.3899320815021974,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3479
+    },
+    {
+      "epoch": 1.390331602077507,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3480
+    },
+    {
+      "epoch": 1.3907311226528165,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 3481
+    },
+    {
+      "epoch": 1.3911306432281263,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 3482
+    },
+    {
+      "epoch": 1.3915301638034359,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 3483
+    },
+    {
+      "epoch": 1.3919296843787454,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3484
+    },
+    {
+      "epoch": 1.3923292049540552,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0796,
+      "step": 3485
+    },
+    {
+      "epoch": 1.3927287255293648,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 3486
+    },
+    {
+      "epoch": 1.3931282461046743,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 3487
+    },
+    {
+      "epoch": 1.3935277666799841,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 3488
+    },
+    {
+      "epoch": 1.3939272872552937,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 3489
+    },
+    {
+      "epoch": 1.3943268078306033,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3490
+    },
+    {
+      "epoch": 1.394726328405913,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3491
+    },
+    {
+      "epoch": 1.3951258489812226,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 3492
+    },
+    {
+      "epoch": 1.3955253695565322,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 3493
+    },
+    {
+      "epoch": 1.3959248901318417,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 3494
+    },
+    {
+      "epoch": 1.3963244107071513,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0795,
+      "step": 3495
+    },
+    {
+      "epoch": 1.396723931282461,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3496
+    },
+    {
+      "epoch": 1.3971234518577706,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 3497
+    },
+    {
+      "epoch": 1.3975229724330802,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 3498
+    },
+    {
+      "epoch": 1.39792249300839,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 3499
+    },
+    {
+      "epoch": 1.3983220135836996,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3987215341590091,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.083,
+      "step": 3501
+    },
+    {
+      "epoch": 1.399121054734319,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3502
+    },
+    {
+      "epoch": 1.3995205753096285,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3503
+    },
+    {
+      "epoch": 1.399920095884938,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 3504
+    },
+    {
+      "epoch": 1.4003196164602478,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0801,
+      "step": 3505
+    },
+    {
+      "epoch": 1.4007191370355574,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3506
+    },
+    {
+      "epoch": 1.401118657610867,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 3507
+    },
+    {
+      "epoch": 1.4015181781861765,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 3508
+    },
+    {
+      "epoch": 1.401917698761486,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3509
+    },
+    {
+      "epoch": 1.4023172193367959,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3510
+    },
+    {
+      "epoch": 1.4027167399121054,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 3511
+    },
+    {
+      "epoch": 1.403116260487415,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 3512
+    },
+    {
+      "epoch": 1.4035157810627248,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3513
+    },
+    {
+      "epoch": 1.4039153016380344,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 3514
+    },
+    {
+      "epoch": 1.404314822213344,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3515
+    },
+    {
+      "epoch": 1.4047143427886537,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3516
+    },
+    {
+      "epoch": 1.4051138633639633,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 3517
+    },
+    {
+      "epoch": 1.4055133839392728,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3518
+    },
+    {
+      "epoch": 1.4059129045145826,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 3519
+    },
+    {
+      "epoch": 1.4063124250898922,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 3520
+    },
+    {
+      "epoch": 1.4067119456652017,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3521
+    },
+    {
+      "epoch": 1.4071114662405113,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 3522
+    },
+    {
+      "epoch": 1.407510986815821,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3523
+    },
+    {
+      "epoch": 1.4079105073911307,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 3524
+    },
+    {
+      "epoch": 1.4083100279664402,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 3525
+    },
+    {
+      "epoch": 1.4087095485417498,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 3526
+    },
+    {
+      "epoch": 1.4091090691170596,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3527
+    },
+    {
+      "epoch": 1.4095085896923691,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 3528
+    },
+    {
+      "epoch": 1.4099081102676787,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0854,
+      "step": 3529
+    },
+    {
+      "epoch": 1.4103076308429885,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 3530
+    },
+    {
+      "epoch": 1.410707151418298,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3531
+    },
+    {
+      "epoch": 1.4111066719936076,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 3532
+    },
+    {
+      "epoch": 1.4115061925689174,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3533
+    },
+    {
+      "epoch": 1.411905713144227,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 3534
+    },
+    {
+      "epoch": 1.4123052337195365,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 3535
+    },
+    {
+      "epoch": 1.4127047542948463,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3536
+    },
+    {
+      "epoch": 1.4131042748701559,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 3537
+    },
+    {
+      "epoch": 1.4135037954454654,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 3538
+    },
+    {
+      "epoch": 1.413903316020775,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3539
+    },
+    {
+      "epoch": 1.4143028365960846,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3540
+    },
+    {
+      "epoch": 1.4147023571713944,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 3541
+    },
+    {
+      "epoch": 1.415101877746704,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3542
+    },
+    {
+      "epoch": 1.4155013983220135,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 3543
+    },
+    {
+      "epoch": 1.4159009188973233,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 3544
+    },
+    {
+      "epoch": 1.4163004394726328,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3545
+    },
+    {
+      "epoch": 1.4166999600479424,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 3546
+    },
+    {
+      "epoch": 1.4170994806232522,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3547
+    },
+    {
+      "epoch": 1.4174990011985618,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3548
+    },
+    {
+      "epoch": 1.4178985217738713,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 3549
+    },
+    {
+      "epoch": 1.418298042349181,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 3550
+    },
+    {
+      "epoch": 1.4186975629244907,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3551
+    },
+    {
+      "epoch": 1.4190970834998002,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 3552
+    },
+    {
+      "epoch": 1.4194966040751098,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3553
+    },
+    {
+      "epoch": 1.4198961246504194,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3554
+    },
+    {
+      "epoch": 1.4202956452257292,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3555
+    },
+    {
+      "epoch": 1.4206951658010387,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 3556
+    },
+    {
+      "epoch": 1.4210946863763483,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 3557
+    },
+    {
+      "epoch": 1.421494206951658,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 3558
+    },
+    {
+      "epoch": 1.4218937275269676,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3559
+    },
+    {
+      "epoch": 1.4222932481022772,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 3560
+    },
+    {
+      "epoch": 1.422692768677587,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3561
+    },
+    {
+      "epoch": 1.4230922892528965,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 3562
+    },
+    {
+      "epoch": 1.4234918098282061,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3563
+    },
+    {
+      "epoch": 1.423891330403516,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3564
+    },
+    {
+      "epoch": 1.4242908509788255,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3565
+    },
+    {
+      "epoch": 1.424690371554135,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 3566
+    },
+    {
+      "epoch": 1.4250898921294446,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 3567
+    },
+    {
+      "epoch": 1.4254894127047544,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 3568
+    },
+    {
+      "epoch": 1.425888933280064,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 3569
+    },
+    {
+      "epoch": 1.4262884538553735,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 3570
+    },
+    {
+      "epoch": 1.426687974430683,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 3571
+    },
+    {
+      "epoch": 1.4270874950059929,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 3572
+    },
+    {
+      "epoch": 1.4274870155813024,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3573
+    },
+    {
+      "epoch": 1.427886536156612,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 3574
+    },
+    {
+      "epoch": 1.4282860567319218,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 3575
+    },
+    {
+      "epoch": 1.4286855773072313,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3576
+    },
+    {
+      "epoch": 1.429085097882541,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 3577
+    },
+    {
+      "epoch": 1.4294846184578507,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3578
+    },
+    {
+      "epoch": 1.4298841390331603,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 3579
+    },
+    {
+      "epoch": 1.4302836596084698,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 3580
+    },
+    {
+      "epoch": 1.4306831801837796,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 3581
+    },
+    {
+      "epoch": 1.4310827007590892,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3582
+    },
+    {
+      "epoch": 1.4314822213343987,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 3583
+    },
+    {
+      "epoch": 1.4318817419097083,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 3584
+    },
+    {
+      "epoch": 1.4322812624850179,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0822,
+      "step": 3585
+    },
+    {
+      "epoch": 1.4326807830603276,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 3586
+    },
+    {
+      "epoch": 1.4330803036356372,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0843,
+      "step": 3587
+    },
+    {
+      "epoch": 1.4334798242109468,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 3588
+    },
+    {
+      "epoch": 1.4338793447862566,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 3589
+    },
+    {
+      "epoch": 1.4342788653615661,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 3590
+    },
+    {
+      "epoch": 1.4346783859368757,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 3591
+    },
+    {
+      "epoch": 1.4350779065121855,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0789,
+      "step": 3592
+    },
+    {
+      "epoch": 1.435477427087495,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3593
+    },
+    {
+      "epoch": 1.4358769476628046,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 3594
+    },
+    {
+      "epoch": 1.4362764682381144,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3595
+    },
+    {
+      "epoch": 1.436675988813424,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3596
+    },
+    {
+      "epoch": 1.4370755093887335,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3597
+    },
+    {
+      "epoch": 1.437475029964043,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 3598
+    },
+    {
+      "epoch": 1.4378745505393526,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 3599
+    },
+    {
+      "epoch": 1.4382740711146624,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 3600
+    },
+    {
+      "epoch": 1.438673591689972,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 3601
+    },
+    {
+      "epoch": 1.4390731122652816,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3602
+    },
+    {
+      "epoch": 1.4394726328405913,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 3603
+    },
+    {
+      "epoch": 1.439872153415901,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3604
+    },
+    {
+      "epoch": 1.4402716739912105,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 3605
+    },
+    {
+      "epoch": 1.4406711945665203,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 3606
+    },
+    {
+      "epoch": 1.4410707151418298,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3607
+    },
+    {
+      "epoch": 1.4414702357171394,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0763,
+      "step": 3608
+    },
+    {
+      "epoch": 1.4418697562924492,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 3609
+    },
+    {
+      "epoch": 1.4422692768677587,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3610
+    },
+    {
+      "epoch": 1.4426687974430683,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3611
+    },
+    {
+      "epoch": 1.4430683180183779,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3612
+    },
+    {
+      "epoch": 1.4434678385936877,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 3613
+    },
+    {
+      "epoch": 1.4438673591689972,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3614
+    },
+    {
+      "epoch": 1.4442668797443068,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 3615
+    },
+    {
+      "epoch": 1.4446664003196164,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3616
+    },
+    {
+      "epoch": 1.4450659208949261,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 3617
+    },
+    {
+      "epoch": 1.4454654414702357,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 3618
+    },
+    {
+      "epoch": 1.4458649620455453,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 3619
+    },
+    {
+      "epoch": 1.446264482620855,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3620
+    },
+    {
+      "epoch": 1.4466640031961646,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 3621
+    },
+    {
+      "epoch": 1.4470635237714742,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 3622
+    },
+    {
+      "epoch": 1.447463044346784,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 3623
+    },
+    {
+      "epoch": 1.4478625649220935,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 3624
+    },
+    {
+      "epoch": 1.448262085497403,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 3625
+    },
+    {
+      "epoch": 1.4486616060727129,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 3626
+    },
+    {
+      "epoch": 1.4490611266480224,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 3627
+    },
+    {
+      "epoch": 1.449460647223332,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0797,
+      "step": 3628
+    },
+    {
+      "epoch": 1.4498601677986416,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3629
+    },
+    {
+      "epoch": 1.4502596883739511,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 3630
+    },
+    {
+      "epoch": 1.450659208949261,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3631
+    },
+    {
+      "epoch": 1.4510587295245705,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3632
+    },
+    {
+      "epoch": 1.45145825009988,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 3633
+    },
+    {
+      "epoch": 1.4518577706751898,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3634
+    },
+    {
+      "epoch": 1.4522572912504994,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 3635
+    },
+    {
+      "epoch": 1.452656811825809,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 3636
+    },
+    {
+      "epoch": 1.4530563324011188,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 3637
+    },
+    {
+      "epoch": 1.4534558529764283,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3638
+    },
+    {
+      "epoch": 1.4538553735517379,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 3639
+    },
+    {
+      "epoch": 1.4542548941270477,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 3640
+    },
+    {
+      "epoch": 1.4546544147023572,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3641
+    },
+    {
+      "epoch": 1.4550539352776668,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 3642
+    },
+    {
+      "epoch": 1.4554534558529764,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3643
+    },
+    {
+      "epoch": 1.455852976428286,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3644
+    },
+    {
+      "epoch": 1.4562524970035957,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 3645
+    },
+    {
+      "epoch": 1.4566520175789053,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 3646
+    },
+    {
+      "epoch": 1.4570515381542148,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 3647
+    },
+    {
+      "epoch": 1.4574510587295246,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 3648
+    },
+    {
+      "epoch": 1.4578505793048342,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 3649
+    },
+    {
+      "epoch": 1.4582500998801438,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 3650
+    },
+    {
+      "epoch": 1.4586496204554535,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3651
+    },
+    {
+      "epoch": 1.459049141030763,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 3652
+    },
+    {
+      "epoch": 1.4594486616060727,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3653
+    },
+    {
+      "epoch": 1.4598481821813825,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3654
+    },
+    {
+      "epoch": 1.460247702756692,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 3655
+    },
+    {
+      "epoch": 1.4606472233320016,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3656
+    },
+    {
+      "epoch": 1.4610467439073112,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 3657
+    },
+    {
+      "epoch": 1.461446264482621,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0799,
+      "step": 3658
+    },
+    {
+      "epoch": 1.4618457850579305,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3659
+    },
+    {
+      "epoch": 1.46224530563324,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3660
+    },
+    {
+      "epoch": 1.4626448262085496,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 3661
+    },
+    {
+      "epoch": 1.4630443467838594,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 3662
+    },
+    {
+      "epoch": 1.463443867359169,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 3663
+    },
+    {
+      "epoch": 1.4638433879344785,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3664
+    },
+    {
+      "epoch": 1.4642429085097883,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3665
+    },
+    {
+      "epoch": 1.464642429085098,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 3666
+    },
+    {
+      "epoch": 1.4650419496604075,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 3667
+    },
+    {
+      "epoch": 1.4654414702357172,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 3668
+    },
+    {
+      "epoch": 1.4658409908110268,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3669
+    },
+    {
+      "epoch": 1.4662405113863364,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 3670
+    },
+    {
+      "epoch": 1.4666400319616462,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 3671
+    },
+    {
+      "epoch": 1.4670395525369557,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 3672
+    },
+    {
+      "epoch": 1.4674390731122653,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 3673
+    },
+    {
+      "epoch": 1.4678385936875749,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 3674
+    },
+    {
+      "epoch": 1.4682381142628844,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 3675
+    },
+    {
+      "epoch": 1.4686376348381942,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.0837,
+      "step": 3676
+    },
+    {
+      "epoch": 1.4690371554135038,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 3677
+    },
+    {
+      "epoch": 1.4694366759888133,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0761,
+      "step": 3678
+    },
+    {
+      "epoch": 1.4698361965641231,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 3679
+    },
+    {
+      "epoch": 1.4702357171394327,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 3680
+    },
+    {
+      "epoch": 1.4706352377147422,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 3681
+    },
+    {
+      "epoch": 1.471034758290052,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 3682
+    },
+    {
+      "epoch": 1.4714342788653616,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 3683
+    },
+    {
+      "epoch": 1.4718337994406712,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3684
+    },
+    {
+      "epoch": 1.472233320015981,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3685
+    },
+    {
+      "epoch": 1.4726328405912905,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 3686
+    },
+    {
+      "epoch": 1.4730323611666,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3687
+    },
+    {
+      "epoch": 1.4734318817419096,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 3688
+    },
+    {
+      "epoch": 1.4738314023172192,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 3689
+    },
+    {
+      "epoch": 1.474230922892529,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 3690
+    },
+    {
+      "epoch": 1.4746304434678386,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 3691
+    },
+    {
+      "epoch": 1.4750299640431481,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3692
+    },
+    {
+      "epoch": 1.475429484618458,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3693
+    },
+    {
+      "epoch": 1.4758290051937675,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3694
+    },
+    {
+      "epoch": 1.476228525769077,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3695
+    },
+    {
+      "epoch": 1.4766280463443868,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 3696
+    },
+    {
+      "epoch": 1.4770275669196964,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0835,
+      "step": 3697
+    },
+    {
+      "epoch": 1.477427087495006,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3698
+    },
+    {
+      "epoch": 1.4778266080703157,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 3699
+    },
+    {
+      "epoch": 1.4782261286456253,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3700
+    },
+    {
+      "epoch": 1.4786256492209349,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 3701
+    },
+    {
+      "epoch": 1.4790251697962444,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3702
+    },
+    {
+      "epoch": 1.479424690371554,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3703
+    },
+    {
+      "epoch": 1.4798242109468638,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3704
+    },
+    {
+      "epoch": 1.4802237315221733,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 3705
+    },
+    {
+      "epoch": 1.480623252097483,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 3706
+    },
+    {
+      "epoch": 1.4810227726727927,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 3707
+    },
+    {
+      "epoch": 1.4814222932481023,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0793,
+      "step": 3708
+    },
+    {
+      "epoch": 1.4818218138234118,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 3709
+    },
+    {
+      "epoch": 1.4822213343987216,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 3710
+    },
+    {
+      "epoch": 1.4826208549740312,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3711
+    },
+    {
+      "epoch": 1.4830203755493407,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3712
+    },
+    {
+      "epoch": 1.4834198961246505,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 3713
+    },
+    {
+      "epoch": 1.48381941669996,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3714
+    },
+    {
+      "epoch": 1.4842189372752697,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 3715
+    },
+    {
+      "epoch": 1.4846184578505794,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 3716
+    },
+    {
+      "epoch": 1.485017978425889,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 3717
+    },
+    {
+      "epoch": 1.4854174990011986,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 3718
+    },
+    {
+      "epoch": 1.4858170195765081,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3719
+    },
+    {
+      "epoch": 1.4862165401518177,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 3720
+    },
+    {
+      "epoch": 1.4866160607271275,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3721
+    },
+    {
+      "epoch": 1.487015581302437,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 3722
+    },
+    {
+      "epoch": 1.4874151018777466,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 3723
+    },
+    {
+      "epoch": 1.4878146224530564,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 3724
+    },
+    {
+      "epoch": 1.488214143028366,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 3725
+    },
+    {
+      "epoch": 1.4886136636036755,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3726
+    },
+    {
+      "epoch": 1.4890131841789853,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3727
+    },
+    {
+      "epoch": 1.4894127047542949,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 3728
+    },
+    {
+      "epoch": 1.4898122253296044,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 3729
+    },
+    {
+      "epoch": 1.4902117459049142,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 3730
+    },
+    {
+      "epoch": 1.4906112664802238,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 3731
+    },
+    {
+      "epoch": 1.4910107870555334,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 3732
+    },
+    {
+      "epoch": 1.491410307630843,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 3733
+    },
+    {
+      "epoch": 1.4918098282061525,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 3734
+    },
+    {
+      "epoch": 1.4922093487814623,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 3735
+    },
+    {
+      "epoch": 1.4926088693567718,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 3736
+    },
+    {
+      "epoch": 1.4930083899320814,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3737
+    },
+    {
+      "epoch": 1.4934079105073912,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 3738
+    },
+    {
+      "epoch": 1.4938074310827008,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 3739
+    },
+    {
+      "epoch": 1.4942069516580103,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 3740
+    },
+    {
+      "epoch": 1.49460647223332,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 3741
+    },
+    {
+      "epoch": 1.4950059928086297,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3742
+    },
+    {
+      "epoch": 1.4954055133839392,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 3743
+    },
+    {
+      "epoch": 1.495805033959249,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3744
+    },
+    {
+      "epoch": 1.4962045545345586,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3745
+    },
+    {
+      "epoch": 1.4966040751098681,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 3746
+    },
+    {
+      "epoch": 1.4970035956851777,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 3747
+    },
+    {
+      "epoch": 1.4974031162604873,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 3748
+    },
+    {
+      "epoch": 1.497802636835797,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 3749
+    },
+    {
+      "epoch": 1.4982021574111066,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 3750
+    },
+    {
+      "epoch": 1.4986016779864162,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 3751
+    },
+    {
+      "epoch": 1.499001198561726,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3752
+    },
+    {
+      "epoch": 1.4994007191370355,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 3753
+    },
+    {
+      "epoch": 1.499800239712345,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0794,
+      "step": 3754
+    },
+    {
+      "epoch": 1.500199760287655,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 3755
+    },
+    {
+      "epoch": 1.5005992808629645,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 3756
+    },
+    {
+      "epoch": 1.500998801438274,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 3757
+    },
+    {
+      "epoch": 1.5013983220135838,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.086,
+      "step": 3758
+    },
+    {
+      "epoch": 1.5017978425888934,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 3759
+    },
+    {
+      "epoch": 1.502197363164203,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0776,
+      "step": 3760
+    },
+    {
+      "epoch": 1.5025968837395127,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 3761
+    },
+    {
+      "epoch": 1.502996404314822,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3762
+    },
+    {
+      "epoch": 1.5033959248901319,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 3763
+    },
+    {
+      "epoch": 1.5037954454654414,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3764
+    },
+    {
+      "epoch": 1.504194966040751,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3765
+    },
+    {
+      "epoch": 1.5045944866160608,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 3766
+    },
+    {
+      "epoch": 1.5049940071913703,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 3767
+    },
+    {
+      "epoch": 1.50539352776668,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 3768
+    },
+    {
+      "epoch": 1.5057930483419897,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 3769
+    },
+    {
+      "epoch": 1.5061925689172992,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0788,
+      "step": 3770
+    },
+    {
+      "epoch": 1.5065920894926088,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 3771
+    },
+    {
+      "epoch": 1.5069916100679186,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 3772
+    },
+    {
+      "epoch": 1.5073911306432282,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 3773
+    },
+    {
+      "epoch": 1.5077906512185377,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 3774
+    },
+    {
+      "epoch": 1.5081901717938475,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 3775
+    },
+    {
+      "epoch": 1.5085896923691569,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 3776
+    },
+    {
+      "epoch": 1.5089892129444666,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 3777
+    },
+    {
+      "epoch": 1.5093887335197764,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 3778
+    },
+    {
+      "epoch": 1.5097882540950858,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 3779
+    },
+    {
+      "epoch": 1.5101877746703956,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 3780
+    },
+    {
+      "epoch": 1.5105872952457051,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0759,
+      "step": 3781
+    },
+    {
+      "epoch": 1.5109868158210147,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 3782
+    },
+    {
+      "epoch": 1.5113863363963245,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3783
+    },
+    {
+      "epoch": 1.511785856971634,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 3784
+    },
+    {
+      "epoch": 1.5121853775469436,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 3785
+    },
+    {
+      "epoch": 1.5125848981222534,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 3786
+    },
+    {
+      "epoch": 1.512984418697563,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 3787
+    },
+    {
+      "epoch": 1.5133839392728725,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0827,
+      "step": 3788
+    },
+    {
+      "epoch": 1.5137834598481823,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 3789
+    },
+    {
+      "epoch": 1.5141829804234919,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3790
+    },
+    {
+      "epoch": 1.5145825009988014,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3791
+    },
+    {
+      "epoch": 1.5149820215741112,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 3792
+    },
+    {
+      "epoch": 1.5153815421494206,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 3793
+    },
+    {
+      "epoch": 1.5157810627247303,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 3794
+    },
+    {
+      "epoch": 1.51618058330004,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 3795
+    },
+    {
+      "epoch": 1.5165801038753495,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 3796
+    },
+    {
+      "epoch": 1.5169796244506593,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 3797
+    },
+    {
+      "epoch": 1.5173791450259688,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 3798
+    },
+    {
+      "epoch": 1.5177786656012784,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 3799
+    },
+    {
+      "epoch": 1.5181781861765882,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 3800
+    },
+    {
+      "epoch": 1.5185777067518977,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3801
+    },
+    {
+      "epoch": 1.5189772273272073,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3802
+    },
+    {
+      "epoch": 1.519376747902517,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 3803
+    },
+    {
+      "epoch": 1.5197762684778267,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 3804
+    },
+    {
+      "epoch": 1.5201757890531362,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3805
+    },
+    {
+      "epoch": 1.520575309628446,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3806
+    },
+    {
+      "epoch": 1.5209748302037553,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 3807
+    },
+    {
+      "epoch": 1.5213743507790651,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 3808
+    },
+    {
+      "epoch": 1.5217738713543747,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 3809
+    },
+    {
+      "epoch": 1.5221733919296843,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 3810
+    },
+    {
+      "epoch": 1.522572912504994,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 3811
+    },
+    {
+      "epoch": 1.5229724330803036,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 3812
+    },
+    {
+      "epoch": 1.5233719536556132,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3813
+    },
+    {
+      "epoch": 1.523771474230923,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 3814
+    },
+    {
+      "epoch": 1.5241709948062325,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 3815
+    },
+    {
+      "epoch": 1.524570515381542,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 3816
+    },
+    {
+      "epoch": 1.5249700359568519,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 3817
+    },
+    {
+      "epoch": 1.5253695565321614,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 3818
+    },
+    {
+      "epoch": 1.525769077107471,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 3819
+    },
+    {
+      "epoch": 1.5261685976827808,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 3820
+    },
+    {
+      "epoch": 1.5265681182580901,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 3821
+    },
+    {
+      "epoch": 1.5269676388334,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0821,
+      "step": 3822
+    },
+    {
+      "epoch": 1.5273671594087097,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 3823
+    },
+    {
+      "epoch": 1.527766679984019,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 3824
+    },
+    {
+      "epoch": 1.5281662005593288,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 3825
+    },
+    {
+      "epoch": 1.5285657211346384,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 3826
+    },
+    {
+      "epoch": 1.528965241709948,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 3827
+    },
+    {
+      "epoch": 1.5293647622852578,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 3828
+    },
+    {
+      "epoch": 1.5297642828605673,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 3829
+    },
+    {
+      "epoch": 1.5301638034358769,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 3830
+    },
+    {
+      "epoch": 1.5305633240111867,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3831
+    },
+    {
+      "epoch": 1.5309628445864962,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 3832
+    },
+    {
+      "epoch": 1.5313623651618058,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 3833
+    },
+    {
+      "epoch": 1.5317618857371156,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 3834
+    },
+    {
+      "epoch": 1.5321614063124251,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 3835
+    },
+    {
+      "epoch": 1.5325609268877347,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 3836
+    },
+    {
+      "epoch": 1.5329604474630445,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 3837
+    },
+    {
+      "epoch": 1.5333599680383538,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 3838
+    },
+    {
+      "epoch": 1.5337594886136636,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 3839
+    },
+    {
+      "epoch": 1.5341590091889732,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 3840
+    },
+    {
+      "epoch": 1.5345585297642828,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 3841
+    },
+    {
+      "epoch": 1.5349580503395925,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 3842
+    },
+    {
+      "epoch": 1.535357570914902,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 3843
+    },
+    {
+      "epoch": 1.5357570914902117,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3844
+    },
+    {
+      "epoch": 1.5361566120655215,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 3845
+    },
+    {
+      "epoch": 1.536556132640831,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3846
+    },
+    {
+      "epoch": 1.5369556532161406,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3847
+    },
+    {
+      "epoch": 1.5373551737914504,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3848
+    },
+    {
+      "epoch": 1.53775469436676,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3849
+    },
+    {
+      "epoch": 1.5381542149420695,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3850
+    },
+    {
+      "epoch": 1.5385537355173793,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 3851
+    },
+    {
+      "epoch": 1.5389532560926886,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 3852
+    },
+    {
+      "epoch": 1.5393527766679984,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 3853
+    },
+    {
+      "epoch": 1.539752297243308,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 3854
+    },
+    {
+      "epoch": 1.5401518178186175,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 3855
+    },
+    {
+      "epoch": 1.5405513383939273,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 3856
+    },
+    {
+      "epoch": 1.540950858969237,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 3857
+    },
+    {
+      "epoch": 1.5413503795445465,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 3858
+    },
+    {
+      "epoch": 1.5417499001198562,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 3859
+    },
+    {
+      "epoch": 1.5421494206951658,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 3860
+    },
+    {
+      "epoch": 1.5425489412704754,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 3861
+    },
+    {
+      "epoch": 1.5429484618457852,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 3862
+    },
+    {
+      "epoch": 1.5433479824210947,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 3863
+    },
+    {
+      "epoch": 1.5437475029964043,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 3864
+    },
+    {
+      "epoch": 1.544147023571714,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 3865
+    },
+    {
+      "epoch": 1.5445465441470234,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 3866
+    },
+    {
+      "epoch": 1.5449460647223332,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 3867
+    },
+    {
+      "epoch": 1.545345585297643,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 3868
+    },
+    {
+      "epoch": 1.5457451058729523,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 3869
+    },
+    {
+      "epoch": 1.5461446264482621,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 3870
+    },
+    {
+      "epoch": 1.5465441470235717,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 3871
+    },
+    {
+      "epoch": 1.5469436675988812,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3872
+    },
+    {
+      "epoch": 1.547343188174191,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 3873
+    },
+    {
+      "epoch": 1.5477427087495006,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 3874
+    },
+    {
+      "epoch": 1.5481422293248102,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 3875
+    },
+    {
+      "epoch": 1.54854174990012,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 3876
+    },
+    {
+      "epoch": 1.5489412704754295,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 3877
+    },
+    {
+      "epoch": 1.549340791050739,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.077,
+      "step": 3878
+    },
+    {
+      "epoch": 1.5497403116260489,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 3879
+    },
+    {
+      "epoch": 1.5501398322013584,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0819,
+      "step": 3880
+    },
+    {
+      "epoch": 1.550539352776668,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 3881
+    },
+    {
+      "epoch": 1.5509388733519778,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 3882
+    },
+    {
+      "epoch": 1.5513383939272871,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 3883
+    },
+    {
+      "epoch": 1.551737914502597,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 3884
+    },
+    {
+      "epoch": 1.5521374350779065,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3885
+    },
+    {
+      "epoch": 1.552536955653216,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 3886
+    },
+    {
+      "epoch": 1.5529364762285258,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 3887
+    },
+    {
+      "epoch": 1.5533359968038354,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0762,
+      "step": 3888
+    },
+    {
+      "epoch": 1.553735517379145,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 3889
+    },
+    {
+      "epoch": 1.5541350379544547,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 3890
+    },
+    {
+      "epoch": 1.5545345585297643,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 3891
+    },
+    {
+      "epoch": 1.5549340791050739,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 3892
+    },
+    {
+      "epoch": 1.5553335996803836,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 3893
+    },
+    {
+      "epoch": 1.5557331202556932,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0759,
+      "step": 3894
+    },
+    {
+      "epoch": 1.5561326408310028,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 3895
+    },
+    {
+      "epoch": 1.5565321614063126,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 3896
+    },
+    {
+      "epoch": 1.556931681981622,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3897
+    },
+    {
+      "epoch": 1.5573312025569317,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 3898
+    },
+    {
+      "epoch": 1.5577307231322413,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 3899
+    },
+    {
+      "epoch": 1.5581302437075508,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3900
+    },
+    {
+      "epoch": 1.5585297642828606,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 3901
+    },
+    {
+      "epoch": 1.5589292848581702,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 3902
+    },
+    {
+      "epoch": 1.5593288054334797,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 3903
+    },
+    {
+      "epoch": 1.5597283260087895,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3904
+    },
+    {
+      "epoch": 1.560127846584099,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3905
+    },
+    {
+      "epoch": 1.5605273671594087,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 3906
+    },
+    {
+      "epoch": 1.5609268877347184,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 3907
+    },
+    {
+      "epoch": 1.561326408310028,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 3908
+    },
+    {
+      "epoch": 1.5617259288853376,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3909
+    },
+    {
+      "epoch": 1.5621254494606474,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3910
+    },
+    {
+      "epoch": 1.5625249700359567,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0833,
+      "step": 3911
+    },
+    {
+      "epoch": 1.5629244906112665,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 3912
+    },
+    {
+      "epoch": 1.5633240111865763,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 3913
+    },
+    {
+      "epoch": 1.5637235317618856,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0864,
+      "step": 3914
+    },
+    {
+      "epoch": 1.5641230523371954,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 3915
+    },
+    {
+      "epoch": 1.564522572912505,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 3916
+    },
+    {
+      "epoch": 1.5649220934878145,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 3917
+    },
+    {
+      "epoch": 1.5653216140631243,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 3918
+    },
+    {
+      "epoch": 1.5657211346384339,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 3919
+    },
+    {
+      "epoch": 1.5661206552137434,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0797,
+      "step": 3920
+    },
+    {
+      "epoch": 1.5665201757890532,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 3921
+    },
+    {
+      "epoch": 1.5669196963643628,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 3922
+    },
+    {
+      "epoch": 1.5673192169396724,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 3923
+    },
+    {
+      "epoch": 1.5677187375149821,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3924
+    },
+    {
+      "epoch": 1.5681182580902915,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 3925
+    },
+    {
+      "epoch": 1.5685177786656013,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 3926
+    },
+    {
+      "epoch": 1.568917299240911,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 3927
+    },
+    {
+      "epoch": 1.5693168198162204,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3928
+    },
+    {
+      "epoch": 1.5697163403915302,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 3929
+    },
+    {
+      "epoch": 1.5701158609668397,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 3930
+    },
+    {
+      "epoch": 1.5705153815421493,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 3931
+    },
+    {
+      "epoch": 1.570914902117459,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 3932
+    },
+    {
+      "epoch": 1.5713144226927687,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 3933
+    },
+    {
+      "epoch": 1.5717139432680782,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 3934
+    },
+    {
+      "epoch": 1.572113463843388,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3935
+    },
+    {
+      "epoch": 1.5725129844186976,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0811,
+      "step": 3936
+    },
+    {
+      "epoch": 1.5729125049940071,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 3937
+    },
+    {
+      "epoch": 1.573312025569317,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 3938
+    },
+    {
+      "epoch": 1.5737115461446265,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 3939
+    },
+    {
+      "epoch": 1.574111066719936,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 3940
+    },
+    {
+      "epoch": 1.5745105872952458,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 3941
+    },
+    {
+      "epoch": 1.5749101078705552,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 3942
+    },
+    {
+      "epoch": 1.575309628445865,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 3943
+    },
+    {
+      "epoch": 1.5757091490211745,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 3944
+    },
+    {
+      "epoch": 1.576108669596484,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 3945
+    },
+    {
+      "epoch": 1.5765081901717939,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 3946
+    },
+    {
+      "epoch": 1.5769077107471035,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 3947
+    },
+    {
+      "epoch": 1.577307231322413,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0808,
+      "step": 3948
+    },
+    {
+      "epoch": 1.5777067518977228,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 3949
+    },
+    {
+      "epoch": 1.5781062724730324,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 3950
+    },
+    {
+      "epoch": 1.578505793048342,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 3951
+    },
+    {
+      "epoch": 1.5789053136236517,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 3952
+    },
+    {
+      "epoch": 1.5793048341989613,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 3953
+    },
+    {
+      "epoch": 1.5797043547742708,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3954
+    },
+    {
+      "epoch": 1.5801038753495806,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 3955
+    },
+    {
+      "epoch": 1.58050339592489,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 3956
+    },
+    {
+      "epoch": 1.5809029165001998,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 3957
+    },
+    {
+      "epoch": 1.5813024370755095,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 3958
+    },
+    {
+      "epoch": 1.581701957650819,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 3959
+    },
+    {
+      "epoch": 1.5821014782261287,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 3960
+    },
+    {
+      "epoch": 1.5825009988014382,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 3961
+    },
+    {
+      "epoch": 1.5829005193767478,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 3962
+    },
+    {
+      "epoch": 1.5833000399520576,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 3963
+    },
+    {
+      "epoch": 1.5836995605273672,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 3964
+    },
+    {
+      "epoch": 1.5840990811026767,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 3965
+    },
+    {
+      "epoch": 1.5844986016779865,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 3966
+    },
+    {
+      "epoch": 1.584898122253296,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 3967
+    },
+    {
+      "epoch": 1.5852976428286056,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 3968
+    },
+    {
+      "epoch": 1.5856971634039154,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 3969
+    },
+    {
+      "epoch": 1.5860966839792248,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 3970
+    },
+    {
+      "epoch": 1.5864962045545346,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 3971
+    },
+    {
+      "epoch": 1.5868957251298443,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 3972
+    },
+    {
+      "epoch": 1.5872952457051537,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 3973
+    },
+    {
+      "epoch": 1.5876947662804635,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 3974
+    },
+    {
+      "epoch": 1.588094286855773,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 3975
+    },
+    {
+      "epoch": 1.5884938074310826,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 3976
+    },
+    {
+      "epoch": 1.5888933280063924,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 3977
+    },
+    {
+      "epoch": 1.589292848581702,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 3978
+    },
+    {
+      "epoch": 1.5896923691570115,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 3979
+    },
+    {
+      "epoch": 1.5900918897323213,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 3980
+    },
+    {
+      "epoch": 1.5904914103076309,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 3981
+    },
+    {
+      "epoch": 1.5908909308829404,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 3982
+    },
+    {
+      "epoch": 1.5912904514582502,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 3983
+    },
+    {
+      "epoch": 1.5916899720335598,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 3984
+    },
+    {
+      "epoch": 1.5920894926088693,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 3985
+    },
+    {
+      "epoch": 1.5924890131841791,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 3986
+    },
+    {
+      "epoch": 1.5928885337594885,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 3987
+    },
+    {
+      "epoch": 1.5932880543347983,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 3988
+    },
+    {
+      "epoch": 1.5936875749101078,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 3989
+    },
+    {
+      "epoch": 1.5940870954854174,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0763,
+      "step": 3990
+    },
+    {
+      "epoch": 1.5944866160607272,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 3991
+    },
+    {
+      "epoch": 1.5948861366360367,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 3992
+    },
+    {
+      "epoch": 1.5952856572113463,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 3993
+    },
+    {
+      "epoch": 1.595685177786656,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 3994
+    },
+    {
+      "epoch": 1.5960846983619656,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 3995
+    },
+    {
+      "epoch": 1.5964842189372752,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 3996
+    },
+    {
+      "epoch": 1.596883739512585,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 3997
+    },
+    {
+      "epoch": 1.5972832600878946,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 3998
+    },
+    {
+      "epoch": 1.5976827806632041,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 3999
+    },
+    {
+      "epoch": 1.598082301238514,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 4000
+    },
+    {
+      "epoch": 1.5984818218138233,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 4001
+    },
+    {
+      "epoch": 1.598881342389133,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4002
+    },
+    {
+      "epoch": 1.5992808629644428,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4003
+    },
+    {
+      "epoch": 1.5996803835397522,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 4004
+    },
+    {
+      "epoch": 1.600079904115062,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4005
+    },
+    {
+      "epoch": 1.6004794246903715,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 4006
+    },
+    {
+      "epoch": 1.600878945265681,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4007
+    },
+    {
+      "epoch": 1.6012784658409909,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4008
+    },
+    {
+      "epoch": 1.6016779864163004,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 4009
+    },
+    {
+      "epoch": 1.60207750699161,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4010
+    },
+    {
+      "epoch": 1.6024770275669198,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4011
+    },
+    {
+      "epoch": 1.6028765481422294,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4012
+    },
+    {
+      "epoch": 1.603276068717539,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 4013
+    },
+    {
+      "epoch": 1.6036755892928487,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4014
+    },
+    {
+      "epoch": 1.604075109868158,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 4015
+    },
+    {
+      "epoch": 1.6044746304434678,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 4016
+    },
+    {
+      "epoch": 1.6048741510187776,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 4017
+    },
+    {
+      "epoch": 1.605273671594087,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 4018
+    },
+    {
+      "epoch": 1.6056731921693967,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 4019
+    },
+    {
+      "epoch": 1.6060727127447063,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 4020
+    },
+    {
+      "epoch": 1.6064722333200159,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 4021
+    },
+    {
+      "epoch": 1.6068717538953257,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 4022
+    },
+    {
+      "epoch": 1.6072712744706352,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 4023
+    },
+    {
+      "epoch": 1.6076707950459448,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 4024
+    },
+    {
+      "epoch": 1.6080703156212546,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 4025
+    },
+    {
+      "epoch": 1.6084698361965641,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 4026
+    },
+    {
+      "epoch": 1.6088693567718737,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 4027
+    },
+    {
+      "epoch": 1.6092688773471835,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4028
+    },
+    {
+      "epoch": 1.609668397922493,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 4029
+    },
+    {
+      "epoch": 1.6100679184978026,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 4030
+    },
+    {
+      "epoch": 1.6104674390731124,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0806,
+      "step": 4031
+    },
+    {
+      "epoch": 1.6108669596484217,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 4032
+    },
+    {
+      "epoch": 1.6112664802237315,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4033
+    },
+    {
+      "epoch": 1.611666000799041,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4034
+    },
+    {
+      "epoch": 1.6120655213743507,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0786,
+      "step": 4035
+    },
+    {
+      "epoch": 1.6124650419496604,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 4036
+    },
+    {
+      "epoch": 1.61286456252497,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 4037
+    },
+    {
+      "epoch": 1.6132640831002796,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 4038
+    },
+    {
+      "epoch": 1.6136636036755894,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 4039
+    },
+    {
+      "epoch": 1.614063124250899,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0818,
+      "step": 4040
+    },
+    {
+      "epoch": 1.6144626448262085,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 4041
+    },
+    {
+      "epoch": 1.6148621654015183,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4042
+    },
+    {
+      "epoch": 1.6152616859768278,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4043
+    },
+    {
+      "epoch": 1.6156612065521374,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4044
+    },
+    {
+      "epoch": 1.6160607271274472,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4045
+    },
+    {
+      "epoch": 1.6164602477027565,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 4046
+    },
+    {
+      "epoch": 1.6168597682780663,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4047
+    },
+    {
+      "epoch": 1.617259288853376,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 4048
+    },
+    {
+      "epoch": 1.6176588094286855,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 4049
+    },
+    {
+      "epoch": 1.6180583300039952,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4050
+    },
+    {
+      "epoch": 1.6184578505793048,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 4051
+    },
+    {
+      "epoch": 1.6188573711546144,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4052
+    },
+    {
+      "epoch": 1.6192568917299242,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 4053
+    },
+    {
+      "epoch": 1.6196564123052337,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 4054
+    },
+    {
+      "epoch": 1.6200559328805433,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4055
+    },
+    {
+      "epoch": 1.620455453455853,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 4056
+    },
+    {
+      "epoch": 1.6208549740311626,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4057
+    },
+    {
+      "epoch": 1.6212544946064722,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 4058
+    },
+    {
+      "epoch": 1.621654015181782,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 4059
+    },
+    {
+      "epoch": 1.6220535357570913,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 4060
+    },
+    {
+      "epoch": 1.6224530563324011,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 4061
+    },
+    {
+      "epoch": 1.622852576907711,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4062
+    },
+    {
+      "epoch": 1.6232520974830202,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 4063
+    },
+    {
+      "epoch": 1.62365161805833,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4064
+    },
+    {
+      "epoch": 1.6240511386336396,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 4065
+    },
+    {
+      "epoch": 1.6244506592089492,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4066
+    },
+    {
+      "epoch": 1.624850179784259,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 4067
+    },
+    {
+      "epoch": 1.6252497003595685,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4068
+    },
+    {
+      "epoch": 1.625649220934878,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 4069
+    },
+    {
+      "epoch": 1.6260487415101879,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 4070
+    },
+    {
+      "epoch": 1.6264482620854974,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 4071
+    },
+    {
+      "epoch": 1.626847782660807,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 4072
+    },
+    {
+      "epoch": 1.6272473032361168,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 4073
+    },
+    {
+      "epoch": 1.6276468238114263,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 4074
+    },
+    {
+      "epoch": 1.628046344386736,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 4075
+    },
+    {
+      "epoch": 1.6284458649620457,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 4076
+    },
+    {
+      "epoch": 1.628845385537355,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4077
+    },
+    {
+      "epoch": 1.6292449061126648,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 4078
+    },
+    {
+      "epoch": 1.6296444266879744,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4079
+    },
+    {
+      "epoch": 1.630043947263284,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 4080
+    },
+    {
+      "epoch": 1.6304434678385937,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 4081
+    },
+    {
+      "epoch": 1.6308429884139033,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4082
+    },
+    {
+      "epoch": 1.6312425089892129,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 4083
+    },
+    {
+      "epoch": 1.6316420295645226,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 4084
+    },
+    {
+      "epoch": 1.6320415501398322,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 4085
+    },
+    {
+      "epoch": 1.6324410707151418,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 4086
+    },
+    {
+      "epoch": 1.6328405912904516,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 4087
+    },
+    {
+      "epoch": 1.6332401118657611,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 4088
+    },
+    {
+      "epoch": 1.6336396324410707,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4089
+    },
+    {
+      "epoch": 1.6340391530163805,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 4090
+    },
+    {
+      "epoch": 1.6344386735916898,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.073,
+      "step": 4091
+    },
+    {
+      "epoch": 1.6348381941669996,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 4092
+    },
+    {
+      "epoch": 1.6352377147423094,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 4093
+    },
+    {
+      "epoch": 1.6356372353176187,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4094
+    },
+    {
+      "epoch": 1.6360367558929285,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 4095
+    },
+    {
+      "epoch": 1.636436276468238,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4096
+    },
+    {
+      "epoch": 1.6368357970435476,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0842,
+      "step": 4097
+    },
+    {
+      "epoch": 1.6372353176188574,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 4098
+    },
+    {
+      "epoch": 1.637634838194167,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4099
+    },
+    {
+      "epoch": 1.6380343587694766,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 4100
+    },
+    {
+      "epoch": 1.6384338793447863,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4101
+    },
+    {
+      "epoch": 1.638833399920096,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4102
+    },
+    {
+      "epoch": 1.6392329204954055,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0775,
+      "step": 4103
+    },
+    {
+      "epoch": 1.6396324410707153,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 4104
+    },
+    {
+      "epoch": 1.6400319616460246,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 4105
+    },
+    {
+      "epoch": 1.6404314822213344,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4106
+    },
+    {
+      "epoch": 1.6408310027966442,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4107
+    },
+    {
+      "epoch": 1.6412305233719535,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 4108
+    },
+    {
+      "epoch": 1.6416300439472633,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 4109
+    },
+    {
+      "epoch": 1.6420295645225729,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 4110
+    },
+    {
+      "epoch": 1.6424290850978824,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 4111
+    },
+    {
+      "epoch": 1.6428286056731922,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 4112
+    },
+    {
+      "epoch": 1.6432281262485018,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 4113
+    },
+    {
+      "epoch": 1.6436276468238114,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0005,
+      "loss": 1.087,
+      "step": 4114
+    },
+    {
+      "epoch": 1.6440271673991211,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0814,
+      "step": 4115
+    },
+    {
+      "epoch": 1.6444266879744307,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4116
+    },
+    {
+      "epoch": 1.6448262085497403,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 4117
+    },
+    {
+      "epoch": 1.64522572912505,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 4118
+    },
+    {
+      "epoch": 1.6456252497003596,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 4119
+    },
+    {
+      "epoch": 1.6460247702756692,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 4120
+    },
+    {
+      "epoch": 1.646424290850979,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4121
+    },
+    {
+      "epoch": 1.6468238114262883,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 4122
+    },
+    {
+      "epoch": 1.647223332001598,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4123
+    },
+    {
+      "epoch": 1.6476228525769077,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 4124
+    },
+    {
+      "epoch": 1.6480223731522172,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 4125
+    },
+    {
+      "epoch": 1.648421893727527,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4126
+    },
+    {
+      "epoch": 1.6488214143028366,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4127
+    },
+    {
+      "epoch": 1.6492209348781461,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4128
+    },
+    {
+      "epoch": 1.649620455453456,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4129
+    },
+    {
+      "epoch": 1.6500199760287655,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 4130
+    },
+    {
+      "epoch": 1.650419496604075,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 4131
+    },
+    {
+      "epoch": 1.6508190171793848,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 4132
+    },
+    {
+      "epoch": 1.6512185377546944,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4133
+    },
+    {
+      "epoch": 1.651618058330004,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 4134
+    },
+    {
+      "epoch": 1.6520175789053138,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 4135
+    },
+    {
+      "epoch": 1.652417099480623,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0903,
+      "step": 4136
+    },
+    {
+      "epoch": 1.6528166200559329,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 4137
+    },
+    {
+      "epoch": 1.6532161406312427,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 4138
+    },
+    {
+      "epoch": 1.653615661206552,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 4139
+    },
+    {
+      "epoch": 1.6540151817818618,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0809,
+      "step": 4140
+    },
+    {
+      "epoch": 1.6544147023571714,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 4141
+    },
+    {
+      "epoch": 1.654814222932481,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 4142
+    },
+    {
+      "epoch": 1.6552137435077907,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 4143
+    },
+    {
+      "epoch": 1.6556132640831003,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 4144
+    },
+    {
+      "epoch": 1.6560127846584098,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 4145
+    },
+    {
+      "epoch": 1.6564123052337196,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 4146
+    },
+    {
+      "epoch": 1.6568118258090292,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 4147
+    },
+    {
+      "epoch": 1.6572113463843388,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 4148
+    },
+    {
+      "epoch": 1.6576108669596485,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4149
+    },
+    {
+      "epoch": 1.6580103875349579,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 4150
+    },
+    {
+      "epoch": 1.6584099081102677,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 4151
+    },
+    {
+      "epoch": 1.6588094286855775,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 4152
+    },
+    {
+      "epoch": 1.6592089492608868,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 4153
+    },
+    {
+      "epoch": 1.6596084698361966,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 4154
+    },
+    {
+      "epoch": 1.6600079904115062,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 4155
+    },
+    {
+      "epoch": 1.6604075109868157,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 4156
+    },
+    {
+      "epoch": 1.6608070315621255,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 4157
+    },
+    {
+      "epoch": 1.661206552137435,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 4158
+    },
+    {
+      "epoch": 1.6616060727127446,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 4159
+    },
+    {
+      "epoch": 1.6620055932880544,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4160
+    },
+    {
+      "epoch": 1.662405113863364,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 4161
+    },
+    {
+      "epoch": 1.6628046344386735,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 4162
+    },
+    {
+      "epoch": 1.6632041550139833,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4163
+    },
+    {
+      "epoch": 1.663603675589293,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 4164
+    },
+    {
+      "epoch": 1.6640031961646025,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 4165
+    },
+    {
+      "epoch": 1.6644027167399122,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 4166
+    },
+    {
+      "epoch": 1.6648022373152216,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 4167
+    },
+    {
+      "epoch": 1.6652017578905314,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 4168
+    },
+    {
+      "epoch": 1.665601278465841,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 4169
+    },
+    {
+      "epoch": 1.6660007990411505,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4170
+    },
+    {
+      "epoch": 1.6664003196164603,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 4171
+    },
+    {
+      "epoch": 1.6667998401917699,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 4172
+    },
+    {
+      "epoch": 1.6671993607670794,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 4173
+    },
+    {
+      "epoch": 1.6675988813423892,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 4174
+    },
+    {
+      "epoch": 1.6679984019176988,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4175
+    },
+    {
+      "epoch": 1.6683979224930083,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 4176
+    },
+    {
+      "epoch": 1.6687974430683181,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 4177
+    },
+    {
+      "epoch": 1.6691969636436277,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 4178
+    },
+    {
+      "epoch": 1.6695964842189372,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4179
+    },
+    {
+      "epoch": 1.669996004794247,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 4180
+    },
+    {
+      "epoch": 1.6703955253695564,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 4181
+    },
+    {
+      "epoch": 1.6707950459448662,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 4182
+    },
+    {
+      "epoch": 1.671194566520176,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4183
+    },
+    {
+      "epoch": 1.6715940870954853,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4184
+    },
+    {
+      "epoch": 1.671993607670795,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4185
+    },
+    {
+      "epoch": 1.6723931282461046,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 4186
+    },
+    {
+      "epoch": 1.6727926488214142,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4187
+    },
+    {
+      "epoch": 1.673192169396724,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4188
+    },
+    {
+      "epoch": 1.6735916899720336,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4189
+    },
+    {
+      "epoch": 1.6739912105473431,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 4190
+    },
+    {
+      "epoch": 1.674390731122653,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0787,
+      "step": 4191
+    },
+    {
+      "epoch": 1.6747902516979625,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 4192
+    },
+    {
+      "epoch": 1.675189772273272,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 4193
+    },
+    {
+      "epoch": 1.6755892928485818,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4194
+    },
+    {
+      "epoch": 1.6759888134238912,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 4195
+    },
+    {
+      "epoch": 1.676388333999201,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 4196
+    },
+    {
+      "epoch": 1.6767878545745107,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 4197
+    },
+    {
+      "epoch": 1.67718737514982,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 4198
+    },
+    {
+      "epoch": 1.6775868957251299,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 4199
+    },
+    {
+      "epoch": 1.6779864163004394,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4200
+    },
+    {
+      "epoch": 1.678385936875749,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 4201
+    },
+    {
+      "epoch": 1.6787854574510588,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 4202
+    },
+    {
+      "epoch": 1.6791849780263683,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4203
+    },
+    {
+      "epoch": 1.679584498601678,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0798,
+      "step": 4204
+    },
+    {
+      "epoch": 1.6799840191769877,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4205
+    },
+    {
+      "epoch": 1.6803835397522973,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 4206
+    },
+    {
+      "epoch": 1.6807830603276068,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 4207
+    },
+    {
+      "epoch": 1.6811825809029166,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 4208
+    },
+    {
+      "epoch": 1.6815821014782262,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4209
+    },
+    {
+      "epoch": 1.6819816220535357,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4210
+    },
+    {
+      "epoch": 1.6823811426288455,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4211
+    },
+    {
+      "epoch": 1.6827806632041549,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 4212
+    },
+    {
+      "epoch": 1.6831801837794647,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 4213
+    },
+    {
+      "epoch": 1.6835797043547742,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 4214
+    },
+    {
+      "epoch": 1.6839792249300838,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0865,
+      "step": 4215
+    },
+    {
+      "epoch": 1.6843787455053936,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4216
+    },
+    {
+      "epoch": 1.6847782660807031,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 4217
+    },
+    {
+      "epoch": 1.6851777866560127,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 4218
+    },
+    {
+      "epoch": 1.6855773072313225,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 4219
+    },
+    {
+      "epoch": 1.685976827806632,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 4220
+    },
+    {
+      "epoch": 1.6863763483819416,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 4221
+    },
+    {
+      "epoch": 1.6867758689572514,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0804,
+      "step": 4222
+    },
+    {
+      "epoch": 1.687175389532561,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 4223
+    },
+    {
+      "epoch": 1.6875749101078705,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4224
+    },
+    {
+      "epoch": 1.6879744306831803,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 4225
+    },
+    {
+      "epoch": 1.6883739512584897,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 4226
+    },
+    {
+      "epoch": 1.6887734718337994,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 4227
+    },
+    {
+      "epoch": 1.6891729924091092,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0746,
+      "step": 4228
+    },
+    {
+      "epoch": 1.6895725129844186,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 4229
+    },
+    {
+      "epoch": 1.6899720335597284,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 4230
+    },
+    {
+      "epoch": 1.690371554135038,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4231
+    },
+    {
+      "epoch": 1.6907710747103475,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4232
+    },
+    {
+      "epoch": 1.6911705952856573,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 4233
+    },
+    {
+      "epoch": 1.6915701158609668,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4234
+    },
+    {
+      "epoch": 1.6919696364362764,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 4235
+    },
+    {
+      "epoch": 1.6923691570115862,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 4236
+    },
+    {
+      "epoch": 1.6927686775868958,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 4237
+    },
+    {
+      "epoch": 1.6931681981622053,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0836,
+      "step": 4238
+    },
+    {
+      "epoch": 1.693567718737515,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4239
+    },
+    {
+      "epoch": 1.6939672393128244,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6943667598881342,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 4241
+    },
+    {
+      "epoch": 1.694766280463444,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 4242
+    },
+    {
+      "epoch": 1.6951658010387534,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 4243
+    },
+    {
+      "epoch": 1.6955653216140631,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 4244
+    },
+    {
+      "epoch": 1.6959648421893727,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 4245
+    },
+    {
+      "epoch": 1.6963643627646823,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 4246
+    },
+    {
+      "epoch": 1.696763883339992,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 4247
+    },
+    {
+      "epoch": 1.6971634039153016,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4248
+    },
+    {
+      "epoch": 1.6975629244906112,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 4249
+    },
+    {
+      "epoch": 1.697962445065921,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 4250
+    },
+    {
+      "epoch": 1.6983619656412305,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 4251
+    },
+    {
+      "epoch": 1.69876148621654,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 4252
+    },
+    {
+      "epoch": 1.69916100679185,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 4253
+    },
+    {
+      "epoch": 1.6995605273671595,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4254
+    },
+    {
+      "epoch": 1.699960047942469,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 4255
+    },
+    {
+      "epoch": 1.7003595685177788,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4256
+    },
+    {
+      "epoch": 1.7007590890930881,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 4257
+    },
+    {
+      "epoch": 1.701158609668398,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 4258
+    },
+    {
+      "epoch": 1.7015581302437075,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 4259
+    },
+    {
+      "epoch": 1.701957650819017,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4260
+    },
+    {
+      "epoch": 1.7023571713943269,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 4261
+    },
+    {
+      "epoch": 1.7027566919696364,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0769,
+      "step": 4262
+    },
+    {
+      "epoch": 1.703156212544946,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 4263
+    },
+    {
+      "epoch": 1.7035557331202558,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 4264
+    },
+    {
+      "epoch": 1.7039552536955653,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4265
+    },
+    {
+      "epoch": 1.704354774270875,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 4266
+    },
+    {
+      "epoch": 1.7047542948461847,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 4267
+    },
+    {
+      "epoch": 1.7051538154214942,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 4268
+    },
+    {
+      "epoch": 1.7055533359968038,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4269
+    },
+    {
+      "epoch": 1.7059528565721136,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4270
+    },
+    {
+      "epoch": 1.706352377147423,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4271
+    },
+    {
+      "epoch": 1.7067518977227327,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 4272
+    },
+    {
+      "epoch": 1.7071514182980425,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4273
+    },
+    {
+      "epoch": 1.7075509388733519,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 4274
+    },
+    {
+      "epoch": 1.7079504594486616,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 4275
+    },
+    {
+      "epoch": 1.7083499800239712,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0803,
+      "step": 4276
+    },
+    {
+      "epoch": 1.7087495005992808,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 4277
+    },
+    {
+      "epoch": 1.7091490211745906,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 4278
+    },
+    {
+      "epoch": 1.7095485417499001,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4279
+    },
+    {
+      "epoch": 1.7099480623252097,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0762,
+      "step": 4280
+    },
+    {
+      "epoch": 1.7103475829005195,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 4281
+    },
+    {
+      "epoch": 1.710747103475829,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 4282
+    },
+    {
+      "epoch": 1.7111466240511386,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 4283
+    },
+    {
+      "epoch": 1.7115461446264484,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4284
+    },
+    {
+      "epoch": 1.7119456652017577,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4285
+    },
+    {
+      "epoch": 1.7123451857770675,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 4286
+    },
+    {
+      "epoch": 1.7127447063523773,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4287
+    },
+    {
+      "epoch": 1.7131442269276866,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 4288
+    },
+    {
+      "epoch": 1.7135437475029964,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 4289
+    },
+    {
+      "epoch": 1.713943268078306,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 4290
+    },
+    {
+      "epoch": 1.7143427886536156,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4291
+    },
+    {
+      "epoch": 1.7147423092289253,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4292
+    },
+    {
+      "epoch": 1.715141829804235,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 4293
+    },
+    {
+      "epoch": 1.7155413503795445,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4294
+    },
+    {
+      "epoch": 1.7159408709548543,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 4295
+    },
+    {
+      "epoch": 1.7163403915301638,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 4296
+    },
+    {
+      "epoch": 1.7167399121054734,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 4297
+    },
+    {
+      "epoch": 1.7171394326807832,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 4298
+    },
+    {
+      "epoch": 1.7175389532560927,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 4299
+    },
+    {
+      "epoch": 1.7179384738314023,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4300
+    },
+    {
+      "epoch": 1.718337994406712,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 4301
+    },
+    {
+      "epoch": 1.7187375149820214,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4302
+    },
+    {
+      "epoch": 1.7191370355573312,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 4303
+    },
+    {
+      "epoch": 1.7195365561326408,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4304
+    },
+    {
+      "epoch": 1.7199360767079503,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 4305
+    },
+    {
+      "epoch": 1.7203355972832601,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 4306
+    },
+    {
+      "epoch": 1.7207351178585697,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4307
+    },
+    {
+      "epoch": 1.7211346384338793,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4308
+    },
+    {
+      "epoch": 1.721534159009189,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4309
+    },
+    {
+      "epoch": 1.7219336795844986,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4310
+    },
+    {
+      "epoch": 1.7223332001598082,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 4311
+    },
+    {
+      "epoch": 1.722732720735118,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 4312
+    },
+    {
+      "epoch": 1.7231322413104275,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 4313
+    },
+    {
+      "epoch": 1.723531761885737,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 4314
+    },
+    {
+      "epoch": 1.7239312824610469,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 4315
+    },
+    {
+      "epoch": 1.7243308030363562,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 4316
+    },
+    {
+      "epoch": 1.724730323611666,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4317
+    },
+    {
+      "epoch": 1.7251298441869758,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4318
+    },
+    {
+      "epoch": 1.7255293647622851,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 4319
+    },
+    {
+      "epoch": 1.725928885337595,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 4320
+    },
+    {
+      "epoch": 1.7263284059129045,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4321
+    },
+    {
+      "epoch": 1.726727926488214,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 4322
+    },
+    {
+      "epoch": 1.7271274470635238,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 4323
+    },
+    {
+      "epoch": 1.7275269676388334,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 4324
+    },
+    {
+      "epoch": 1.727926488214143,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 4325
+    },
+    {
+      "epoch": 1.7283260087894528,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4326
+    },
+    {
+      "epoch": 1.7287255293647623,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 4327
+    },
+    {
+      "epoch": 1.7291250499400719,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 4328
+    },
+    {
+      "epoch": 1.7295245705153817,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 4329
+    },
+    {
+      "epoch": 1.729924091090691,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4330
+    },
+    {
+      "epoch": 1.7303236116660008,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 4331
+    },
+    {
+      "epoch": 1.7307231322413106,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 4332
+    },
+    {
+      "epoch": 1.73112265281662,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 4333
+    },
+    {
+      "epoch": 1.7315221733919297,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4334
+    },
+    {
+      "epoch": 1.7319216939672393,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 4335
+    },
+    {
+      "epoch": 1.7323212145425488,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 4336
+    },
+    {
+      "epoch": 1.7327207351178586,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 4337
+    },
+    {
+      "epoch": 1.7331202556931682,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 4338
+    },
+    {
+      "epoch": 1.7335197762684778,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 4339
+    },
+    {
+      "epoch": 1.7339192968437875,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 4340
+    },
+    {
+      "epoch": 1.734318817419097,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 4341
+    },
+    {
+      "epoch": 1.7347183379944067,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 4342
+    },
+    {
+      "epoch": 1.7351178585697165,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 4343
+    },
+    {
+      "epoch": 1.735517379145026,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 4344
+    },
+    {
+      "epoch": 1.7359168997203356,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 4345
+    },
+    {
+      "epoch": 1.7363164202956454,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 4346
+    },
+    {
+      "epoch": 1.7367159408709547,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 4347
+    },
+    {
+      "epoch": 1.7371154614462645,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4348
+    },
+    {
+      "epoch": 1.737514982021574,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 4349
+    },
+    {
+      "epoch": 1.7379145025968836,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 4350
+    },
+    {
+      "epoch": 1.7383140231721934,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4351
+    },
+    {
+      "epoch": 1.738713543747503,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4352
+    },
+    {
+      "epoch": 1.7391130643228125,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 4353
+    },
+    {
+      "epoch": 1.7395125848981223,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0729,
+      "step": 4354
+    },
+    {
+      "epoch": 1.739912105473432,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4355
+    },
+    {
+      "epoch": 1.7403116260487415,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 4356
+    },
+    {
+      "epoch": 1.7407111466240512,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.08,
+      "step": 4357
+    },
+    {
+      "epoch": 1.7411106671993608,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 4358
+    },
+    {
+      "epoch": 1.7415101877746704,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 4359
+    },
+    {
+      "epoch": 1.7419097083499802,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 4360
+    },
+    {
+      "epoch": 1.7423092289252895,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 4361
+    },
+    {
+      "epoch": 1.7427087495005993,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4362
+    },
+    {
+      "epoch": 1.743108270075909,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 4363
+    },
+    {
+      "epoch": 1.7435077906512184,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 4364
+    },
+    {
+      "epoch": 1.7439073112265282,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0819,
+      "step": 4365
+    },
+    {
+      "epoch": 1.7443068318018378,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 4366
+    },
+    {
+      "epoch": 1.7447063523771473,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4367
+    },
+    {
+      "epoch": 1.7451058729524571,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4368
+    },
+    {
+      "epoch": 1.7455053935277667,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4369
+    },
+    {
+      "epoch": 1.7459049141030762,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4370
+    },
+    {
+      "epoch": 1.746304434678386,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 4371
+    },
+    {
+      "epoch": 1.7467039552536956,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 4372
+    },
+    {
+      "epoch": 1.7471034758290052,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 4373
+    },
+    {
+      "epoch": 1.747502996404315,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 4374
+    },
+    {
+      "epoch": 1.7479025169796243,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 4375
+    },
+    {
+      "epoch": 1.748302037554934,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 4376
+    },
+    {
+      "epoch": 1.7487015581302439,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4377
+    },
+    {
+      "epoch": 1.7491010787055532,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 4378
+    },
+    {
+      "epoch": 1.749500599280863,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 4379
+    },
+    {
+      "epoch": 1.7499001198561726,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4380
+    },
+    {
+      "epoch": 1.7502996404314821,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4381
+    },
+    {
+      "epoch": 1.750699161006792,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 4382
+    },
+    {
+      "epoch": 1.7510986815821015,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 4383
+    },
+    {
+      "epoch": 1.751498202157411,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0744,
+      "step": 4384
+    },
+    {
+      "epoch": 1.7518977227327208,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0714,
+      "step": 4385
+    },
+    {
+      "epoch": 1.7522972433080304,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 4386
+    },
+    {
+      "epoch": 1.75269676388334,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4387
+    },
+    {
+      "epoch": 1.7530962844586497,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 4388
+    },
+    {
+      "epoch": 1.7534958050339593,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0784,
+      "step": 4389
+    },
+    {
+      "epoch": 1.7538953256092689,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 4390
+    },
+    {
+      "epoch": 1.7542948461845786,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 4391
+    },
+    {
+      "epoch": 1.754694366759888,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 4392
+    },
+    {
+      "epoch": 1.7550938873351978,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 4393
+    },
+    {
+      "epoch": 1.7554934079105073,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 4394
+    },
+    {
+      "epoch": 1.755892928485817,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 4395
+    },
+    {
+      "epoch": 1.7562924490611267,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4396
+    },
+    {
+      "epoch": 1.7566919696364363,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4397
+    },
+    {
+      "epoch": 1.7570914902117458,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4398
+    },
+    {
+      "epoch": 1.7574910107870556,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4399
+    },
+    {
+      "epoch": 1.7578905313623652,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 4400
+    },
+    {
+      "epoch": 1.7582900519376747,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 4401
+    },
+    {
+      "epoch": 1.7586895725129845,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4402
+    },
+    {
+      "epoch": 1.759089093088294,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4403
+    },
+    {
+      "epoch": 1.7594886136636037,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 4404
+    },
+    {
+      "epoch": 1.7598881342389134,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 4405
+    },
+    {
+      "epoch": 1.7602876548142228,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 4406
+    },
+    {
+      "epoch": 1.7606871753895326,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 4407
+    },
+    {
+      "epoch": 1.7610866959648424,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4408
+    },
+    {
+      "epoch": 1.7614862165401517,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 4409
+    },
+    {
+      "epoch": 1.7618857371154615,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 4410
+    },
+    {
+      "epoch": 1.762285257690771,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 4411
+    },
+    {
+      "epoch": 1.7626847782660806,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 4412
+    },
+    {
+      "epoch": 1.7630842988413904,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4413
+    },
+    {
+      "epoch": 1.7634838194167,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 4414
+    },
+    {
+      "epoch": 1.7638833399920095,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4415
+    },
+    {
+      "epoch": 1.7642828605673193,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 4416
+    },
+    {
+      "epoch": 1.7646823811426289,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 4417
+    },
+    {
+      "epoch": 1.7650819017179384,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 4418
+    },
+    {
+      "epoch": 1.7654814222932482,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 4419
+    },
+    {
+      "epoch": 1.7658809428685576,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 4420
+    },
+    {
+      "epoch": 1.7662804634438674,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4421
+    },
+    {
+      "epoch": 1.7666799840191771,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4422
+    },
+    {
+      "epoch": 1.7670795045944865,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 4423
+    },
+    {
+      "epoch": 1.7674790251697963,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4424
+    },
+    {
+      "epoch": 1.7678785457451058,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 4425
+    },
+    {
+      "epoch": 1.7682780663204154,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 4426
+    },
+    {
+      "epoch": 1.7686775868957252,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 4427
+    },
+    {
+      "epoch": 1.7690771074710347,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4428
+    },
+    {
+      "epoch": 1.7694766280463443,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4429
+    },
+    {
+      "epoch": 1.769876148621654,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 4430
+    },
+    {
+      "epoch": 1.7702756691969637,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 4431
+    },
+    {
+      "epoch": 1.7706751897722732,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 4432
+    },
+    {
+      "epoch": 1.771074710347583,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 4433
+    },
+    {
+      "epoch": 1.7714742309228926,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4434
+    },
+    {
+      "epoch": 1.7718737514982021,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4435
+    },
+    {
+      "epoch": 1.772273272073512,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 4436
+    },
+    {
+      "epoch": 1.7726727926488213,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4437
+    },
+    {
+      "epoch": 1.773072313224131,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 4438
+    },
+    {
+      "epoch": 1.7734718337994406,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 4439
+    },
+    {
+      "epoch": 1.7738713543747502,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 4440
+    },
+    {
+      "epoch": 1.77427087495006,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 4441
+    },
+    {
+      "epoch": 1.7746703955253695,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4442
+    },
+    {
+      "epoch": 1.775069916100679,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 4443
+    },
+    {
+      "epoch": 1.7754694366759889,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 4444
+    },
+    {
+      "epoch": 1.7758689572512985,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 4445
+    },
+    {
+      "epoch": 1.776268477826608,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 4446
+    },
+    {
+      "epoch": 1.7766679984019178,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 4447
+    },
+    {
+      "epoch": 1.7770675189772274,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 4448
+    },
+    {
+      "epoch": 1.777467039552537,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0362,
+      "step": 4449
+    },
+    {
+      "epoch": 1.7778665601278467,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 4450
+    },
+    {
+      "epoch": 1.778266080703156,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 4451
+    },
+    {
+      "epoch": 1.7786656012784658,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 4452
+    },
+    {
+      "epoch": 1.7790651218537756,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 4453
+    },
+    {
+      "epoch": 1.779464642429085,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4454
+    },
+    {
+      "epoch": 1.7798641630043948,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4455
+    },
+    {
+      "epoch": 1.7802636835797043,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 4456
+    },
+    {
+      "epoch": 1.780663204155014,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 4457
+    },
+    {
+      "epoch": 1.7810627247303237,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4458
+    },
+    {
+      "epoch": 1.7814622453056332,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 4459
+    },
+    {
+      "epoch": 1.7818617658809428,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 4460
+    },
+    {
+      "epoch": 1.7822612864562526,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4461
+    },
+    {
+      "epoch": 1.7826608070315622,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4462
+    },
+    {
+      "epoch": 1.7830603276068717,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4463
+    },
+    {
+      "epoch": 1.7834598481821815,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0826,
+      "step": 4464
+    },
+    {
+      "epoch": 1.7838593687574908,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 4465
+    },
+    {
+      "epoch": 1.7842588893328006,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 4466
+    },
+    {
+      "epoch": 1.7846584099081104,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 4467
+    },
+    {
+      "epoch": 1.7850579304834198,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 4468
+    },
+    {
+      "epoch": 1.7854574510587295,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 4469
+    },
+    {
+      "epoch": 1.7858569716340391,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 4470
+    },
+    {
+      "epoch": 1.7862564922093487,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 4471
+    },
+    {
+      "epoch": 1.7866560127846585,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4472
+    },
+    {
+      "epoch": 1.787055533359968,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 4473
+    },
+    {
+      "epoch": 1.7874550539352776,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 4474
+    },
+    {
+      "epoch": 1.7878545745105874,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4475
+    },
+    {
+      "epoch": 1.788254095085897,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 4476
+    },
+    {
+      "epoch": 1.7886536156612065,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 4477
+    },
+    {
+      "epoch": 1.7890531362365163,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 4478
+    },
+    {
+      "epoch": 1.7894526568118259,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 4479
+    },
+    {
+      "epoch": 1.7898521773871354,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 4480
+    },
+    {
+      "epoch": 1.7902516979624452,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 4481
+    },
+    {
+      "epoch": 1.7906512185377546,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0802,
+      "step": 4482
+    },
+    {
+      "epoch": 1.7910507391130643,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4483
+    },
+    {
+      "epoch": 1.791450259688374,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 4484
+    },
+    {
+      "epoch": 1.7918497802636835,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 4485
+    },
+    {
+      "epoch": 1.7922493008389933,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 4486
+    },
+    {
+      "epoch": 1.7926488214143028,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4487
+    },
+    {
+      "epoch": 1.7930483419896124,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 4488
+    },
+    {
+      "epoch": 1.7934478625649222,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 4489
+    },
+    {
+      "epoch": 1.7938473831402317,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 4490
+    },
+    {
+      "epoch": 1.7942469037155413,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 4491
+    },
+    {
+      "epoch": 1.794646424290851,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 4492
+    },
+    {
+      "epoch": 1.7950459448661606,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0823,
+      "step": 4493
+    },
+    {
+      "epoch": 1.7954454654414702,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 4494
+    },
+    {
+      "epoch": 1.79584498601678,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 4495
+    },
+    {
+      "epoch": 1.7962445065920893,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4496
+    },
+    {
+      "epoch": 1.7966440271673991,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 4497
+    },
+    {
+      "epoch": 1.797043547742709,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 4498
+    },
+    {
+      "epoch": 1.7974430683180183,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 4499
+    },
+    {
+      "epoch": 1.797842588893328,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0742,
+      "step": 4500
+    },
+    {
+      "epoch": 1.7982421094686376,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 4501
+    },
+    {
+      "epoch": 1.7986416300439472,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 4502
+    },
+    {
+      "epoch": 1.799041150619257,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4503
+    },
+    {
+      "epoch": 1.7994406711945665,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 4504
+    },
+    {
+      "epoch": 1.799840191769876,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 4505
+    },
+    {
+      "epoch": 1.8002397123451859,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 4506
+    },
+    {
+      "epoch": 1.8006392329204954,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 4507
+    },
+    {
+      "epoch": 1.801038753495805,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 4508
+    },
+    {
+      "epoch": 1.8014382740711148,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 4509
+    },
+    {
+      "epoch": 1.8018377946464241,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 4510
+    },
+    {
+      "epoch": 1.802237315221734,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 4511
+    },
+    {
+      "epoch": 1.8026368357970437,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4512
+    },
+    {
+      "epoch": 1.803036356372353,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 4513
+    },
+    {
+      "epoch": 1.8034358769476628,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 4514
+    },
+    {
+      "epoch": 1.8038353975229724,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 4515
+    },
+    {
+      "epoch": 1.804234918098282,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0779,
+      "step": 4516
+    },
+    {
+      "epoch": 1.8046344386735917,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 4517
+    },
+    {
+      "epoch": 1.8050339592489013,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4518
+    },
+    {
+      "epoch": 1.8054334798242109,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4519
+    },
+    {
+      "epoch": 1.8058330003995207,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 4520
+    },
+    {
+      "epoch": 1.8062325209748302,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4521
+    },
+    {
+      "epoch": 1.8066320415501398,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 4522
+    },
+    {
+      "epoch": 1.8070315621254496,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 4523
+    },
+    {
+      "epoch": 1.8074310827007591,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4524
+    },
+    {
+      "epoch": 1.8078306032760687,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 4525
+    },
+    {
+      "epoch": 1.8082301238513785,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4526
+    },
+    {
+      "epoch": 1.8086296444266878,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 4527
+    },
+    {
+      "epoch": 1.8090291650019976,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 4528
+    },
+    {
+      "epoch": 1.8094286855773072,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4529
+    },
+    {
+      "epoch": 1.8098282061526167,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4530
+    },
+    {
+      "epoch": 1.8102277267279265,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 4531
+    },
+    {
+      "epoch": 1.810627247303236,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4532
+    },
+    {
+      "epoch": 1.8110267678785457,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4533
+    },
+    {
+      "epoch": 1.8114262884538554,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4534
+    },
+    {
+      "epoch": 1.811825809029165,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 4535
+    },
+    {
+      "epoch": 1.8122253296044746,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4536
+    },
+    {
+      "epoch": 1.8126248501797844,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4537
+    },
+    {
+      "epoch": 1.813024370755094,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 4538
+    },
+    {
+      "epoch": 1.8134238913304035,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4539
+    },
+    {
+      "epoch": 1.8138234119057133,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 4540
+    },
+    {
+      "epoch": 1.8142229324810226,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.082,
+      "step": 4541
+    },
+    {
+      "epoch": 1.8146224530563324,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 4542
+    },
+    {
+      "epoch": 1.8150219736316422,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4543
+    },
+    {
+      "epoch": 1.8154214942069515,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 4544
+    },
+    {
+      "epoch": 1.8158210147822613,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.075,
+      "step": 4545
+    },
+    {
+      "epoch": 1.8162205353575709,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0778,
+      "step": 4546
+    },
+    {
+      "epoch": 1.8166200559328805,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4547
+    },
+    {
+      "epoch": 1.8170195765081902,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 4548
+    },
+    {
+      "epoch": 1.8174190970834998,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4549
+    },
+    {
+      "epoch": 1.8178186176588094,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4550
+    },
+    {
+      "epoch": 1.8182181382341192,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 4551
+    },
+    {
+      "epoch": 1.8186176588094287,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 4552
+    },
+    {
+      "epoch": 1.8190171793847383,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 4553
+    },
+    {
+      "epoch": 1.819416699960048,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4554
+    },
+    {
+      "epoch": 1.8198162205353574,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 4555
+    },
+    {
+      "epoch": 1.8202157411106672,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 4556
+    },
+    {
+      "epoch": 1.820615261685977,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 4557
+    },
+    {
+      "epoch": 1.8210147822612863,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 4558
+    },
+    {
+      "epoch": 1.821414302836596,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 4559
+    },
+    {
+      "epoch": 1.8218138234119057,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 4560
+    },
+    {
+      "epoch": 1.8222133439872152,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 4561
+    },
+    {
+      "epoch": 1.822612864562525,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 4562
+    },
+    {
+      "epoch": 1.8230123851378346,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4563
+    },
+    {
+      "epoch": 1.8234119057131442,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 4564
+    },
+    {
+      "epoch": 1.823811426288454,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 4565
+    },
+    {
+      "epoch": 1.8242109468637635,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4566
+    },
+    {
+      "epoch": 1.824610467439073,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4567
+    },
+    {
+      "epoch": 1.8250099880143829,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 4568
+    },
+    {
+      "epoch": 1.8254095085896924,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 4569
+    },
+    {
+      "epoch": 1.825809029165002,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4570
+    },
+    {
+      "epoch": 1.8262085497403118,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 4571
+    },
+    {
+      "epoch": 1.8266080703156211,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4572
+    },
+    {
+      "epoch": 1.827007590890931,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 4573
+    },
+    {
+      "epoch": 1.8274071114662405,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 4574
+    },
+    {
+      "epoch": 1.82780663204155,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4575
+    },
+    {
+      "epoch": 1.8282061526168598,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 4576
+    },
+    {
+      "epoch": 1.8286056731921694,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0785,
+      "step": 4577
+    },
+    {
+      "epoch": 1.829005193767479,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4578
+    },
+    {
+      "epoch": 1.8294047143427887,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 4579
+    },
+    {
+      "epoch": 1.8298042349180983,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0757,
+      "step": 4580
+    },
+    {
+      "epoch": 1.8302037554934079,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4581
+    },
+    {
+      "epoch": 1.8306032760687176,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 4582
+    },
+    {
+      "epoch": 1.8310027966440272,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4583
+    },
+    {
+      "epoch": 1.8314023172193368,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4584
+    },
+    {
+      "epoch": 1.8318018377946466,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 4585
+    },
+    {
+      "epoch": 1.832201358369956,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 4586
+    },
+    {
+      "epoch": 1.8326008789452657,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 4587
+    },
+    {
+      "epoch": 1.8330003995205755,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 4588
+    },
+    {
+      "epoch": 1.8333999200958848,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4589
+    },
+    {
+      "epoch": 1.8337994406711946,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4590
+    },
+    {
+      "epoch": 1.8341989612465042,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 4591
+    },
+    {
+      "epoch": 1.8345984818218137,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 4592
+    },
+    {
+      "epoch": 1.8349980023971235,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 4593
+    },
+    {
+      "epoch": 1.835397522972433,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4594
+    },
+    {
+      "epoch": 1.8357970435477426,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 4595
+    },
+    {
+      "epoch": 1.8361965641230524,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4596
+    },
+    {
+      "epoch": 1.836596084698362,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 4597
+    },
+    {
+      "epoch": 1.8369956052736716,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4598
+    },
+    {
+      "epoch": 1.8373951258489813,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 4599
+    },
+    {
+      "epoch": 1.8377946464242907,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4600
+    },
+    {
+      "epoch": 1.8381941669996005,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 4601
+    },
+    {
+      "epoch": 1.8385936875749103,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4602
+    },
+    {
+      "epoch": 1.8389932081502196,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 4603
+    },
+    {
+      "epoch": 1.8393927287255294,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 4604
+    },
+    {
+      "epoch": 1.839792249300839,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 4605
+    },
+    {
+      "epoch": 1.8401917698761485,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4606
+    },
+    {
+      "epoch": 1.8405912904514583,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 4607
+    },
+    {
+      "epoch": 1.8409908110267679,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 4608
+    },
+    {
+      "epoch": 1.8413903316020774,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 4609
+    },
+    {
+      "epoch": 1.8417898521773872,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 4610
+    },
+    {
+      "epoch": 1.8421893727526968,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 4611
+    },
+    {
+      "epoch": 1.8425888933280063,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 4612
+    },
+    {
+      "epoch": 1.8429884139033161,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4613
+    },
+    {
+      "epoch": 1.8433879344786257,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 4614
+    },
+    {
+      "epoch": 1.8437874550539353,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 4615
+    },
+    {
+      "epoch": 1.844186975629245,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4616
+    },
+    {
+      "epoch": 1.8445864962045544,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 4617
+    },
+    {
+      "epoch": 1.8449860167798642,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4618
+    },
+    {
+      "epoch": 1.8453855373551737,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 4619
+    },
+    {
+      "epoch": 1.8457850579304833,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 4620
+    },
+    {
+      "epoch": 1.846184578505793,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0735,
+      "step": 4621
+    },
+    {
+      "epoch": 1.8465840990811027,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 4622
+    },
+    {
+      "epoch": 1.8469836196564122,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 4623
+    },
+    {
+      "epoch": 1.847383140231722,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 4624
+    },
+    {
+      "epoch": 1.8477826608070316,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 4625
+    },
+    {
+      "epoch": 1.8481821813823411,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 4626
+    },
+    {
+      "epoch": 1.848581701957651,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4627
+    },
+    {
+      "epoch": 1.8489812225329605,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 4628
+    },
+    {
+      "epoch": 1.84938074310827,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0792,
+      "step": 4629
+    },
+    {
+      "epoch": 1.8497802636835798,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 4630
+    },
+    {
+      "epoch": 1.8501797842588892,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 4631
+    },
+    {
+      "epoch": 1.850579304834199,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 4632
+    },
+    {
+      "epoch": 1.8509788254095088,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 4633
+    },
+    {
+      "epoch": 1.851378345984818,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4634
+    },
+    {
+      "epoch": 1.8517778665601279,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0817,
+      "step": 4635
+    },
+    {
+      "epoch": 1.8521773871354374,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 4636
+    },
+    {
+      "epoch": 1.852576907710747,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 4637
+    },
+    {
+      "epoch": 1.8529764282860568,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 4638
+    },
+    {
+      "epoch": 1.8533759488613664,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 4639
+    },
+    {
+      "epoch": 1.853775469436676,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 4640
+    },
+    {
+      "epoch": 1.8541749900119857,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4641
+    },
+    {
+      "epoch": 1.8545745105872953,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4642
+    },
+    {
+      "epoch": 1.8549740311626048,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 4643
+    },
+    {
+      "epoch": 1.8553735517379146,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 4644
+    },
+    {
+      "epoch": 1.855773072313224,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 4645
+    },
+    {
+      "epoch": 1.8561725928885338,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 4646
+    },
+    {
+      "epoch": 1.8565721134638435,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 4647
+    },
+    {
+      "epoch": 1.8569716340391529,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4648
+    },
+    {
+      "epoch": 1.8573711546144627,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 4649
+    },
+    {
+      "epoch": 1.8577706751897722,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4650
+    },
+    {
+      "epoch": 1.8581701957650818,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 4651
+    },
+    {
+      "epoch": 1.8585697163403916,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 4652
+    },
+    {
+      "epoch": 1.8589692369157012,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 4653
+    },
+    {
+      "epoch": 1.8593687574910107,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 4654
+    },
+    {
+      "epoch": 1.8597682780663205,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 4655
+    },
+    {
+      "epoch": 1.86016779864163,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 4656
+    },
+    {
+      "epoch": 1.8605673192169396,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 4657
+    },
+    {
+      "epoch": 1.8609668397922494,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4658
+    },
+    {
+      "epoch": 1.861366360367559,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 4659
+    },
+    {
+      "epoch": 1.8617658809428685,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 4660
+    },
+    {
+      "epoch": 1.8621654015181783,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4661
+    },
+    {
+      "epoch": 1.8625649220934877,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 4662
+    },
+    {
+      "epoch": 1.8629644426687975,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 4663
+    },
+    {
+      "epoch": 1.863363963244107,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4664
+    },
+    {
+      "epoch": 1.8637634838194166,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 4665
+    },
+    {
+      "epoch": 1.8641630043947264,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0857,
+      "step": 4666
+    },
+    {
+      "epoch": 1.864562524970036,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 4667
+    },
+    {
+      "epoch": 1.8649620455453455,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 4668
+    },
+    {
+      "epoch": 1.8653615661206553,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4669
+    },
+    {
+      "epoch": 1.8657610866959649,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 4670
+    },
+    {
+      "epoch": 1.8661606072712744,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4671
+    },
+    {
+      "epoch": 1.8665601278465842,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4672
+    },
+    {
+      "epoch": 1.8669596484218938,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 4673
+    },
+    {
+      "epoch": 1.8673591689972033,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4674
+    },
+    {
+      "epoch": 1.8677586895725131,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4675
+    },
+    {
+      "epoch": 1.8681582101478225,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4676
+    },
+    {
+      "epoch": 1.8685577307231322,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 4677
+    },
+    {
+      "epoch": 1.868957251298442,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 4678
+    },
+    {
+      "epoch": 1.8693567718737514,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 4679
+    },
+    {
+      "epoch": 1.8697562924490612,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 4680
+    },
+    {
+      "epoch": 1.8701558130243707,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 4681
+    },
+    {
+      "epoch": 1.8705553335996803,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 4682
+    },
+    {
+      "epoch": 1.87095485417499,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 4683
+    },
+    {
+      "epoch": 1.8713543747502996,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4684
+    },
+    {
+      "epoch": 1.8717538953256092,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 4685
+    },
+    {
+      "epoch": 1.872153415900919,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 4686
+    },
+    {
+      "epoch": 1.8725529364762286,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 4687
+    },
+    {
+      "epoch": 1.8729524570515381,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 4688
+    },
+    {
+      "epoch": 1.873351977626848,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0383,
+      "step": 4689
+    },
+    {
+      "epoch": 1.8737514982021573,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 4690
+    },
+    {
+      "epoch": 1.874151018777467,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 4691
+    },
+    {
+      "epoch": 1.8745505393527768,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 4692
+    },
+    {
+      "epoch": 1.8749500599280862,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 4693
+    },
+    {
+      "epoch": 1.875349580503396,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 4694
+    },
+    {
+      "epoch": 1.8757491010787055,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 4695
+    },
+    {
+      "epoch": 1.876148621654015,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 4696
+    },
+    {
+      "epoch": 1.8765481422293249,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 4697
+    },
+    {
+      "epoch": 1.8769476628046344,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 4698
+    },
+    {
+      "epoch": 1.877347183379944,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4699
+    },
+    {
+      "epoch": 1.8777467039552538,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 4700
+    },
+    {
+      "epoch": 1.8781462245305633,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 4701
+    },
+    {
+      "epoch": 1.878545745105873,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4702
+    },
+    {
+      "epoch": 1.8789452656811827,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 4703
+    },
+    {
+      "epoch": 1.8793447862564923,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.081,
+      "step": 4704
+    },
+    {
+      "epoch": 1.8797443068318018,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 4705
+    },
+    {
+      "epoch": 1.8801438274071116,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 4706
+    },
+    {
+      "epoch": 1.880543347982421,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 4707
+    },
+    {
+      "epoch": 1.8809428685577307,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 4708
+    },
+    {
+      "epoch": 1.8813423891330403,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 4709
+    },
+    {
+      "epoch": 1.8817419097083499,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 4710
+    },
+    {
+      "epoch": 1.8821414302836597,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4711
+    },
+    {
+      "epoch": 1.8825409508589692,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 4712
+    },
+    {
+      "epoch": 1.8829404714342788,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 4713
+    },
+    {
+      "epoch": 1.8833399920095886,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 4714
+    },
+    {
+      "epoch": 1.8837395125848981,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4715
+    },
+    {
+      "epoch": 1.8841390331602077,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 4716
+    },
+    {
+      "epoch": 1.8845385537355175,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0764,
+      "step": 4717
+    },
+    {
+      "epoch": 1.884938074310827,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 4718
+    },
+    {
+      "epoch": 1.8853375948861366,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 4719
+    },
+    {
+      "epoch": 1.8857371154614464,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 4720
+    },
+    {
+      "epoch": 1.8861366360367557,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4721
+    },
+    {
+      "epoch": 1.8865361566120655,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 4722
+    },
+    {
+      "epoch": 1.8869356771873753,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 4723
+    },
+    {
+      "epoch": 1.8873351977626847,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 4724
+    },
+    {
+      "epoch": 1.8877347183379944,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 4725
+    },
+    {
+      "epoch": 1.888134238913304,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4726
+    },
+    {
+      "epoch": 1.8885337594886136,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4727
+    },
+    {
+      "epoch": 1.8889332800639234,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4728
+    },
+    {
+      "epoch": 1.889332800639233,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4729
+    },
+    {
+      "epoch": 1.8897323212145425,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4730
+    },
+    {
+      "epoch": 1.8901318417898523,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4731
+    },
+    {
+      "epoch": 1.8905313623651618,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4732
+    },
+    {
+      "epoch": 1.8909308829404714,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 4733
+    },
+    {
+      "epoch": 1.8913304035157812,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 4734
+    },
+    {
+      "epoch": 1.8917299240910905,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 4735
+    },
+    {
+      "epoch": 1.8921294446664003,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4736
+    },
+    {
+      "epoch": 1.89252896524171,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 4737
+    },
+    {
+      "epoch": 1.8929284858170194,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0773,
+      "step": 4738
+    },
+    {
+      "epoch": 1.8933280063923292,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4739
+    },
+    {
+      "epoch": 1.8937275269676388,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 4740
+    },
+    {
+      "epoch": 1.8941270475429484,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 4741
+    },
+    {
+      "epoch": 1.8945265681182581,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 4742
+    },
+    {
+      "epoch": 1.8949260886935677,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 4743
+    },
+    {
+      "epoch": 1.8953256092688773,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 4744
+    },
+    {
+      "epoch": 1.895725129844187,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 4745
+    },
+    {
+      "epoch": 1.8961246504194966,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4746
+    },
+    {
+      "epoch": 1.8965241709948062,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4747
+    },
+    {
+      "epoch": 1.896923691570116,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 4748
+    },
+    {
+      "epoch": 1.8973232121454255,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 4749
+    },
+    {
+      "epoch": 1.897722732720735,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4750
+    },
+    {
+      "epoch": 1.898122253296045,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4751
+    },
+    {
+      "epoch": 1.8985217738713542,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 4752
+    },
+    {
+      "epoch": 1.898921294446664,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4753
+    },
+    {
+      "epoch": 1.8993208150219736,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 4754
+    },
+    {
+      "epoch": 1.8997203355972831,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 4755
+    },
+    {
+      "epoch": 1.900119856172593,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 4756
+    },
+    {
+      "epoch": 1.9005193767479025,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4757
+    },
+    {
+      "epoch": 1.900918897323212,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4758
+    },
+    {
+      "epoch": 1.9013184178985219,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 4759
+    },
+    {
+      "epoch": 1.9017179384738314,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 4760
+    },
+    {
+      "epoch": 1.902117459049141,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 4761
+    },
+    {
+      "epoch": 1.9025169796244508,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 4762
+    },
+    {
+      "epoch": 1.9029165001997603,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 4763
+    },
+    {
+      "epoch": 1.90331602077507,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 4764
+    },
+    {
+      "epoch": 1.9037155413503797,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 4765
+    },
+    {
+      "epoch": 1.904115061925689,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 4766
+    },
+    {
+      "epoch": 1.9045145825009988,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 4767
+    },
+    {
+      "epoch": 1.9049141030763086,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 4768
+    },
+    {
+      "epoch": 1.905313623651618,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 4769
+    },
+    {
+      "epoch": 1.9057131442269277,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 4770
+    },
+    {
+      "epoch": 1.9061126648022373,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 4771
+    },
+    {
+      "epoch": 1.9065121853775469,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 4772
+    },
+    {
+      "epoch": 1.9069117059528566,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4773
+    },
+    {
+      "epoch": 1.9073112265281662,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 4774
+    },
+    {
+      "epoch": 1.9077107471034758,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 4775
+    },
+    {
+      "epoch": 1.9081102676787856,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 4776
+    },
+    {
+      "epoch": 1.9085097882540951,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 4777
+    },
+    {
+      "epoch": 1.9089093088294047,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 4778
+    },
+    {
+      "epoch": 1.9093088294047145,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 4779
+    },
+    {
+      "epoch": 1.9097083499800238,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 4780
+    },
+    {
+      "epoch": 1.9101078705553336,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 4781
+    },
+    {
+      "epoch": 1.9105073911306434,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 4782
+    },
+    {
+      "epoch": 1.9109069117059527,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 4783
+    },
+    {
+      "epoch": 1.9113064322812625,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 4784
+    },
+    {
+      "epoch": 1.911705952856572,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 4785
+    },
+    {
+      "epoch": 1.9121054734318816,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 4786
+    },
+    {
+      "epoch": 1.9125049940071914,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 4787
+    },
+    {
+      "epoch": 1.912904514582501,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4788
+    },
+    {
+      "epoch": 1.9133040351578106,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 4789
+    },
+    {
+      "epoch": 1.9137035557331203,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 4790
+    },
+    {
+      "epoch": 1.91410307630843,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 4791
+    },
+    {
+      "epoch": 1.9145025968837395,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 4792
+    },
+    {
+      "epoch": 1.9149021174590493,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 4793
+    },
+    {
+      "epoch": 1.9153016380343588,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 4794
+    },
+    {
+      "epoch": 1.9157011586096684,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 4795
+    },
+    {
+      "epoch": 1.9161006791849782,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0778,
+      "step": 4796
+    },
+    {
+      "epoch": 1.9165001997602875,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4797
+    },
+    {
+      "epoch": 1.9168997203355973,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4798
+    },
+    {
+      "epoch": 1.9172992409109069,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 4799
+    },
+    {
+      "epoch": 1.9176987614862164,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4800
+    },
+    {
+      "epoch": 1.9180982820615262,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 4801
+    },
+    {
+      "epoch": 1.9184978026368358,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 4802
+    },
+    {
+      "epoch": 1.9188973232121453,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 4803
+    },
+    {
+      "epoch": 1.9192968437874551,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 4804
+    },
+    {
+      "epoch": 1.9196963643627647,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 4805
+    },
+    {
+      "epoch": 1.9200958849380743,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 4806
+    },
+    {
+      "epoch": 1.920495405513384,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 4807
+    },
+    {
+      "epoch": 1.9208949260886936,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4808
+    },
+    {
+      "epoch": 1.9212944466640032,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 4809
+    },
+    {
+      "epoch": 1.921693967239313,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 4810
+    },
+    {
+      "epoch": 1.9220934878146223,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 4811
+    },
+    {
+      "epoch": 1.922493008389932,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 4812
+    },
+    {
+      "epoch": 1.9228925289652419,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4813
+    },
+    {
+      "epoch": 1.9232920495405512,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4814
+    },
+    {
+      "epoch": 1.923691570115861,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 4815
+    },
+    {
+      "epoch": 1.9240910906911706,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 4816
+    },
+    {
+      "epoch": 1.9244906112664801,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 4817
+    },
+    {
+      "epoch": 1.92489013184179,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 4818
+    },
+    {
+      "epoch": 1.9252896524170995,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4819
+    },
+    {
+      "epoch": 1.925689172992409,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4820
+    },
+    {
+      "epoch": 1.9260886935677188,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 4821
+    },
+    {
+      "epoch": 1.9264882141430284,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4822
+    },
+    {
+      "epoch": 1.926887734718338,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 4823
+    },
+    {
+      "epoch": 1.9272872552936477,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4824
+    },
+    {
+      "epoch": 1.927686775868957,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 4825
+    },
+    {
+      "epoch": 1.9280862964442669,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 4826
+    },
+    {
+      "epoch": 1.9284858170195767,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 4827
+    },
+    {
+      "epoch": 1.928885337594886,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 4828
+    },
+    {
+      "epoch": 1.9292848581701958,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4829
+    },
+    {
+      "epoch": 1.9296843787455054,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4830
+    },
+    {
+      "epoch": 1.930083899320815,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 4831
+    },
+    {
+      "epoch": 1.9304834198961247,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 4832
+    },
+    {
+      "epoch": 1.9308829404714343,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 4833
+    },
+    {
+      "epoch": 1.9312824610467438,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 4834
+    },
+    {
+      "epoch": 1.9316819816220536,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 4835
+    },
+    {
+      "epoch": 1.9320815021973632,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 4836
+    },
+    {
+      "epoch": 1.9324810227726728,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4837
+    },
+    {
+      "epoch": 1.9328805433479825,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 4838
+    },
+    {
+      "epoch": 1.933280063923292,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 4839
+    },
+    {
+      "epoch": 1.9336795844986017,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 4840
+    },
+    {
+      "epoch": 1.9340791050739115,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 4841
+    },
+    {
+      "epoch": 1.9344786256492208,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 4842
+    },
+    {
+      "epoch": 1.9348781462245306,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4843
+    },
+    {
+      "epoch": 1.9352776667998401,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 4844
+    },
+    {
+      "epoch": 1.9356771873751497,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 4845
+    },
+    {
+      "epoch": 1.9360767079504595,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 4846
+    },
+    {
+      "epoch": 1.936476228525769,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4847
+    },
+    {
+      "epoch": 1.9368757491010786,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 4848
+    },
+    {
+      "epoch": 1.9372752696763884,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 4849
+    },
+    {
+      "epoch": 1.937674790251698,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 4850
+    },
+    {
+      "epoch": 1.9380743108270075,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 4851
+    },
+    {
+      "epoch": 1.9384738314023173,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 4852
+    },
+    {
+      "epoch": 1.938873351977627,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 4853
+    },
+    {
+      "epoch": 1.9392728725529365,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 4854
+    },
+    {
+      "epoch": 1.9396723931282462,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 4855
+    },
+    {
+      "epoch": 1.9400719137035556,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4856
+    },
+    {
+      "epoch": 1.9404714342788654,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 4857
+    },
+    {
+      "epoch": 1.940870954854175,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 4858
+    },
+    {
+      "epoch": 1.9412704754294845,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 4859
+    },
+    {
+      "epoch": 1.9416699960047943,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 4860
+    },
+    {
+      "epoch": 1.9420695165801038,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4861
+    },
+    {
+      "epoch": 1.9424690371554134,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4862
+    },
+    {
+      "epoch": 1.9428685577307232,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 4863
+    },
+    {
+      "epoch": 1.9432680783060328,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 4864
+    },
+    {
+      "epoch": 1.9436675988813423,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 4865
+    },
+    {
+      "epoch": 1.9440671194566521,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4866
+    },
+    {
+      "epoch": 1.9444666400319617,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 4867
+    },
+    {
+      "epoch": 1.9448661606072712,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 4868
+    },
+    {
+      "epoch": 1.945265681182581,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 4869
+    },
+    {
+      "epoch": 1.9456652017578904,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 4870
+    },
+    {
+      "epoch": 1.9460647223332002,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 4871
+    },
+    {
+      "epoch": 1.94646424290851,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 4872
+    },
+    {
+      "epoch": 1.9468637634838193,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4873
+    },
+    {
+      "epoch": 1.947263284059129,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4874
+    },
+    {
+      "epoch": 1.9476628046344386,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 4875
+    },
+    {
+      "epoch": 1.9480623252097482,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 4876
+    },
+    {
+      "epoch": 1.948461845785058,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 4877
+    },
+    {
+      "epoch": 1.9488613663603676,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 4878
+    },
+    {
+      "epoch": 1.9492608869356771,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 4879
+    },
+    {
+      "epoch": 1.949660407510987,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4880
+    },
+    {
+      "epoch": 1.9500599280862965,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 4881
+    },
+    {
+      "epoch": 1.950459448661606,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 4882
+    },
+    {
+      "epoch": 1.9508589692369158,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 4883
+    },
+    {
+      "epoch": 1.9512584898122254,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 4884
+    },
+    {
+      "epoch": 1.951658010387535,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 4885
+    },
+    {
+      "epoch": 1.9520575309628447,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 4886
+    },
+    {
+      "epoch": 1.952457051538154,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 4887
+    },
+    {
+      "epoch": 1.9528565721134639,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 4888
+    },
+    {
+      "epoch": 1.9532560926887734,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 4889
+    },
+    {
+      "epoch": 1.953655613264083,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 4890
+    },
+    {
+      "epoch": 1.9540551338393928,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 4891
+    },
+    {
+      "epoch": 1.9544546544147023,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 4892
+    },
+    {
+      "epoch": 1.954854174990012,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 4893
+    },
+    {
+      "epoch": 1.9552536955653217,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 4894
+    },
+    {
+      "epoch": 1.9556532161406313,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 4895
+    },
+    {
+      "epoch": 1.9560527367159408,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 4896
+    },
+    {
+      "epoch": 1.9564522572912506,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 4897
+    },
+    {
+      "epoch": 1.9568517778665602,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 4898
+    },
+    {
+      "epoch": 1.9572512984418697,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 4899
+    },
+    {
+      "epoch": 1.9576508190171795,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4900
+    },
+    {
+      "epoch": 1.9580503395924889,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 4901
+    },
+    {
+      "epoch": 1.9584498601677987,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 4902
+    },
+    {
+      "epoch": 1.9588493807431082,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 4903
+    },
+    {
+      "epoch": 1.9592489013184178,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 4904
+    },
+    {
+      "epoch": 1.9596484218937276,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 4905
+    },
+    {
+      "epoch": 1.9600479424690371,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 4906
+    },
+    {
+      "epoch": 1.9604474630443467,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 4907
+    },
+    {
+      "epoch": 1.9608469836196565,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 4908
+    },
+    {
+      "epoch": 1.961246504194966,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 4909
+    },
+    {
+      "epoch": 1.9616460247702756,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4910
+    },
+    {
+      "epoch": 1.9620455453455854,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4911
+    },
+    {
+      "epoch": 1.962445065920895,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 4912
+    },
+    {
+      "epoch": 1.9628445864962045,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4913
+    },
+    {
+      "epoch": 1.9632441070715143,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 4914
+    },
+    {
+      "epoch": 1.9636436276468237,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 4915
+    },
+    {
+      "epoch": 1.9640431482221334,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 4916
+    },
+    {
+      "epoch": 1.9644426687974432,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 4917
+    },
+    {
+      "epoch": 1.9648421893727526,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 4918
+    },
+    {
+      "epoch": 1.9652417099480624,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 4919
+    },
+    {
+      "epoch": 1.965641230523372,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 4920
+    },
+    {
+      "epoch": 1.9660407510986815,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4921
+    },
+    {
+      "epoch": 1.9664402716739913,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 4922
+    },
+    {
+      "epoch": 1.9668397922493008,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 4923
+    },
+    {
+      "epoch": 1.9672393128246104,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 4924
+    },
+    {
+      "epoch": 1.9676388333999202,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 4925
+    },
+    {
+      "epoch": 1.9680383539752297,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4926
+    },
+    {
+      "epoch": 1.9684378745505393,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 4927
+    },
+    {
+      "epoch": 1.968837395125849,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 4928
+    },
+    {
+      "epoch": 1.9692369157011587,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 4929
+    },
+    {
+      "epoch": 1.9696364362764682,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 4930
+    },
+    {
+      "epoch": 1.970035956851778,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 4931
+    },
+    {
+      "epoch": 1.9704354774270874,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4932
+    },
+    {
+      "epoch": 1.9708349980023971,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 4933
+    },
+    {
+      "epoch": 1.9712345185777067,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 4934
+    },
+    {
+      "epoch": 1.9716340391530163,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4935
+    },
+    {
+      "epoch": 1.972033559728326,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0756,
+      "step": 4936
+    },
+    {
+      "epoch": 1.9724330803036356,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0731,
+      "step": 4937
+    },
+    {
+      "epoch": 1.9728326008789452,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 4938
+    },
+    {
+      "epoch": 1.973232121454255,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 4939
+    },
+    {
+      "epoch": 1.9736316420295645,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 4940
+    },
+    {
+      "epoch": 1.974031162604874,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 4941
+    },
+    {
+      "epoch": 1.9744306831801839,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 4942
+    },
+    {
+      "epoch": 1.9748302037554935,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 4943
+    },
+    {
+      "epoch": 1.975229724330803,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4944
+    },
+    {
+      "epoch": 1.9756292449061128,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 4945
+    },
+    {
+      "epoch": 1.9760287654814221,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 4946
+    },
+    {
+      "epoch": 1.976428286056732,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 4947
+    },
+    {
+      "epoch": 1.9768278066320415,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 4948
+    },
+    {
+      "epoch": 1.977227327207351,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 4949
+    },
+    {
+      "epoch": 1.9776268477826608,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 4950
+    },
+    {
+      "epoch": 1.9780263683579704,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 4951
+    },
+    {
+      "epoch": 1.97842588893328,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 4952
+    },
+    {
+      "epoch": 1.9788254095085898,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0777,
+      "step": 4953
+    },
+    {
+      "epoch": 1.9792249300838993,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 4954
+    },
+    {
+      "epoch": 1.979624450659209,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 4955
+    },
+    {
+      "epoch": 1.9800239712345187,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 4956
+    },
+    {
+      "epoch": 1.9804234918098282,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 4957
+    },
+    {
+      "epoch": 1.9808230123851378,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 4958
+    },
+    {
+      "epoch": 1.9812225329604476,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 4959
+    },
+    {
+      "epoch": 1.981622053535757,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 4960
+    },
+    {
+      "epoch": 1.9820215741110667,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 4961
+    },
+    {
+      "epoch": 1.9824210946863765,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 4962
+    },
+    {
+      "epoch": 1.9828206152616858,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 4963
+    },
+    {
+      "epoch": 1.9832201358369956,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4964
+    },
+    {
+      "epoch": 1.9836196564123052,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 4965
+    },
+    {
+      "epoch": 1.9840191769876148,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 4966
+    },
+    {
+      "epoch": 1.9844186975629245,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 4967
+    },
+    {
+      "epoch": 1.9848182181382341,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 4968
+    },
+    {
+      "epoch": 1.9852177387135437,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 4969
+    },
+    {
+      "epoch": 1.9856172592888535,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 4970
+    },
+    {
+      "epoch": 1.986016779864163,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 4971
+    },
+    {
+      "epoch": 1.9864163004394726,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 4972
+    },
+    {
+      "epoch": 1.9868158210147824,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 4973
+    },
+    {
+      "epoch": 1.987215341590092,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4974
+    },
+    {
+      "epoch": 1.9876148621654015,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 4975
+    },
+    {
+      "epoch": 1.9880143827407113,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 4976
+    },
+    {
+      "epoch": 1.9884139033160206,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 4977
+    },
+    {
+      "epoch": 1.9888134238913304,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 4978
+    },
+    {
+      "epoch": 1.98921294446664,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 4979
+    },
+    {
+      "epoch": 1.9896124650419496,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 4980
+    },
+    {
+      "epoch": 1.9900119856172593,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 4981
+    },
+    {
+      "epoch": 1.990411506192569,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 4982
+    },
+    {
+      "epoch": 1.9908110267678785,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 4983
+    },
+    {
+      "epoch": 1.9912105473431883,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 4984
+    },
+    {
+      "epoch": 1.9916100679184978,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 4985
+    },
+    {
+      "epoch": 1.9920095884938074,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 4986
+    },
+    {
+      "epoch": 1.9924091090691172,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 4987
+    },
+    {
+      "epoch": 1.9928086296444267,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 4988
+    },
+    {
+      "epoch": 1.9932081502197363,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 4989
+    },
+    {
+      "epoch": 1.993607670795046,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 4990
+    },
+    {
+      "epoch": 1.9940071913703554,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 4991
+    },
+    {
+      "epoch": 1.9944067119456652,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 4992
+    },
+    {
+      "epoch": 1.9948062325209748,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 4993
+    },
+    {
+      "epoch": 1.9952057530962843,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 4994
+    },
+    {
+      "epoch": 1.9956052736715941,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 4995
+    },
+    {
+      "epoch": 1.9960047942469037,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 4996
+    },
+    {
+      "epoch": 1.9964043148222133,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 4997
+    },
+    {
+      "epoch": 1.996803835397523,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 4998
+    },
+    {
+      "epoch": 1.9972033559728326,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 4999
+    },
+    {
+      "epoch": 1.9976028765481422,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 5000
+    },
+    {
+      "epoch": 1.998002397123452,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 5001
+    },
+    {
+      "epoch": 1.9984019176987615,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5002
+    },
+    {
+      "epoch": 1.998801438274071,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 5003
+    },
+    {
+      "epoch": 1.9992009588493809,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5004
+    },
+    {
+      "epoch": 1.9996004794246902,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 5005
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 5006
+    },
+    {
+      "epoch": 2.00039952057531,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 5007
+    },
+    {
+      "epoch": 2.000799041150619,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 5008
+    },
+    {
+      "epoch": 2.001198561725929,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 5009
+    },
+    {
+      "epoch": 2.0015980823012387,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5010
+    },
+    {
+      "epoch": 2.001997602876548,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 5011
+    },
+    {
+      "epoch": 2.002397123451858,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 5012
+    },
+    {
+      "epoch": 2.0027966440271676,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 5013
+    },
+    {
+      "epoch": 2.003196164602477,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0759,
+      "step": 5014
+    },
+    {
+      "epoch": 2.0035956851777867,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 5015
+    },
+    {
+      "epoch": 2.003995205753096,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5016
+    },
+    {
+      "epoch": 2.004394726328406,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5017
+    },
+    {
+      "epoch": 2.0047942469037157,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 5018
+    },
+    {
+      "epoch": 2.005193767479025,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 5019
+    },
+    {
+      "epoch": 2.005593288054335,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5020
+    },
+    {
+      "epoch": 2.0059928086296446,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5021
+    },
+    {
+      "epoch": 2.006392329204954,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 5022
+    },
+    {
+      "epoch": 2.0067918497802637,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5023
+    },
+    {
+      "epoch": 2.0071913703555735,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 5024
+    },
+    {
+      "epoch": 2.007590890930883,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5025
+    },
+    {
+      "epoch": 2.0079904115061926,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 5026
+    },
+    {
+      "epoch": 2.0083899320815024,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5027
+    },
+    {
+      "epoch": 2.0087894526568117,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5028
+    },
+    {
+      "epoch": 2.0091889732321215,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 5029
+    },
+    {
+      "epoch": 2.009588493807431,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5030
+    },
+    {
+      "epoch": 2.0099880143827407,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 5031
+    },
+    {
+      "epoch": 2.0103875349580504,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 5032
+    },
+    {
+      "epoch": 2.01078705553336,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 5033
+    },
+    {
+      "epoch": 2.0111865761086696,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 5034
+    },
+    {
+      "epoch": 2.0115860966839794,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5035
+    },
+    {
+      "epoch": 2.0119856172592887,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 5036
+    },
+    {
+      "epoch": 2.0123851378345985,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 5037
+    },
+    {
+      "epoch": 2.0127846584099083,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5038
+    },
+    {
+      "epoch": 2.0131841789852176,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 5039
+    },
+    {
+      "epoch": 2.0135836995605274,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5040
+    },
+    {
+      "epoch": 2.013983220135837,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 5041
+    },
+    {
+      "epoch": 2.0143827407111465,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 5042
+    },
+    {
+      "epoch": 2.0147822612864563,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.079,
+      "step": 5043
+    },
+    {
+      "epoch": 2.0151817818617657,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5044
+    },
+    {
+      "epoch": 2.0155813024370755,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 5045
+    },
+    {
+      "epoch": 2.0159808230123852,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 5046
+    },
+    {
+      "epoch": 2.0163803435876946,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 5047
+    },
+    {
+      "epoch": 2.0167798641630044,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 5048
+    },
+    {
+      "epoch": 2.017179384738314,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 5049
+    },
+    {
+      "epoch": 2.0175789053136235,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 5050
+    },
+    {
+      "epoch": 2.0179784258889333,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 5051
+    },
+    {
+      "epoch": 2.018377946464243,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 5052
+    },
+    {
+      "epoch": 2.0187774670395524,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5053
+    },
+    {
+      "epoch": 2.019176987614862,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 5054
+    },
+    {
+      "epoch": 2.019576508190172,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 5055
+    },
+    {
+      "epoch": 2.0199760287654813,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 5056
+    },
+    {
+      "epoch": 2.020375549340791,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 5057
+    },
+    {
+      "epoch": 2.020775069916101,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 5058
+    },
+    {
+      "epoch": 2.0211745904914102,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5059
+    },
+    {
+      "epoch": 2.02157411106672,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 5060
+    },
+    {
+      "epoch": 2.0219736316420294,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5061
+    },
+    {
+      "epoch": 2.022373152217339,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 5062
+    },
+    {
+      "epoch": 2.022772672792649,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5063
+    },
+    {
+      "epoch": 2.0231721933679583,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5064
+    },
+    {
+      "epoch": 2.023571713943268,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 5065
+    },
+    {
+      "epoch": 2.023971234518578,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 5066
+    },
+    {
+      "epoch": 2.024370755093887,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5067
+    },
+    {
+      "epoch": 2.024770275669197,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 5068
+    },
+    {
+      "epoch": 2.0251697962445068,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 5069
+    },
+    {
+      "epoch": 2.025569316819816,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 5070
+    },
+    {
+      "epoch": 2.025968837395126,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 5071
+    },
+    {
+      "epoch": 2.0263683579704357,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0774,
+      "step": 5072
+    },
+    {
+      "epoch": 2.026767878545745,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 5073
+    },
+    {
+      "epoch": 2.027167399121055,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 5074
+    },
+    {
+      "epoch": 2.027566919696364,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5075
+    },
+    {
+      "epoch": 2.027966440271674,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5076
+    },
+    {
+      "epoch": 2.0283659608469837,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5077
+    },
+    {
+      "epoch": 2.028765481422293,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 5078
+    },
+    {
+      "epoch": 2.029165001997603,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 5079
+    },
+    {
+      "epoch": 2.0295645225729126,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 5080
+    },
+    {
+      "epoch": 2.029964043148222,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 5081
+    },
+    {
+      "epoch": 2.0303635637235318,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 5082
+    },
+    {
+      "epoch": 2.0307630842988416,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5083
+    },
+    {
+      "epoch": 2.031162604874151,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 5084
+    },
+    {
+      "epoch": 2.0315621254494607,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 5085
+    },
+    {
+      "epoch": 2.0319616460247705,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5086
+    },
+    {
+      "epoch": 2.03236116660008,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 5087
+    },
+    {
+      "epoch": 2.0327606871753896,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5088
+    },
+    {
+      "epoch": 2.033160207750699,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 5089
+    },
+    {
+      "epoch": 2.0335597283260087,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5090
+    },
+    {
+      "epoch": 2.0339592489013185,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5091
+    },
+    {
+      "epoch": 2.034358769476628,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 5092
+    },
+    {
+      "epoch": 2.0347582900519376,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5093
+    },
+    {
+      "epoch": 2.0351578106272474,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 5094
+    },
+    {
+      "epoch": 2.0355573312025568,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 5095
+    },
+    {
+      "epoch": 2.0359568517778666,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 5096
+    },
+    {
+      "epoch": 2.0363563723531763,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 5097
+    },
+    {
+      "epoch": 2.0367558929284857,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 5098
+    },
+    {
+      "epoch": 2.0371554135037955,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 5099
+    },
+    {
+      "epoch": 2.0375549340791053,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 5100
+    },
+    {
+      "epoch": 2.0379544546544146,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 5101
+    },
+    {
+      "epoch": 2.0383539752297244,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 5102
+    },
+    {
+      "epoch": 2.038753495805034,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 5103
+    },
+    {
+      "epoch": 2.0391530163803435,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 5104
+    },
+    {
+      "epoch": 2.0395525369556533,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5105
+    },
+    {
+      "epoch": 2.0399520575309626,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 5106
+    },
+    {
+      "epoch": 2.0403515781062724,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0762,
+      "step": 5107
+    },
+    {
+      "epoch": 2.040751098681582,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 5108
+    },
+    {
+      "epoch": 2.0411506192568916,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 5109
+    },
+    {
+      "epoch": 2.0415501398322013,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 5110
+    },
+    {
+      "epoch": 2.041949660407511,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0738,
+      "step": 5111
+    },
+    {
+      "epoch": 2.0423491809828205,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 5112
+    },
+    {
+      "epoch": 2.0427487015581303,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 5113
+    },
+    {
+      "epoch": 2.04314822213344,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 5114
+    },
+    {
+      "epoch": 2.0435477427087494,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 5115
+    },
+    {
+      "epoch": 2.043947263284059,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5116
+    },
+    {
+      "epoch": 2.044346783859369,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0751,
+      "step": 5117
+    },
+    {
+      "epoch": 2.0447463044346783,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5118
+    },
+    {
+      "epoch": 2.045145825009988,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 5119
+    },
+    {
+      "epoch": 2.0455453455852974,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 5120
+    },
+    {
+      "epoch": 2.0459448661606072,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5121
+    },
+    {
+      "epoch": 2.046344386735917,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 5122
+    },
+    {
+      "epoch": 2.0467439073112264,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 5123
+    },
+    {
+      "epoch": 2.047143427886536,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 5124
+    },
+    {
+      "epoch": 2.047542948461846,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 5125
+    },
+    {
+      "epoch": 2.0479424690371553,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 5126
+    },
+    {
+      "epoch": 2.048341989612465,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 5127
+    },
+    {
+      "epoch": 2.048741510187775,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 5128
+    },
+    {
+      "epoch": 2.049141030763084,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5129
+    },
+    {
+      "epoch": 2.049540551338394,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 5130
+    },
+    {
+      "epoch": 2.0499400719137038,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5131
+    },
+    {
+      "epoch": 2.050339592489013,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 5132
+    },
+    {
+      "epoch": 2.050739113064323,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 5133
+    },
+    {
+      "epoch": 2.0511386336396322,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 5134
+    },
+    {
+      "epoch": 2.051538154214942,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 5135
+    },
+    {
+      "epoch": 2.051937674790252,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 5136
+    },
+    {
+      "epoch": 2.052337195365561,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 5137
+    },
+    {
+      "epoch": 2.052736715940871,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0782,
+      "step": 5138
+    },
+    {
+      "epoch": 2.0531362365161807,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 5139
+    },
+    {
+      "epoch": 2.05353575709149,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 5140
+    },
+    {
+      "epoch": 2.0539352776668,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 5141
+    },
+    {
+      "epoch": 2.0543347982421096,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 5142
+    },
+    {
+      "epoch": 2.054734318817419,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 5143
+    },
+    {
+      "epoch": 2.0551338393927288,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 5144
+    },
+    {
+      "epoch": 2.0555333599680385,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 5145
+    },
+    {
+      "epoch": 2.055932880543348,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 5146
+    },
+    {
+      "epoch": 2.0563324011186577,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 5147
+    },
+    {
+      "epoch": 2.0567319216939675,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5148
+    },
+    {
+      "epoch": 2.057131442269277,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 5149
+    },
+    {
+      "epoch": 2.0575309628445866,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5150
+    },
+    {
+      "epoch": 2.057930483419896,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 5151
+    },
+    {
+      "epoch": 2.0583300039952057,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 5152
+    },
+    {
+      "epoch": 2.0587295245705155,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 5153
+    },
+    {
+      "epoch": 2.059129045145825,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5154
+    },
+    {
+      "epoch": 2.0595285657211346,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 5155
+    },
+    {
+      "epoch": 2.0599280862964444,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5156
+    },
+    {
+      "epoch": 2.0603276068717538,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 5157
+    },
+    {
+      "epoch": 2.0607271274470635,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 5158
+    },
+    {
+      "epoch": 2.0611266480223733,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 5159
+    },
+    {
+      "epoch": 2.0615261685976827,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 5160
+    },
+    {
+      "epoch": 2.0619256891729925,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 5161
+    },
+    {
+      "epoch": 2.0623252097483022,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 5162
+    },
+    {
+      "epoch": 2.0627247303236116,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5163
+    },
+    {
+      "epoch": 2.0631242508989214,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 5164
+    },
+    {
+      "epoch": 2.0635237714742307,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5165
+    },
+    {
+      "epoch": 2.0639232920495405,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 5166
+    },
+    {
+      "epoch": 2.0643228126248503,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 5167
+    },
+    {
+      "epoch": 2.0647223332001596,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 5168
+    },
+    {
+      "epoch": 2.0651218537754694,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5169
+    },
+    {
+      "epoch": 2.065521374350779,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 5170
+    },
+    {
+      "epoch": 2.0659208949260885,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 5171
+    },
+    {
+      "epoch": 2.0663204155013983,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 5172
+    },
+    {
+      "epoch": 2.066719936076708,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 5173
+    },
+    {
+      "epoch": 2.0671194566520175,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 5174
+    },
+    {
+      "epoch": 2.0675189772273272,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 5175
+    },
+    {
+      "epoch": 2.067918497802637,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 5176
+    },
+    {
+      "epoch": 2.0683180183779464,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 5177
+    },
+    {
+      "epoch": 2.068717538953256,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5178
+    },
+    {
+      "epoch": 2.0691170595285655,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5179
+    },
+    {
+      "epoch": 2.0695165801038753,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0847,
+      "step": 5180
+    },
+    {
+      "epoch": 2.069916100679185,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 5181
+    },
+    {
+      "epoch": 2.0703156212544944,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 5182
+    },
+    {
+      "epoch": 2.070715141829804,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 5183
+    },
+    {
+      "epoch": 2.071114662405114,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5184
+    },
+    {
+      "epoch": 2.0715141829804233,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 5185
+    },
+    {
+      "epoch": 2.071913703555733,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5186
+    },
+    {
+      "epoch": 2.072313224131043,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 5187
+    },
+    {
+      "epoch": 2.0727127447063523,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5188
+    },
+    {
+      "epoch": 2.073112265281662,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5189
+    },
+    {
+      "epoch": 2.073511785856972,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0752,
+      "step": 5190
+    },
+    {
+      "epoch": 2.073911306432281,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 5191
+    },
+    {
+      "epoch": 2.074310827007591,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 5192
+    },
+    {
+      "epoch": 2.0747103475829007,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 5193
+    },
+    {
+      "epoch": 2.07510986815821,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5194
+    },
+    {
+      "epoch": 2.07550938873352,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 5195
+    },
+    {
+      "epoch": 2.075908909308829,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 5196
+    },
+    {
+      "epoch": 2.076308429884139,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5197
+    },
+    {
+      "epoch": 2.076707950459449,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 5198
+    },
+    {
+      "epoch": 2.077107471034758,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 5199
+    },
+    {
+      "epoch": 2.077506991610068,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0766,
+      "step": 5200
+    },
+    {
+      "epoch": 2.0779065121853777,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5201
+    },
+    {
+      "epoch": 2.078306032760687,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 5202
+    },
+    {
+      "epoch": 2.078705553335997,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 5203
+    },
+    {
+      "epoch": 2.0791050739113066,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 5204
+    },
+    {
+      "epoch": 2.079504594486616,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 5205
+    },
+    {
+      "epoch": 2.0799041150619257,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 5206
+    },
+    {
+      "epoch": 2.0803036356372355,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 5207
+    },
+    {
+      "epoch": 2.080703156212545,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5208
+    },
+    {
+      "epoch": 2.0811026767878547,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 5209
+    },
+    {
+      "epoch": 2.081502197363164,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 5210
+    },
+    {
+      "epoch": 2.081901717938474,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 5211
+    },
+    {
+      "epoch": 2.0823012385137836,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 5212
+    },
+    {
+      "epoch": 2.082700759089093,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 5213
+    },
+    {
+      "epoch": 2.0831002796644027,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 5214
+    },
+    {
+      "epoch": 2.0834998002397125,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 5215
+    },
+    {
+      "epoch": 2.083899320815022,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 5216
+    },
+    {
+      "epoch": 2.0842988413903316,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 5217
+    },
+    {
+      "epoch": 2.0846983619656414,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 5218
+    },
+    {
+      "epoch": 2.0850978825409507,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5219
+    },
+    {
+      "epoch": 2.0854974031162605,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 5220
+    },
+    {
+      "epoch": 2.0858969236915703,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5221
+    },
+    {
+      "epoch": 2.0862964442668797,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.078,
+      "step": 5222
+    },
+    {
+      "epoch": 2.0866959648421894,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 5223
+    },
+    {
+      "epoch": 2.087095485417499,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 5224
+    },
+    {
+      "epoch": 2.0874950059928086,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5225
+    },
+    {
+      "epoch": 2.0878945265681184,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 5226
+    },
+    {
+      "epoch": 2.0882940471434277,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 5227
+    },
+    {
+      "epoch": 2.0886935677187375,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 5228
+    },
+    {
+      "epoch": 2.0890930882940473,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 5229
+    },
+    {
+      "epoch": 2.0894926088693566,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 5230
+    },
+    {
+      "epoch": 2.0898921294446664,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5231
+    },
+    {
+      "epoch": 2.090291650019976,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 5232
+    },
+    {
+      "epoch": 2.0906911705952855,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 5233
+    },
+    {
+      "epoch": 2.0910906911705953,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 5234
+    },
+    {
+      "epoch": 2.091490211745905,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 5235
+    },
+    {
+      "epoch": 2.0918897323212144,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 5236
+    },
+    {
+      "epoch": 2.0922892528965242,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 5237
+    },
+    {
+      "epoch": 2.092688773471834,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 5238
+    },
+    {
+      "epoch": 2.0930882940471434,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5239
+    },
+    {
+      "epoch": 2.093487814622453,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 5240
+    },
+    {
+      "epoch": 2.0938873351977625,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 5241
+    },
+    {
+      "epoch": 2.0942868557730723,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5242
+    },
+    {
+      "epoch": 2.094686376348382,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5243
+    },
+    {
+      "epoch": 2.0950858969236914,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 5244
+    },
+    {
+      "epoch": 2.095485417499001,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 5245
+    },
+    {
+      "epoch": 2.095884938074311,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 5246
+    },
+    {
+      "epoch": 2.0962844586496203,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 5247
+    },
+    {
+      "epoch": 2.09668397922493,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0767,
+      "step": 5248
+    },
+    {
+      "epoch": 2.09708349980024,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 5249
+    },
+    {
+      "epoch": 2.0974830203755492,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 5250
+    },
+    {
+      "epoch": 2.097882540950859,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5251
+    },
+    {
+      "epoch": 2.098282061526169,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 5252
+    },
+    {
+      "epoch": 2.098681582101478,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 5253
+    },
+    {
+      "epoch": 2.099081102676788,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5254
+    },
+    {
+      "epoch": 2.0994806232520973,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 5255
+    },
+    {
+      "epoch": 2.099880143827407,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 5256
+    },
+    {
+      "epoch": 2.100279664402717,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 5257
+    },
+    {
+      "epoch": 2.100679184978026,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 5258
+    },
+    {
+      "epoch": 2.101078705553336,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 5259
+    },
+    {
+      "epoch": 2.1014782261286458,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 5260
+    },
+    {
+      "epoch": 2.101877746703955,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 5261
+    },
+    {
+      "epoch": 2.102277267279265,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0771,
+      "step": 5262
+    },
+    {
+      "epoch": 2.1026767878545747,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5263
+    },
+    {
+      "epoch": 2.103076308429884,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 5264
+    },
+    {
+      "epoch": 2.103475829005194,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 5265
+    },
+    {
+      "epoch": 2.1038753495805036,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 5266
+    },
+    {
+      "epoch": 2.104274870155813,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 5267
+    },
+    {
+      "epoch": 2.1046743907311227,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 5268
+    },
+    {
+      "epoch": 2.105073911306432,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 5269
+    },
+    {
+      "epoch": 2.105473431881742,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 5270
+    },
+    {
+      "epoch": 2.1058729524570516,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 5271
+    },
+    {
+      "epoch": 2.106272473032361,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 5272
+    },
+    {
+      "epoch": 2.1066719936076708,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5273
+    },
+    {
+      "epoch": 2.1070715141829806,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 5274
+    },
+    {
+      "epoch": 2.10747103475829,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 5275
+    },
+    {
+      "epoch": 2.1078705553335997,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5276
+    },
+    {
+      "epoch": 2.1082700759089095,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 5277
+    },
+    {
+      "epoch": 2.108669596484219,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5278
+    },
+    {
+      "epoch": 2.1090691170595286,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 5279
+    },
+    {
+      "epoch": 2.1094686376348384,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 5280
+    },
+    {
+      "epoch": 2.1098681582101477,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 5281
+    },
+    {
+      "epoch": 2.1102676787854575,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 5282
+    },
+    {
+      "epoch": 2.1106671993607673,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 5283
+    },
+    {
+      "epoch": 2.1110667199360766,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0716,
+      "step": 5284
+    },
+    {
+      "epoch": 2.1114662405113864,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 5285
+    },
+    {
+      "epoch": 2.1118657610866958,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 5286
+    },
+    {
+      "epoch": 2.1122652816620056,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5287
+    },
+    {
+      "epoch": 2.1126648022373153,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 5288
+    },
+    {
+      "epoch": 2.1130643228126247,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 5289
+    },
+    {
+      "epoch": 2.1134638433879345,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5290
+    },
+    {
+      "epoch": 2.1138633639632443,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 5291
+    },
+    {
+      "epoch": 2.1142628845385536,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 5292
+    },
+    {
+      "epoch": 2.1146624051138634,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 5293
+    },
+    {
+      "epoch": 2.115061925689173,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 5294
+    },
+    {
+      "epoch": 2.1154614462644825,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 5295
+    },
+    {
+      "epoch": 2.1158609668397923,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 5296
+    },
+    {
+      "epoch": 2.116260487415102,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0406,
+      "step": 5297
+    },
+    {
+      "epoch": 2.1166600079904114,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 5298
+    },
+    {
+      "epoch": 2.117059528565721,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 5299
+    },
+    {
+      "epoch": 2.1174590491410306,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 5300
+    },
+    {
+      "epoch": 2.1178585697163403,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5301
+    },
+    {
+      "epoch": 2.11825809029165,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 5302
+    },
+    {
+      "epoch": 2.1186576108669595,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 5303
+    },
+    {
+      "epoch": 2.1190571314422693,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 5304
+    },
+    {
+      "epoch": 2.119456652017579,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.076,
+      "step": 5305
+    },
+    {
+      "epoch": 2.1198561725928884,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 5306
+    },
+    {
+      "epoch": 2.120255693168198,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5307
+    },
+    {
+      "epoch": 2.120655213743508,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5308
+    },
+    {
+      "epoch": 2.1210547343188173,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 5309
+    },
+    {
+      "epoch": 2.121454254894127,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5310
+    },
+    {
+      "epoch": 2.121853775469437,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 5311
+    },
+    {
+      "epoch": 2.122253296044746,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 5312
+    },
+    {
+      "epoch": 2.122652816620056,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 5313
+    },
+    {
+      "epoch": 2.1230523371953653,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 5314
+    },
+    {
+      "epoch": 2.123451857770675,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 5315
+    },
+    {
+      "epoch": 2.123851378345985,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 5316
+    },
+    {
+      "epoch": 2.1242508989212943,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 5317
+    },
+    {
+      "epoch": 2.124650419496604,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5318
+    },
+    {
+      "epoch": 2.125049940071914,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5319
+    },
+    {
+      "epoch": 2.125449460647223,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 5320
+    },
+    {
+      "epoch": 2.125848981222533,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5321
+    },
+    {
+      "epoch": 2.1262485017978427,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 5322
+    },
+    {
+      "epoch": 2.126648022373152,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 5323
+    },
+    {
+      "epoch": 2.127047542948462,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 5324
+    },
+    {
+      "epoch": 2.1274470635237717,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5325
+    },
+    {
+      "epoch": 2.127846584099081,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 5326
+    },
+    {
+      "epoch": 2.128246104674391,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 5327
+    },
+    {
+      "epoch": 2.1286456252497006,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 5328
+    },
+    {
+      "epoch": 2.12904514582501,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 5329
+    },
+    {
+      "epoch": 2.1294446664003197,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 5330
+    },
+    {
+      "epoch": 2.129844186975629,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 5331
+    },
+    {
+      "epoch": 2.130243707550939,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 5332
+    },
+    {
+      "epoch": 2.1306432281262486,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5333
+    },
+    {
+      "epoch": 2.131042748701558,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 5334
+    },
+    {
+      "epoch": 2.1314422692768678,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 5335
+    },
+    {
+      "epoch": 2.1318417898521775,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 5336
+    },
+    {
+      "epoch": 2.132241310427487,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 5337
+    },
+    {
+      "epoch": 2.1326408310027967,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 5338
+    },
+    {
+      "epoch": 2.1330403515781065,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 5339
+    },
+    {
+      "epoch": 2.133439872153416,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 5340
+    },
+    {
+      "epoch": 2.1338393927287256,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5341
+    },
+    {
+      "epoch": 2.134238913304035,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5342
+    },
+    {
+      "epoch": 2.1346384338793447,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0783,
+      "step": 5343
+    },
+    {
+      "epoch": 2.1350379544546545,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 5344
+    },
+    {
+      "epoch": 2.135437475029964,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 5345
+    },
+    {
+      "epoch": 2.1358369956052736,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 5346
+    },
+    {
+      "epoch": 2.1362365161805834,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 5347
+    },
+    {
+      "epoch": 2.1366360367558928,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 5348
+    },
+    {
+      "epoch": 2.1370355573312025,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5349
+    },
+    {
+      "epoch": 2.1374350779065123,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 5350
+    },
+    {
+      "epoch": 2.1378345984818217,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 5351
+    },
+    {
+      "epoch": 2.1382341190571315,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5352
+    },
+    {
+      "epoch": 2.1386336396324412,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5353
+    },
+    {
+      "epoch": 2.1390331602077506,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0745,
+      "step": 5354
+    },
+    {
+      "epoch": 2.1394326807830604,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 5355
+    },
+    {
+      "epoch": 2.13983220135837,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5356
+    },
+    {
+      "epoch": 2.1402317219336795,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 5357
+    },
+    {
+      "epoch": 2.1406312425089893,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5358
+    },
+    {
+      "epoch": 2.1410307630842986,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5359
+    },
+    {
+      "epoch": 2.1414302836596084,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 5360
+    },
+    {
+      "epoch": 2.141829804234918,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 5361
+    },
+    {
+      "epoch": 2.1422293248102275,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 5362
+    },
+    {
+      "epoch": 2.1426288453855373,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 5363
+    },
+    {
+      "epoch": 2.143028365960847,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 5364
+    },
+    {
+      "epoch": 2.1434278865361565,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 5365
+    },
+    {
+      "epoch": 2.1438274071114662,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5366
+    },
+    {
+      "epoch": 2.144226927686776,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 5367
+    },
+    {
+      "epoch": 2.1446264482620854,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 5368
+    },
+    {
+      "epoch": 2.145025968837395,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 5369
+    },
+    {
+      "epoch": 2.145425489412705,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5370
+    },
+    {
+      "epoch": 2.1458250099880143,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5371
+    },
+    {
+      "epoch": 2.146224530563324,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 5372
+    },
+    {
+      "epoch": 2.146624051138634,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 5373
+    },
+    {
+      "epoch": 2.147023571713943,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 5374
+    },
+    {
+      "epoch": 2.147423092289253,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 5375
+    },
+    {
+      "epoch": 2.1478226128645623,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.074,
+      "step": 5376
+    },
+    {
+      "epoch": 2.148222133439872,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 5377
+    },
+    {
+      "epoch": 2.148621654015182,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 5378
+    },
+    {
+      "epoch": 2.1490211745904912,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5379
+    },
+    {
+      "epoch": 2.149420695165801,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 5380
+    },
+    {
+      "epoch": 2.149820215741111,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5381
+    },
+    {
+      "epoch": 2.15021973631642,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 5382
+    },
+    {
+      "epoch": 2.15061925689173,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 5383
+    },
+    {
+      "epoch": 2.1510187774670397,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 5384
+    },
+    {
+      "epoch": 2.151418298042349,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 5385
+    },
+    {
+      "epoch": 2.151817818617659,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5386
+    },
+    {
+      "epoch": 2.152217339192968,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 5387
+    },
+    {
+      "epoch": 2.152616859768278,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 5388
+    },
+    {
+      "epoch": 2.1530163803435878,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 5389
+    },
+    {
+      "epoch": 2.153415900918897,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 5390
+    },
+    {
+      "epoch": 2.153815421494207,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5391
+    },
+    {
+      "epoch": 2.1542149420695167,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 5392
+    },
+    {
+      "epoch": 2.154614462644826,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 5393
+    },
+    {
+      "epoch": 2.155013983220136,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5394
+    },
+    {
+      "epoch": 2.1554135037954456,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5395
+    },
+    {
+      "epoch": 2.155813024370755,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5396
+    },
+    {
+      "epoch": 2.1562125449460647,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5397
+    },
+    {
+      "epoch": 2.1566120655213745,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 5398
+    },
+    {
+      "epoch": 2.157011586096684,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5399
+    },
+    {
+      "epoch": 2.1574111066719937,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5400
+    },
+    {
+      "epoch": 2.1578106272473034,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 5401
+    },
+    {
+      "epoch": 2.158210147822613,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 5402
+    },
+    {
+      "epoch": 2.1586096683979226,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5403
+    },
+    {
+      "epoch": 2.159009188973232,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 5404
+    },
+    {
+      "epoch": 2.1594087095485417,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5405
+    },
+    {
+      "epoch": 2.1598082301238515,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 5406
+    },
+    {
+      "epoch": 2.160207750699161,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 5407
+    },
+    {
+      "epoch": 2.1606072712744706,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5408
+    },
+    {
+      "epoch": 2.1610067918497804,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 5409
+    },
+    {
+      "epoch": 2.1614063124250897,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 5410
+    },
+    {
+      "epoch": 2.1618058330003995,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5411
+    },
+    {
+      "epoch": 2.1622053535757093,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5412
+    },
+    {
+      "epoch": 2.1626048741510187,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 5413
+    },
+    {
+      "epoch": 2.1630043947263284,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5414
+    },
+    {
+      "epoch": 2.1634039153016382,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5415
+    },
+    {
+      "epoch": 2.1638034358769476,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 5416
+    },
+    {
+      "epoch": 2.1642029564522574,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5417
+    },
+    {
+      "epoch": 2.164602477027567,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5418
+    },
+    {
+      "epoch": 2.1650019976028765,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5419
+    },
+    {
+      "epoch": 2.1654015181781863,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 5420
+    },
+    {
+      "epoch": 2.1658010387534956,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 5421
+    },
+    {
+      "epoch": 2.1662005593288054,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 5422
+    },
+    {
+      "epoch": 2.166600079904115,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 5423
+    },
+    {
+      "epoch": 2.1669996004794245,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 5424
+    },
+    {
+      "epoch": 2.1673991210547343,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 5425
+    },
+    {
+      "epoch": 2.167798641630044,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 5426
+    },
+    {
+      "epoch": 2.1681981622053534,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5427
+    },
+    {
+      "epoch": 2.1685976827806632,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 5428
+    },
+    {
+      "epoch": 2.168997203355973,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 5429
+    },
+    {
+      "epoch": 2.1693967239312824,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 5430
+    },
+    {
+      "epoch": 2.169796244506592,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0743,
+      "step": 5431
+    },
+    {
+      "epoch": 2.1701957650819015,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 5432
+    },
+    {
+      "epoch": 2.1705952856572113,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0719,
+      "step": 5433
+    },
+    {
+      "epoch": 2.170994806232521,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 5434
+    },
+    {
+      "epoch": 2.1713943268078304,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5435
+    },
+    {
+      "epoch": 2.17179384738314,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 5436
+    },
+    {
+      "epoch": 2.17219336795845,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5437
+    },
+    {
+      "epoch": 2.1725928885337593,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 5438
+    },
+    {
+      "epoch": 2.172992409109069,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 5439
+    },
+    {
+      "epoch": 2.173391929684379,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 5440
+    },
+    {
+      "epoch": 2.1737914502596882,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 5441
+    },
+    {
+      "epoch": 2.174190970834998,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5442
+    },
+    {
+      "epoch": 2.174590491410308,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 5443
+    },
+    {
+      "epoch": 2.174990011985617,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 5444
+    },
+    {
+      "epoch": 2.175389532560927,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 5445
+    },
+    {
+      "epoch": 2.1757890531362367,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 5446
+    },
+    {
+      "epoch": 2.176188573711546,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 5447
+    },
+    {
+      "epoch": 2.176588094286856,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 5448
+    },
+    {
+      "epoch": 2.176987614862165,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 5449
+    },
+    {
+      "epoch": 2.177387135437475,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5450
+    },
+    {
+      "epoch": 2.1777866560127848,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 5451
+    },
+    {
+      "epoch": 2.178186176588094,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5452
+    },
+    {
+      "epoch": 2.178585697163404,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 5453
+    },
+    {
+      "epoch": 2.1789852177387137,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 5454
+    },
+    {
+      "epoch": 2.179384738314023,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0741,
+      "step": 5455
+    },
+    {
+      "epoch": 2.179784258889333,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 5456
+    },
+    {
+      "epoch": 2.1801837794646426,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 5457
+    },
+    {
+      "epoch": 2.180583300039952,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5458
+    },
+    {
+      "epoch": 2.1809828206152617,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 5459
+    },
+    {
+      "epoch": 2.1813823411905715,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 5460
+    },
+    {
+      "epoch": 2.181781861765881,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5461
+    },
+    {
+      "epoch": 2.1821813823411906,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 5462
+    },
+    {
+      "epoch": 2.1825809029165004,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5463
+    },
+    {
+      "epoch": 2.1829804234918098,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 5464
+    },
+    {
+      "epoch": 2.1833799440671195,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 5465
+    },
+    {
+      "epoch": 2.183779464642429,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 5466
+    },
+    {
+      "epoch": 2.1841789852177387,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 5467
+    },
+    {
+      "epoch": 2.1845785057930485,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5468
+    },
+    {
+      "epoch": 2.184978026368358,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 5469
+    },
+    {
+      "epoch": 2.1853775469436676,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 5470
+    },
+    {
+      "epoch": 2.1857770675189774,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 5471
+    },
+    {
+      "epoch": 2.1861765880942867,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 5472
+    },
+    {
+      "epoch": 2.1865761086695965,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 5473
+    },
+    {
+      "epoch": 2.1869756292449063,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 5474
+    },
+    {
+      "epoch": 2.1873751498202156,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 5475
+    },
+    {
+      "epoch": 2.1877746703955254,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0755,
+      "step": 5476
+    },
+    {
+      "epoch": 2.1881741909708348,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 5477
+    },
+    {
+      "epoch": 2.1885737115461446,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5478
+    },
+    {
+      "epoch": 2.1889732321214543,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5479
+    },
+    {
+      "epoch": 2.1893727526967637,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 5480
+    },
+    {
+      "epoch": 2.1897722732720735,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0703,
+      "step": 5481
+    },
+    {
+      "epoch": 2.1901717938473833,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5482
+    },
+    {
+      "epoch": 2.1905713144226926,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 5483
+    },
+    {
+      "epoch": 2.1909708349980024,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 5484
+    },
+    {
+      "epoch": 2.191370355573312,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 5485
+    },
+    {
+      "epoch": 2.1917698761486215,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 5486
+    },
+    {
+      "epoch": 2.1921693967239313,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0445,
+      "step": 5487
+    },
+    {
+      "epoch": 2.192568917299241,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 5488
+    },
+    {
+      "epoch": 2.1929684378745504,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 5489
+    },
+    {
+      "epoch": 2.19336795844986,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 5490
+    },
+    {
+      "epoch": 2.19376747902517,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 5491
+    },
+    {
+      "epoch": 2.1941669996004793,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 5492
+    },
+    {
+      "epoch": 2.194566520175789,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5493
+    },
+    {
+      "epoch": 2.1949660407510985,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5494
+    },
+    {
+      "epoch": 2.1953655613264083,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 5495
+    },
+    {
+      "epoch": 2.195765081901718,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 5496
+    },
+    {
+      "epoch": 2.1961646024770274,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 5497
+    },
+    {
+      "epoch": 2.196564123052337,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 5498
+    },
+    {
+      "epoch": 2.196963643627647,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5499
+    },
+    {
+      "epoch": 2.1973631642029563,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 5500
+    },
+    {
+      "epoch": 2.197762684778266,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 5501
+    },
+    {
+      "epoch": 2.198162205353576,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 5502
+    },
+    {
+      "epoch": 2.198561725928885,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 5503
+    },
+    {
+      "epoch": 2.198961246504195,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5504
+    },
+    {
+      "epoch": 2.199360767079505,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 5505
+    },
+    {
+      "epoch": 2.199760287654814,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 5506
+    },
+    {
+      "epoch": 2.200159808230124,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 5507
+    },
+    {
+      "epoch": 2.2005593288054337,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 5508
+    },
+    {
+      "epoch": 2.200958849380743,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5509
+    },
+    {
+      "epoch": 2.201358369956053,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 5510
+    },
+    {
+      "epoch": 2.201757890531362,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 5511
+    },
+    {
+      "epoch": 2.202157411106672,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 5512
+    },
+    {
+      "epoch": 2.2025569316819817,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 5513
+    },
+    {
+      "epoch": 2.202956452257291,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 5514
+    },
+    {
+      "epoch": 2.203355972832601,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 5515
+    },
+    {
+      "epoch": 2.2037554934079107,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5516
+    },
+    {
+      "epoch": 2.20415501398322,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 5517
+    },
+    {
+      "epoch": 2.20455453455853,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 5518
+    },
+    {
+      "epoch": 2.2049540551338396,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5519
+    },
+    {
+      "epoch": 2.205353575709149,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 5520
+    },
+    {
+      "epoch": 2.2057530962844587,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 5521
+    },
+    {
+      "epoch": 2.206152616859768,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 5522
+    },
+    {
+      "epoch": 2.206552137435078,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 5523
+    },
+    {
+      "epoch": 2.2069516580103876,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5524
+    },
+    {
+      "epoch": 2.207351178585697,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5525
+    },
+    {
+      "epoch": 2.2077506991610067,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 5526
+    },
+    {
+      "epoch": 2.2081502197363165,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 5527
+    },
+    {
+      "epoch": 2.208549740311626,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 5528
+    },
+    {
+      "epoch": 2.2089492608869357,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 5529
+    },
+    {
+      "epoch": 2.2093487814622454,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 5530
+    },
+    {
+      "epoch": 2.209748302037555,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 5531
+    },
+    {
+      "epoch": 2.2101478226128646,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5532
+    },
+    {
+      "epoch": 2.2105473431881744,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 5533
+    },
+    {
+      "epoch": 2.2109468637634837,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 5534
+    },
+    {
+      "epoch": 2.2113463843387935,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 5535
+    },
+    {
+      "epoch": 2.2117459049141033,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 5536
+    },
+    {
+      "epoch": 2.2121454254894126,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 5537
+    },
+    {
+      "epoch": 2.2125449460647224,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 5538
+    },
+    {
+      "epoch": 2.2129444666400317,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5539
+    },
+    {
+      "epoch": 2.2133439872153415,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 5540
+    },
+    {
+      "epoch": 2.2137435077906513,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 5541
+    },
+    {
+      "epoch": 2.2141430283659607,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 5542
+    },
+    {
+      "epoch": 2.2145425489412705,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 5543
+    },
+    {
+      "epoch": 2.2149420695165802,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 5544
+    },
+    {
+      "epoch": 2.2153415900918896,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 5545
+    },
+    {
+      "epoch": 2.2157411106671994,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5546
+    },
+    {
+      "epoch": 2.216140631242509,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 5547
+    },
+    {
+      "epoch": 2.2165401518178185,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5548
+    },
+    {
+      "epoch": 2.2169396723931283,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5549
+    },
+    {
+      "epoch": 2.217339192968438,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5550
+    },
+    {
+      "epoch": 2.2177387135437474,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 5551
+    },
+    {
+      "epoch": 2.218138234119057,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 5552
+    },
+    {
+      "epoch": 2.218537754694367,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 5553
+    },
+    {
+      "epoch": 2.2189372752696763,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0758,
+      "step": 5554
+    },
+    {
+      "epoch": 2.219336795844986,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 5555
+    },
+    {
+      "epoch": 2.2197363164202955,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5556
+    },
+    {
+      "epoch": 2.2201358369956052,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5557
+    },
+    {
+      "epoch": 2.220535357570915,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 5558
+    },
+    {
+      "epoch": 2.2209348781462244,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 5559
+    },
+    {
+      "epoch": 2.221334398721534,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 5560
+    },
+    {
+      "epoch": 2.221733919296844,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 5561
+    },
+    {
+      "epoch": 2.2221334398721533,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 5562
+    },
+    {
+      "epoch": 2.222532960447463,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 5563
+    },
+    {
+      "epoch": 2.222932481022773,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 5564
+    },
+    {
+      "epoch": 2.223332001598082,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5565
+    },
+    {
+      "epoch": 2.223731522173392,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 5566
+    },
+    {
+      "epoch": 2.2241310427487013,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 5567
+    },
+    {
+      "epoch": 2.224530563324011,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 5568
+    },
+    {
+      "epoch": 2.224930083899321,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 5569
+    },
+    {
+      "epoch": 2.2253296044746302,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 5570
+    },
+    {
+      "epoch": 2.22572912504994,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 5571
+    },
+    {
+      "epoch": 2.22612864562525,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5572
+    },
+    {
+      "epoch": 2.226528166200559,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 5573
+    },
+    {
+      "epoch": 2.226927686775869,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 5574
+    },
+    {
+      "epoch": 2.2273272073511787,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 5575
+    },
+    {
+      "epoch": 2.227726727926488,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 5576
+    },
+    {
+      "epoch": 2.228126248501798,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 5577
+    },
+    {
+      "epoch": 2.2285257690771076,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5578
+    },
+    {
+      "epoch": 2.228925289652417,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 5579
+    },
+    {
+      "epoch": 2.2293248102277268,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5580
+    },
+    {
+      "epoch": 2.2297243308030366,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5581
+    },
+    {
+      "epoch": 2.230123851378346,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5582
+    },
+    {
+      "epoch": 2.2305233719536557,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 5583
+    },
+    {
+      "epoch": 2.230922892528965,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 5584
+    },
+    {
+      "epoch": 2.231322413104275,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 5585
+    },
+    {
+      "epoch": 2.2317219336795846,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 5586
+    },
+    {
+      "epoch": 2.232121454254894,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 5587
+    },
+    {
+      "epoch": 2.2325209748302037,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 5588
+    },
+    {
+      "epoch": 2.2329204954055135,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 5589
+    },
+    {
+      "epoch": 2.233320015980823,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 5590
+    },
+    {
+      "epoch": 2.2337195365561326,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 5591
+    },
+    {
+      "epoch": 2.2341190571314424,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 5592
+    },
+    {
+      "epoch": 2.2345185777067518,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0805,
+      "step": 5593
+    },
+    {
+      "epoch": 2.2349180982820616,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 5594
+    },
+    {
+      "epoch": 2.2353176188573713,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 5595
+    },
+    {
+      "epoch": 2.2357171394326807,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5596
+    },
+    {
+      "epoch": 2.2361166600079905,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 5597
+    },
+    {
+      "epoch": 2.2365161805833003,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 5598
+    },
+    {
+      "epoch": 2.2369157011586096,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5599
+    },
+    {
+      "epoch": 2.2373152217339194,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 5600
+    },
+    {
+      "epoch": 2.2377147423092287,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5601
+    },
+    {
+      "epoch": 2.2381142628845385,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 5602
+    },
+    {
+      "epoch": 2.2385137834598483,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 5603
+    },
+    {
+      "epoch": 2.2389133040351576,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5604
+    },
+    {
+      "epoch": 2.2393128246104674,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 5605
+    },
+    {
+      "epoch": 2.239712345185777,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 5606
+    },
+    {
+      "epoch": 2.2401118657610866,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 5607
+    },
+    {
+      "epoch": 2.2405113863363963,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 5608
+    },
+    {
+      "epoch": 2.240910906911706,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 5609
+    },
+    {
+      "epoch": 2.2413104274870155,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 5610
+    },
+    {
+      "epoch": 2.2417099480623253,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 5611
+    },
+    {
+      "epoch": 2.2421094686376346,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 5612
+    },
+    {
+      "epoch": 2.2425089892129444,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5613
+    },
+    {
+      "epoch": 2.242908509788254,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 5614
+    },
+    {
+      "epoch": 2.2433080303635635,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 5615
+    },
+    {
+      "epoch": 2.2437075509388733,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5616
+    },
+    {
+      "epoch": 2.244107071514183,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 5617
+    },
+    {
+      "epoch": 2.2445065920894924,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 5618
+    },
+    {
+      "epoch": 2.2449061126648022,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 5619
+    },
+    {
+      "epoch": 2.245305633240112,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 5620
+    },
+    {
+      "epoch": 2.2457051538154214,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 5621
+    },
+    {
+      "epoch": 2.246104674390731,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 5622
+    },
+    {
+      "epoch": 2.246504194966041,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 5623
+    },
+    {
+      "epoch": 2.2469037155413503,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 5624
+    },
+    {
+      "epoch": 2.24730323611666,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 5625
+    },
+    {
+      "epoch": 2.24770275669197,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 5626
+    },
+    {
+      "epoch": 2.248102277267279,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 5627
+    },
+    {
+      "epoch": 2.248501797842589,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5628
+    },
+    {
+      "epoch": 2.2489013184178983,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 5629
+    },
+    {
+      "epoch": 2.249300838993208,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5630
+    },
+    {
+      "epoch": 2.249700359568518,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 5631
+    },
+    {
+      "epoch": 2.2500998801438272,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 5632
+    },
+    {
+      "epoch": 2.250499400719137,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5633
+    },
+    {
+      "epoch": 2.250898921294447,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 5634
+    },
+    {
+      "epoch": 2.251298441869756,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 5635
+    },
+    {
+      "epoch": 2.251697962445066,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 5636
+    },
+    {
+      "epoch": 2.2520974830203757,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5637
+    },
+    {
+      "epoch": 2.252497003595685,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 5638
+    },
+    {
+      "epoch": 2.252896524170995,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 5639
+    },
+    {
+      "epoch": 2.2532960447463046,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5640
+    },
+    {
+      "epoch": 2.253695565321614,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5641
+    },
+    {
+      "epoch": 2.2540950858969238,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 5642
+    },
+    {
+      "epoch": 2.2544946064722335,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 5643
+    },
+    {
+      "epoch": 2.254894127047543,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 5644
+    },
+    {
+      "epoch": 2.2552936476228527,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 5645
+    },
+    {
+      "epoch": 2.255693168198162,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5646
+    },
+    {
+      "epoch": 2.256092688773472,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5647
+    },
+    {
+      "epoch": 2.2564922093487816,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 5648
+    },
+    {
+      "epoch": 2.256891729924091,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 5649
+    },
+    {
+      "epoch": 2.2572912504994007,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 5650
+    },
+    {
+      "epoch": 2.2576907710747105,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0753,
+      "step": 5651
+    },
+    {
+      "epoch": 2.25809029165002,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 5652
+    },
+    {
+      "epoch": 2.2584898122253296,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0768,
+      "step": 5653
+    },
+    {
+      "epoch": 2.2588893328006394,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5654
+    },
+    {
+      "epoch": 2.2592888533759488,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 5655
+    },
+    {
+      "epoch": 2.2596883739512585,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 5656
+    },
+    {
+      "epoch": 2.260087894526568,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 5657
+    },
+    {
+      "epoch": 2.2604874151018777,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5658
+    },
+    {
+      "epoch": 2.2608869356771875,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 5659
+    },
+    {
+      "epoch": 2.2612864562524972,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 5660
+    },
+    {
+      "epoch": 2.2616859768278066,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 5661
+    },
+    {
+      "epoch": 2.2620854974031164,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5662
+    },
+    {
+      "epoch": 2.2624850179784257,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0781,
+      "step": 5663
+    },
+    {
+      "epoch": 2.2628845385537355,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 5664
+    },
+    {
+      "epoch": 2.2632840591290453,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 5665
+    },
+    {
+      "epoch": 2.2636835797043546,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5666
+    },
+    {
+      "epoch": 2.2640831002796644,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 5667
+    },
+    {
+      "epoch": 2.264482620854974,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5668
+    },
+    {
+      "epoch": 2.2648821414302835,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 5669
+    },
+    {
+      "epoch": 2.2652816620055933,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 5670
+    },
+    {
+      "epoch": 2.265681182580903,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5671
+    },
+    {
+      "epoch": 2.2660807031562125,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 5672
+    },
+    {
+      "epoch": 2.2664802237315222,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 5673
+    },
+    {
+      "epoch": 2.2668797443068316,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 5674
+    },
+    {
+      "epoch": 2.2672792648821414,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5675
+    },
+    {
+      "epoch": 2.267678785457451,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 5676
+    },
+    {
+      "epoch": 2.2680783060327605,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 5677
+    },
+    {
+      "epoch": 2.2684778266080703,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 5678
+    },
+    {
+      "epoch": 2.26887734718338,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 5679
+    },
+    {
+      "epoch": 2.2692768677586894,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 5680
+    },
+    {
+      "epoch": 2.269676388333999,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 5681
+    },
+    {
+      "epoch": 2.270075908909309,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 5682
+    },
+    {
+      "epoch": 2.2704754294846183,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 5683
+    },
+    {
+      "epoch": 2.270874950059928,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 5684
+    },
+    {
+      "epoch": 2.271274470635238,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 5685
+    },
+    {
+      "epoch": 2.2716739912105472,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 5686
+    },
+    {
+      "epoch": 2.272073511785857,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 5687
+    },
+    {
+      "epoch": 2.272473032361167,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 5688
+    },
+    {
+      "epoch": 2.272872552936476,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5689
+    },
+    {
+      "epoch": 2.273272073511786,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0403,
+      "step": 5690
+    },
+    {
+      "epoch": 2.2736715940870953,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 5691
+    },
+    {
+      "epoch": 2.274071114662405,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 5692
+    },
+    {
+      "epoch": 2.274470635237715,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5693
+    },
+    {
+      "epoch": 2.274870155813024,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 5694
+    },
+    {
+      "epoch": 2.275269676388334,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 5695
+    },
+    {
+      "epoch": 2.275669196963644,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0739,
+      "step": 5696
+    },
+    {
+      "epoch": 2.276068717538953,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 5697
+    },
+    {
+      "epoch": 2.276468238114263,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 5698
+    },
+    {
+      "epoch": 2.2768677586895727,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5699
+    },
+    {
+      "epoch": 2.277267279264882,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0736,
+      "step": 5700
+    },
+    {
+      "epoch": 2.277666799840192,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 5701
+    },
+    {
+      "epoch": 2.278066320415501,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 5702
+    },
+    {
+      "epoch": 2.278465840990811,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 5703
+    },
+    {
+      "epoch": 2.2788653615661207,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5704
+    },
+    {
+      "epoch": 2.2792648821414305,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 5705
+    },
+    {
+      "epoch": 2.27966440271674,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 5706
+    },
+    {
+      "epoch": 2.2800639232920497,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5707
+    },
+    {
+      "epoch": 2.280463443867359,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5708
+    },
+    {
+      "epoch": 2.280862964442669,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 5709
+    },
+    {
+      "epoch": 2.2812624850179786,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0405,
+      "step": 5710
+    },
+    {
+      "epoch": 2.281662005593288,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 5711
+    },
+    {
+      "epoch": 2.2820615261685977,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 5712
+    },
+    {
+      "epoch": 2.2824610467439075,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 5713
+    },
+    {
+      "epoch": 2.282860567319217,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 5714
+    },
+    {
+      "epoch": 2.2832600878945266,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 5715
+    },
+    {
+      "epoch": 2.2836596084698364,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 5716
+    },
+    {
+      "epoch": 2.2840591290451457,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 5717
+    },
+    {
+      "epoch": 2.2844586496204555,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 5718
+    },
+    {
+      "epoch": 2.284858170195765,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5719
+    },
+    {
+      "epoch": 2.2852576907710747,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5720
+    },
+    {
+      "epoch": 2.2856572113463844,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 5721
+    },
+    {
+      "epoch": 2.286056731921694,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 5722
+    },
+    {
+      "epoch": 2.2864562524970036,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 5723
+    },
+    {
+      "epoch": 2.2868557730723134,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5724
+    },
+    {
+      "epoch": 2.2872552936476227,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 5725
+    },
+    {
+      "epoch": 2.2876548142229325,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 5726
+    },
+    {
+      "epoch": 2.2880543347982423,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 5727
+    },
+    {
+      "epoch": 2.2884538553735516,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0712,
+      "step": 5728
+    },
+    {
+      "epoch": 2.2888533759488614,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 5729
+    },
+    {
+      "epoch": 2.289252896524171,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 5730
+    },
+    {
+      "epoch": 2.2896524170994805,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 5731
+    },
+    {
+      "epoch": 2.2900519376747903,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5732
+    },
+    {
+      "epoch": 2.2904514582501,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 5733
+    },
+    {
+      "epoch": 2.2908509788254094,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 5734
+    },
+    {
+      "epoch": 2.2912504994007192,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 5735
+    },
+    {
+      "epoch": 2.2916500199760286,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0421,
+      "step": 5736
+    },
+    {
+      "epoch": 2.2920495405513384,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 5737
+    },
+    {
+      "epoch": 2.292449061126648,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 5738
+    },
+    {
+      "epoch": 2.2928485817019575,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 5739
+    },
+    {
+      "epoch": 2.2932481022772673,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 5740
+    },
+    {
+      "epoch": 2.293647622852577,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 5741
+    },
+    {
+      "epoch": 2.2940471434278864,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 5742
+    },
+    {
+      "epoch": 2.294446664003196,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5743
+    },
+    {
+      "epoch": 2.294846184578506,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5744
+    },
+    {
+      "epoch": 2.2952457051538153,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5745
+    },
+    {
+      "epoch": 2.295645225729125,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 5746
+    },
+    {
+      "epoch": 2.2960447463044344,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 5747
+    },
+    {
+      "epoch": 2.2964442668797442,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 5748
+    },
+    {
+      "epoch": 2.296843787455054,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5749
+    },
+    {
+      "epoch": 2.297243308030364,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5750
+    },
+    {
+      "epoch": 2.297642828605673,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 5751
+    },
+    {
+      "epoch": 2.298042349180983,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 5752
+    },
+    {
+      "epoch": 2.2984418697562923,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 5753
+    },
+    {
+      "epoch": 2.298841390331602,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 5754
+    },
+    {
+      "epoch": 2.299240910906912,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5755
+    },
+    {
+      "epoch": 2.299640431482221,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 5756
+    },
+    {
+      "epoch": 2.300039952057531,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5757
+    },
+    {
+      "epoch": 2.3004394726328408,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 5758
+    },
+    {
+      "epoch": 2.30083899320815,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 5759
+    },
+    {
+      "epoch": 2.30123851378346,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5760
+    },
+    {
+      "epoch": 2.3016380343587697,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 5761
+    },
+    {
+      "epoch": 2.302037554934079,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 5762
+    },
+    {
+      "epoch": 2.302437075509389,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5763
+    },
+    {
+      "epoch": 2.302836596084698,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 5764
+    },
+    {
+      "epoch": 2.303236116660008,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5765
+    },
+    {
+      "epoch": 2.3036356372353177,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5766
+    },
+    {
+      "epoch": 2.304035157810627,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5767
+    },
+    {
+      "epoch": 2.304434678385937,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5768
+    },
+    {
+      "epoch": 2.3048341989612466,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 5769
+    },
+    {
+      "epoch": 2.305233719536556,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 5770
+    },
+    {
+      "epoch": 2.3056332401118658,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 5771
+    },
+    {
+      "epoch": 2.3060327606871756,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 5772
+    },
+    {
+      "epoch": 2.306432281262485,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 5773
+    },
+    {
+      "epoch": 2.3068318018377947,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5774
+    },
+    {
+      "epoch": 2.3072313224131045,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 5775
+    },
+    {
+      "epoch": 2.307630842988414,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 5776
+    },
+    {
+      "epoch": 2.3080303635637236,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 5777
+    },
+    {
+      "epoch": 2.3084298841390334,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 5778
+    },
+    {
+      "epoch": 2.3088294047143427,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5779
+    },
+    {
+      "epoch": 2.3092289252896525,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 5780
+    },
+    {
+      "epoch": 2.309628445864962,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 5781
+    },
+    {
+      "epoch": 2.3100279664402716,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 5782
+    },
+    {
+      "epoch": 2.3104274870155814,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5783
+    },
+    {
+      "epoch": 2.3108270075908908,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 5784
+    },
+    {
+      "epoch": 2.3112265281662006,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5785
+    },
+    {
+      "epoch": 2.3116260487415103,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5786
+    },
+    {
+      "epoch": 2.3120255693168197,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 5787
+    },
+    {
+      "epoch": 2.3124250898921295,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5788
+    },
+    {
+      "epoch": 2.3128246104674393,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5789
+    },
+    {
+      "epoch": 2.3132241310427486,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 5790
+    },
+    {
+      "epoch": 2.3136236516180584,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 5791
+    },
+    {
+      "epoch": 2.3140231721933677,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 5792
+    },
+    {
+      "epoch": 2.3144226927686775,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 5793
+    },
+    {
+      "epoch": 2.3148222133439873,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 5794
+    },
+    {
+      "epoch": 2.315221733919297,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 5795
+    },
+    {
+      "epoch": 2.3156212544946064,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 5796
+    },
+    {
+      "epoch": 2.316020775069916,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 5797
+    },
+    {
+      "epoch": 2.3164202956452256,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 5798
+    },
+    {
+      "epoch": 2.3168198162205353,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 5799
+    },
+    {
+      "epoch": 2.317219336795845,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 5800
+    },
+    {
+      "epoch": 2.3176188573711545,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5801
+    },
+    {
+      "epoch": 2.3180183779464643,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 5802
+    },
+    {
+      "epoch": 2.318417898521774,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 5803
+    },
+    {
+      "epoch": 2.3188174190970834,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5804
+    },
+    {
+      "epoch": 2.319216939672393,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 5805
+    },
+    {
+      "epoch": 2.319616460247703,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5806
+    },
+    {
+      "epoch": 2.3200159808230123,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0728,
+      "step": 5807
+    },
+    {
+      "epoch": 2.320415501398322,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5808
+    },
+    {
+      "epoch": 2.3208150219736314,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 5809
+    },
+    {
+      "epoch": 2.321214542548941,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 5810
+    },
+    {
+      "epoch": 2.321614063124251,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5811
+    },
+    {
+      "epoch": 2.3220135836995603,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 5812
+    },
+    {
+      "epoch": 2.32241310427487,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 5813
+    },
+    {
+      "epoch": 2.32281262485018,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 5814
+    },
+    {
+      "epoch": 2.3232121454254893,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 5815
+    },
+    {
+      "epoch": 2.323611666000799,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 5816
+    },
+    {
+      "epoch": 2.324011186576109,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0711,
+      "step": 5817
+    },
+    {
+      "epoch": 2.324410707151418,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 5818
+    },
+    {
+      "epoch": 2.324810227726728,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5819
+    },
+    {
+      "epoch": 2.3252097483020377,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 5820
+    },
+    {
+      "epoch": 2.325609268877347,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5821
+    },
+    {
+      "epoch": 2.326008789452657,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 5822
+    },
+    {
+      "epoch": 2.3264083100279667,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 5823
+    },
+    {
+      "epoch": 2.326807830603276,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 5824
+    },
+    {
+      "epoch": 2.327207351178586,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5825
+    },
+    {
+      "epoch": 2.327606871753895,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5826
+    },
+    {
+      "epoch": 2.328006392329205,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 5827
+    },
+    {
+      "epoch": 2.3284059129045147,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 5828
+    },
+    {
+      "epoch": 2.328805433479824,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 5829
+    },
+    {
+      "epoch": 2.329204954055134,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 5830
+    },
+    {
+      "epoch": 2.3296044746304436,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 5831
+    },
+    {
+      "epoch": 2.330003995205753,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 5832
+    },
+    {
+      "epoch": 2.3304035157810628,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 5833
+    },
+    {
+      "epoch": 2.3308030363563725,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 5834
+    },
+    {
+      "epoch": 2.331202556931682,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 5835
+    },
+    {
+      "epoch": 2.3316020775069917,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 5836
+    },
+    {
+      "epoch": 2.332001598082301,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 5837
+    },
+    {
+      "epoch": 2.332401118657611,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 5838
+    },
+    {
+      "epoch": 2.3328006392329206,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 5839
+    },
+    {
+      "epoch": 2.3332001598082304,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5840
+    },
+    {
+      "epoch": 2.3335996803835397,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5841
+    },
+    {
+      "epoch": 2.3339992009588495,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 5842
+    },
+    {
+      "epoch": 2.334398721534159,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 5843
+    },
+    {
+      "epoch": 2.3347982421094686,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 5844
+    },
+    {
+      "epoch": 2.3351977626847784,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 5845
+    },
+    {
+      "epoch": 2.3355972832600878,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0772,
+      "step": 5846
+    },
+    {
+      "epoch": 2.3359968038353975,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 5847
+    },
+    {
+      "epoch": 2.3363963244107073,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 5848
+    },
+    {
+      "epoch": 2.3367958449860167,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5849
+    },
+    {
+      "epoch": 2.3371953655613265,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5850
+    },
+    {
+      "epoch": 2.3375948861366362,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 5851
+    },
+    {
+      "epoch": 2.3379944067119456,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5852
+    },
+    {
+      "epoch": 2.3383939272872554,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 5853
+    },
+    {
+      "epoch": 2.3387934478625647,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 5854
+    },
+    {
+      "epoch": 2.3391929684378745,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 5855
+    },
+    {
+      "epoch": 2.3395924890131843,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5856
+    },
+    {
+      "epoch": 2.3399920095884936,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 5857
+    },
+    {
+      "epoch": 2.3403915301638034,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 5858
+    },
+    {
+      "epoch": 2.340791050739113,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 5859
+    },
+    {
+      "epoch": 2.3411905713144225,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 5860
+    },
+    {
+      "epoch": 2.3415900918897323,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5861
+    },
+    {
+      "epoch": 2.341989612465042,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 5862
+    },
+    {
+      "epoch": 2.3423891330403515,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 5863
+    },
+    {
+      "epoch": 2.3427886536156612,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 5864
+    },
+    {
+      "epoch": 2.343188174190971,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 5865
+    },
+    {
+      "epoch": 2.3435876947662804,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 5866
+    },
+    {
+      "epoch": 2.34398721534159,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 5867
+    },
+    {
+      "epoch": 2.3443867359169,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 5868
+    },
+    {
+      "epoch": 2.3447862564922093,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5869
+    },
+    {
+      "epoch": 2.345185777067519,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 5870
+    },
+    {
+      "epoch": 2.3455852976428284,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 5871
+    },
+    {
+      "epoch": 2.345984818218138,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 5872
+    },
+    {
+      "epoch": 2.346384338793448,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 5873
+    },
+    {
+      "epoch": 2.3467838593687573,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0685,
+      "step": 5874
+    },
+    {
+      "epoch": 2.347183379944067,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 5875
+    },
+    {
+      "epoch": 2.347582900519377,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 5876
+    },
+    {
+      "epoch": 2.3479824210946862,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 5877
+    },
+    {
+      "epoch": 2.348381941669996,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 5878
+    },
+    {
+      "epoch": 2.348781462245306,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 5879
+    },
+    {
+      "epoch": 2.349180982820615,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 5880
+    },
+    {
+      "epoch": 2.349580503395925,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 5881
+    },
+    {
+      "epoch": 2.3499800239712343,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 5882
+    },
+    {
+      "epoch": 2.350379544546544,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 5883
+    },
+    {
+      "epoch": 2.350779065121854,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 5884
+    },
+    {
+      "epoch": 2.3511785856971636,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 5885
+    },
+    {
+      "epoch": 2.351578106272473,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 5886
+    },
+    {
+      "epoch": 2.3519776268477828,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 5887
+    },
+    {
+      "epoch": 2.352377147423092,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 5888
+    },
+    {
+      "epoch": 2.352776667998402,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 5889
+    },
+    {
+      "epoch": 2.3531761885737117,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 5890
+    },
+    {
+      "epoch": 2.353575709149021,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 5891
+    },
+    {
+      "epoch": 2.353975229724331,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 5892
+    },
+    {
+      "epoch": 2.3543747502996406,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0726,
+      "step": 5893
+    },
+    {
+      "epoch": 2.35477427087495,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 5894
+    },
+    {
+      "epoch": 2.3551737914502597,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 5895
+    },
+    {
+      "epoch": 2.3555733120255695,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 5896
+    },
+    {
+      "epoch": 2.355972832600879,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 5897
+    },
+    {
+      "epoch": 2.3563723531761886,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 5898
+    },
+    {
+      "epoch": 2.356771873751498,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 5899
+    },
+    {
+      "epoch": 2.357171394326808,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 5900
+    },
+    {
+      "epoch": 2.3575709149021176,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 5901
+    },
+    {
+      "epoch": 2.357970435477427,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 5902
+    },
+    {
+      "epoch": 2.3583699560527367,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 5903
+    },
+    {
+      "epoch": 2.3587694766280465,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 5904
+    },
+    {
+      "epoch": 2.359168997203356,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 5905
+    },
+    {
+      "epoch": 2.3595685177786656,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 5906
+    },
+    {
+      "epoch": 2.3599680383539754,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 5907
+    },
+    {
+      "epoch": 2.3603675589292847,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 5908
+    },
+    {
+      "epoch": 2.3607670795045945,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 5909
+    },
+    {
+      "epoch": 2.361166600079904,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 5910
+    },
+    {
+      "epoch": 2.3615661206552137,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 5911
+    },
+    {
+      "epoch": 2.3619656412305234,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 5912
+    },
+    {
+      "epoch": 2.3623651618058332,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5913
+    },
+    {
+      "epoch": 2.3627646823811426,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 5914
+    },
+    {
+      "epoch": 2.3631642029564524,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 5915
+    },
+    {
+      "epoch": 2.3635637235317617,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 5916
+    },
+    {
+      "epoch": 2.3639632441070715,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5917
+    },
+    {
+      "epoch": 2.3643627646823813,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 5918
+    },
+    {
+      "epoch": 2.3647622852576906,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 5919
+    },
+    {
+      "epoch": 2.3651618058330004,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 5920
+    },
+    {
+      "epoch": 2.36556132640831,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 5921
+    },
+    {
+      "epoch": 2.3659608469836195,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 5922
+    },
+    {
+      "epoch": 2.3663603675589293,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 5923
+    },
+    {
+      "epoch": 2.366759888134239,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 5924
+    },
+    {
+      "epoch": 2.3671594087095484,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 5925
+    },
+    {
+      "epoch": 2.3675589292848582,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 5926
+    },
+    {
+      "epoch": 2.3679584498601676,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 5927
+    },
+    {
+      "epoch": 2.3683579704354774,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5928
+    },
+    {
+      "epoch": 2.368757491010787,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 5929
+    },
+    {
+      "epoch": 2.369157011586097,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 5930
+    },
+    {
+      "epoch": 2.3695565321614063,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 5931
+    },
+    {
+      "epoch": 2.369956052736716,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 5932
+    },
+    {
+      "epoch": 2.3703555733120254,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 5933
+    },
+    {
+      "epoch": 2.370755093887335,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 5934
+    },
+    {
+      "epoch": 2.371154614462645,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5935
+    },
+    {
+      "epoch": 2.3715541350379543,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 5936
+    },
+    {
+      "epoch": 2.371953655613264,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 5937
+    },
+    {
+      "epoch": 2.372353176188574,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 5938
+    },
+    {
+      "epoch": 2.3727526967638832,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 5939
+    },
+    {
+      "epoch": 2.373152217339193,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 5940
+    },
+    {
+      "epoch": 2.373551737914503,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 5941
+    },
+    {
+      "epoch": 2.373951258489812,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 5942
+    },
+    {
+      "epoch": 2.374350779065122,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 5943
+    },
+    {
+      "epoch": 2.3747502996404313,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 5944
+    },
+    {
+      "epoch": 2.375149820215741,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0733,
+      "step": 5945
+    },
+    {
+      "epoch": 2.375549340791051,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 5946
+    },
+    {
+      "epoch": 2.37594886136636,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 5947
+    },
+    {
+      "epoch": 2.37634838194167,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 5948
+    },
+    {
+      "epoch": 2.3767479025169798,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 5949
+    },
+    {
+      "epoch": 2.377147423092289,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 5950
+    },
+    {
+      "epoch": 2.377546943667599,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 5951
+    },
+    {
+      "epoch": 2.3779464642429087,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 5952
+    },
+    {
+      "epoch": 2.378345984818218,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 5953
+    },
+    {
+      "epoch": 2.378745505393528,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 5954
+    },
+    {
+      "epoch": 2.379145025968837,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 5955
+    },
+    {
+      "epoch": 2.379544546544147,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 5956
+    },
+    {
+      "epoch": 2.3799440671194567,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 5957
+    },
+    {
+      "epoch": 2.3803435876947665,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 5958
+    },
+    {
+      "epoch": 2.380743108270076,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 5959
+    },
+    {
+      "epoch": 2.3811426288453856,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 5960
+    },
+    {
+      "epoch": 2.381542149420695,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 5961
+    },
+    {
+      "epoch": 2.3819416699960048,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 5962
+    },
+    {
+      "epoch": 2.3823411905713145,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 5963
+    },
+    {
+      "epoch": 2.382740711146624,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 5964
+    },
+    {
+      "epoch": 2.3831402317219337,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5965
+    },
+    {
+      "epoch": 2.3835397522972435,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 5966
+    },
+    {
+      "epoch": 2.383939272872553,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 5967
+    },
+    {
+      "epoch": 2.3843387934478626,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 5968
+    },
+    {
+      "epoch": 2.3847383140231724,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 5969
+    },
+    {
+      "epoch": 2.3851378345984817,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 5970
+    },
+    {
+      "epoch": 2.3855373551737915,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 5971
+    },
+    {
+      "epoch": 2.385936875749101,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 5972
+    },
+    {
+      "epoch": 2.3863363963244106,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0402,
+      "step": 5973
+    },
+    {
+      "epoch": 2.3867359168997204,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 5974
+    },
+    {
+      "epoch": 2.38713543747503,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 5975
+    },
+    {
+      "epoch": 2.3875349580503396,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 5976
+    },
+    {
+      "epoch": 2.3879344786256493,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 5977
+    },
+    {
+      "epoch": 2.3883339992009587,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 5978
+    },
+    {
+      "epoch": 2.3887335197762685,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 5979
+    },
+    {
+      "epoch": 2.3891330403515783,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 5980
+    },
+    {
+      "epoch": 2.3895325609268876,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 5981
+    },
+    {
+      "epoch": 2.3899320815021974,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 5982
+    },
+    {
+      "epoch": 2.390331602077507,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 5983
+    },
+    {
+      "epoch": 2.3907311226528165,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 5984
+    },
+    {
+      "epoch": 2.3911306432281263,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 5985
+    },
+    {
+      "epoch": 2.391530163803436,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 5986
+    },
+    {
+      "epoch": 2.3919296843787454,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 5987
+    },
+    {
+      "epoch": 2.392329204954055,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 5988
+    },
+    {
+      "epoch": 2.3927287255293646,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 5989
+    },
+    {
+      "epoch": 2.3931282461046743,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 5990
+    },
+    {
+      "epoch": 2.393527766679984,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 5991
+    },
+    {
+      "epoch": 2.3939272872552935,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 5992
+    },
+    {
+      "epoch": 2.3943268078306033,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 5993
+    },
+    {
+      "epoch": 2.394726328405913,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 5994
+    },
+    {
+      "epoch": 2.3951258489812224,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 5995
+    },
+    {
+      "epoch": 2.395525369556532,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 5996
+    },
+    {
+      "epoch": 2.395924890131842,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 5997
+    },
+    {
+      "epoch": 2.3963244107071513,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 5998
+    },
+    {
+      "epoch": 2.396723931282461,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 5999
+    },
+    {
+      "epoch": 2.3971234518577704,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6000
+    },
+    {
+      "epoch": 2.39752297243308,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6001
+    },
+    {
+      "epoch": 2.39792249300839,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 6002
+    },
+    {
+      "epoch": 2.3983220135837,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 6003
+    },
+    {
+      "epoch": 2.398721534159009,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6004
+    },
+    {
+      "epoch": 2.399121054734319,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 6005
+    },
+    {
+      "epoch": 2.3995205753096283,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6006
+    },
+    {
+      "epoch": 2.399920095884938,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6007
+    },
+    {
+      "epoch": 2.400319616460248,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 6008
+    },
+    {
+      "epoch": 2.400719137035557,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6009
+    },
+    {
+      "epoch": 2.401118657610867,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6010
+    },
+    {
+      "epoch": 2.4015181781861767,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6011
+    },
+    {
+      "epoch": 2.401917698761486,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 6012
+    },
+    {
+      "epoch": 2.402317219336796,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.043,
+      "step": 6013
+    },
+    {
+      "epoch": 2.4027167399121057,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6014
+    },
+    {
+      "epoch": 2.403116260487415,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 6015
+    },
+    {
+      "epoch": 2.403515781062725,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6016
+    },
+    {
+      "epoch": 2.403915301638034,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 6017
+    },
+    {
+      "epoch": 2.404314822213344,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 6018
+    },
+    {
+      "epoch": 2.4047143427886537,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6019
+    },
+    {
+      "epoch": 2.4051138633639635,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6020
+    },
+    {
+      "epoch": 2.405513383939273,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 6021
+    },
+    {
+      "epoch": 2.4059129045145826,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 6022
+    },
+    {
+      "epoch": 2.406312425089892,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6023
+    },
+    {
+      "epoch": 2.4067119456652017,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 6024
+    },
+    {
+      "epoch": 2.4071114662405115,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 6025
+    },
+    {
+      "epoch": 2.407510986815821,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 6026
+    },
+    {
+      "epoch": 2.4079105073911307,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 6027
+    },
+    {
+      "epoch": 2.4083100279664404,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 6028
+    },
+    {
+      "epoch": 2.40870954854175,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 6029
+    },
+    {
+      "epoch": 2.4091090691170596,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 6030
+    },
+    {
+      "epoch": 2.4095085896923694,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6031
+    },
+    {
+      "epoch": 2.4099081102676787,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 6032
+    },
+    {
+      "epoch": 2.4103076308429885,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 6033
+    },
+    {
+      "epoch": 2.410707151418298,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 6034
+    },
+    {
+      "epoch": 2.4111066719936076,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6035
+    },
+    {
+      "epoch": 2.4115061925689174,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 6036
+    },
+    {
+      "epoch": 2.4119057131442267,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 6037
+    },
+    {
+      "epoch": 2.4123052337195365,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 6038
+    },
+    {
+      "epoch": 2.4127047542948463,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6039
+    },
+    {
+      "epoch": 2.4131042748701557,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 6040
+    },
+    {
+      "epoch": 2.4135037954454654,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6041
+    },
+    {
+      "epoch": 2.4139033160207752,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 6042
+    },
+    {
+      "epoch": 2.4143028365960846,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6043
+    },
+    {
+      "epoch": 2.4147023571713944,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6044
+    },
+    {
+      "epoch": 2.4151018777467037,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 6045
+    },
+    {
+      "epoch": 2.4155013983220135,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6046
+    },
+    {
+      "epoch": 2.4159009188973233,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 6047
+    },
+    {
+      "epoch": 2.416300439472633,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 6048
+    },
+    {
+      "epoch": 2.4166999600479424,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6049
+    },
+    {
+      "epoch": 2.417099480623252,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 6050
+    },
+    {
+      "epoch": 2.4174990011985615,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 6051
+    },
+    {
+      "epoch": 2.4178985217738713,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 6052
+    },
+    {
+      "epoch": 2.418298042349181,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 6053
+    },
+    {
+      "epoch": 2.4186975629244905,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6054
+    },
+    {
+      "epoch": 2.4190970834998002,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 6055
+    },
+    {
+      "epoch": 2.41949660407511,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 6056
+    },
+    {
+      "epoch": 2.4198961246504194,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6057
+    },
+    {
+      "epoch": 2.420295645225729,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 6058
+    },
+    {
+      "epoch": 2.420695165801039,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 6059
+    },
+    {
+      "epoch": 2.4210946863763483,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 6060
+    },
+    {
+      "epoch": 2.421494206951658,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6061
+    },
+    {
+      "epoch": 2.4218937275269674,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 6062
+    },
+    {
+      "epoch": 2.422293248102277,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 6063
+    },
+    {
+      "epoch": 2.422692768677587,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6064
+    },
+    {
+      "epoch": 2.4230922892528968,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6065
+    },
+    {
+      "epoch": 2.423491809828206,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 6066
+    },
+    {
+      "epoch": 2.423891330403516,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 6067
+    },
+    {
+      "epoch": 2.4242908509788252,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 6068
+    },
+    {
+      "epoch": 2.424690371554135,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6069
+    },
+    {
+      "epoch": 2.425089892129445,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 6070
+    },
+    {
+      "epoch": 2.425489412704754,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 6071
+    },
+    {
+      "epoch": 2.425888933280064,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 6072
+    },
+    {
+      "epoch": 2.4262884538553737,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6073
+    },
+    {
+      "epoch": 2.426687974430683,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 6074
+    },
+    {
+      "epoch": 2.427087495005993,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 6075
+    },
+    {
+      "epoch": 2.4274870155813026,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 6076
+    },
+    {
+      "epoch": 2.427886536156612,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 6077
+    },
+    {
+      "epoch": 2.4282860567319218,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6078
+    },
+    {
+      "epoch": 2.428685577307231,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 6079
+    },
+    {
+      "epoch": 2.429085097882541,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 6080
+    },
+    {
+      "epoch": 2.4294846184578507,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6081
+    },
+    {
+      "epoch": 2.42988413903316,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 6082
+    },
+    {
+      "epoch": 2.43028365960847,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 6083
+    },
+    {
+      "epoch": 2.4306831801837796,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6084
+    },
+    {
+      "epoch": 2.431082700759089,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 6085
+    },
+    {
+      "epoch": 2.4314822213343987,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6086
+    },
+    {
+      "epoch": 2.4318817419097085,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6087
+    },
+    {
+      "epoch": 2.432281262485018,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6088
+    },
+    {
+      "epoch": 2.4326807830603276,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 6089
+    },
+    {
+      "epoch": 2.433080303635637,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6090
+    },
+    {
+      "epoch": 2.4334798242109468,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 6091
+    },
+    {
+      "epoch": 2.4338793447862566,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0406,
+      "step": 6092
+    },
+    {
+      "epoch": 2.4342788653615663,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 6093
+    },
+    {
+      "epoch": 2.4346783859368757,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 6094
+    },
+    {
+      "epoch": 2.4350779065121855,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0824,
+      "step": 6095
+    },
+    {
+      "epoch": 2.435477427087495,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 6096
+    },
+    {
+      "epoch": 2.4358769476628046,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 6097
+    },
+    {
+      "epoch": 2.4362764682381144,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 6098
+    },
+    {
+      "epoch": 2.4366759888134237,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6099
+    },
+    {
+      "epoch": 2.4370755093887335,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 6100
+    },
+    {
+      "epoch": 2.4374750299640433,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 6101
+    },
+    {
+      "epoch": 2.4378745505393526,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6102
+    },
+    {
+      "epoch": 2.4382740711146624,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6103
+    },
+    {
+      "epoch": 2.438673591689972,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6104
+    },
+    {
+      "epoch": 2.4390731122652816,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 6105
+    },
+    {
+      "epoch": 2.4394726328405913,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0367,
+      "step": 6106
+    },
+    {
+      "epoch": 2.4398721534159007,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 6107
+    },
+    {
+      "epoch": 2.4402716739912105,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 6108
+    },
+    {
+      "epoch": 2.4406711945665203,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 6109
+    },
+    {
+      "epoch": 2.44107071514183,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6110
+    },
+    {
+      "epoch": 2.4414702357171394,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6111
+    },
+    {
+      "epoch": 2.441869756292449,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 6112
+    },
+    {
+      "epoch": 2.4422692768677585,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6113
+    },
+    {
+      "epoch": 2.4426687974430683,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 6114
+    },
+    {
+      "epoch": 2.443068318018378,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6115
+    },
+    {
+      "epoch": 2.4434678385936874,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6116
+    },
+    {
+      "epoch": 2.443867359168997,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 6117
+    },
+    {
+      "epoch": 2.444266879744307,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 6118
+    },
+    {
+      "epoch": 2.4446664003196164,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 6119
+    },
+    {
+      "epoch": 2.445065920894926,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6120
+    },
+    {
+      "epoch": 2.445465441470236,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 6121
+    },
+    {
+      "epoch": 2.4458649620455453,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6122
+    },
+    {
+      "epoch": 2.446264482620855,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6123
+    },
+    {
+      "epoch": 2.4466640031961644,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 6124
+    },
+    {
+      "epoch": 2.447063523771474,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 6125
+    },
+    {
+      "epoch": 2.447463044346784,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 6126
+    },
+    {
+      "epoch": 2.4478625649220933,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 6127
+    },
+    {
+      "epoch": 2.448262085497403,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 6128
+    },
+    {
+      "epoch": 2.448661606072713,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 6129
+    },
+    {
+      "epoch": 2.4490611266480222,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6130
+    },
+    {
+      "epoch": 2.449460647223332,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 6131
+    },
+    {
+      "epoch": 2.449860167798642,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 6132
+    },
+    {
+      "epoch": 2.450259688373951,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 6133
+    },
+    {
+      "epoch": 2.450659208949261,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 6134
+    },
+    {
+      "epoch": 2.4510587295245703,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6135
+    },
+    {
+      "epoch": 2.45145825009988,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 6136
+    },
+    {
+      "epoch": 2.45185777067519,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 6137
+    },
+    {
+      "epoch": 2.4522572912504996,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 6138
+    },
+    {
+      "epoch": 2.452656811825809,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 6139
+    },
+    {
+      "epoch": 2.4530563324011188,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6140
+    },
+    {
+      "epoch": 2.453455852976428,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 6141
+    },
+    {
+      "epoch": 2.453855373551738,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6142
+    },
+    {
+      "epoch": 2.4542548941270477,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 6143
+    },
+    {
+      "epoch": 2.454654414702357,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 6144
+    },
+    {
+      "epoch": 2.455053935277667,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6145
+    },
+    {
+      "epoch": 2.4554534558529766,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 6146
+    },
+    {
+      "epoch": 2.455852976428286,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 6147
+    },
+    {
+      "epoch": 2.4562524970035957,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0665,
+      "step": 6148
+    },
+    {
+      "epoch": 2.4566520175789055,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6149
+    },
+    {
+      "epoch": 2.457051538154215,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6150
+    },
+    {
+      "epoch": 2.4574510587295246,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 6151
+    },
+    {
+      "epoch": 2.457850579304834,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 6152
+    },
+    {
+      "epoch": 2.4582500998801438,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 6153
+    },
+    {
+      "epoch": 2.4586496204554535,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 6154
+    },
+    {
+      "epoch": 2.4590491410307633,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 6155
+    },
+    {
+      "epoch": 2.4594486616060727,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6156
+    },
+    {
+      "epoch": 2.4598481821813825,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 6157
+    },
+    {
+      "epoch": 2.460247702756692,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 6158
+    },
+    {
+      "epoch": 2.4606472233320016,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6159
+    },
+    {
+      "epoch": 2.4610467439073114,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 6160
+    },
+    {
+      "epoch": 2.4614462644826207,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 6161
+    },
+    {
+      "epoch": 2.4618457850579305,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6162
+    },
+    {
+      "epoch": 2.4622453056332403,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 6163
+    },
+    {
+      "epoch": 2.4626448262085496,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 6164
+    },
+    {
+      "epoch": 2.4630443467838594,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 6165
+    },
+    {
+      "epoch": 2.463443867359169,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 6166
+    },
+    {
+      "epoch": 2.4638433879344785,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 6167
+    },
+    {
+      "epoch": 2.4642429085097883,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 6168
+    },
+    {
+      "epoch": 2.4646424290850977,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 6169
+    },
+    {
+      "epoch": 2.4650419496604075,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 6170
+    },
+    {
+      "epoch": 2.4654414702357172,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 6171
+    },
+    {
+      "epoch": 2.4658409908110266,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6172
+    },
+    {
+      "epoch": 2.4662405113863364,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6173
+    },
+    {
+      "epoch": 2.466640031961646,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 6174
+    },
+    {
+      "epoch": 2.4670395525369555,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6175
+    },
+    {
+      "epoch": 2.4674390731122653,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6176
+    },
+    {
+      "epoch": 2.467838593687575,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6177
+    },
+    {
+      "epoch": 2.4682381142628844,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 6178
+    },
+    {
+      "epoch": 2.468637634838194,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 6179
+    },
+    {
+      "epoch": 2.4690371554135035,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 6180
+    },
+    {
+      "epoch": 2.4694366759888133,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 6181
+    },
+    {
+      "epoch": 2.469836196564123,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 6182
+    },
+    {
+      "epoch": 2.470235717139433,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 6183
+    },
+    {
+      "epoch": 2.4706352377147422,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 6184
+    },
+    {
+      "epoch": 2.471034758290052,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6185
+    },
+    {
+      "epoch": 2.4714342788653614,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 6186
+    },
+    {
+      "epoch": 2.471833799440671,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6187
+    },
+    {
+      "epoch": 2.472233320015981,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 6188
+    },
+    {
+      "epoch": 2.4726328405912903,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 6189
+    },
+    {
+      "epoch": 2.4730323611666,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6190
+    },
+    {
+      "epoch": 2.47343188174191,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6191
+    },
+    {
+      "epoch": 2.473831402317219,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6192
+    },
+    {
+      "epoch": 2.474230922892529,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 6193
+    },
+    {
+      "epoch": 2.474630443467839,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 6194
+    },
+    {
+      "epoch": 2.475029964043148,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6195
+    },
+    {
+      "epoch": 2.475429484618458,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 6196
+    },
+    {
+      "epoch": 2.4758290051937673,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6197
+    },
+    {
+      "epoch": 2.476228525769077,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 6198
+    },
+    {
+      "epoch": 2.476628046344387,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6199
+    },
+    {
+      "epoch": 2.4770275669196966,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 6200
+    },
+    {
+      "epoch": 2.477427087495006,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6201
+    },
+    {
+      "epoch": 2.4778266080703157,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 6202
+    },
+    {
+      "epoch": 2.478226128645625,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 6203
+    },
+    {
+      "epoch": 2.478625649220935,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 6204
+    },
+    {
+      "epoch": 2.4790251697962447,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 6205
+    },
+    {
+      "epoch": 2.479424690371554,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 6206
+    },
+    {
+      "epoch": 2.479824210946864,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.07,
+      "step": 6207
+    },
+    {
+      "epoch": 2.4802237315221736,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 6208
+    },
+    {
+      "epoch": 2.480623252097483,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 6209
+    },
+    {
+      "epoch": 2.4810227726727927,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 6210
+    },
+    {
+      "epoch": 2.4814222932481025,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6211
+    },
+    {
+      "epoch": 2.481821813823412,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6212
+    },
+    {
+      "epoch": 2.4822213343987216,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 6213
+    },
+    {
+      "epoch": 2.482620854974031,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6214
+    },
+    {
+      "epoch": 2.4830203755493407,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 6215
+    },
+    {
+      "epoch": 2.4834198961246505,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6216
+    },
+    {
+      "epoch": 2.48381941669996,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 6217
+    },
+    {
+      "epoch": 2.4842189372752697,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 6218
+    },
+    {
+      "epoch": 2.4846184578505794,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 6219
+    },
+    {
+      "epoch": 2.485017978425889,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 6220
+    },
+    {
+      "epoch": 2.4854174990011986,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6221
+    },
+    {
+      "epoch": 2.4858170195765084,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 6222
+    },
+    {
+      "epoch": 2.4862165401518177,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 6223
+    },
+    {
+      "epoch": 2.4866160607271275,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 6224
+    },
+    {
+      "epoch": 2.487015581302437,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6225
+    },
+    {
+      "epoch": 2.4874151018777466,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 6226
+    },
+    {
+      "epoch": 2.4878146224530564,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 6227
+    },
+    {
+      "epoch": 2.488214143028366,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0715,
+      "step": 6228
+    },
+    {
+      "epoch": 2.4886136636036755,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 6229
+    },
+    {
+      "epoch": 2.4890131841789853,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 6230
+    },
+    {
+      "epoch": 2.4894127047542947,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 6231
+    },
+    {
+      "epoch": 2.4898122253296044,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 6232
+    },
+    {
+      "epoch": 2.4902117459049142,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 6233
+    },
+    {
+      "epoch": 2.4906112664802236,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 6234
+    },
+    {
+      "epoch": 2.4910107870555334,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 6235
+    },
+    {
+      "epoch": 2.491410307630843,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6236
+    },
+    {
+      "epoch": 2.4918098282061525,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 6237
+    },
+    {
+      "epoch": 2.4922093487814623,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 6238
+    },
+    {
+      "epoch": 2.492608869356772,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6239
+    },
+    {
+      "epoch": 2.4930083899320814,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 6240
+    },
+    {
+      "epoch": 2.493407910507391,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 6241
+    },
+    {
+      "epoch": 2.4938074310827005,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 6242
+    },
+    {
+      "epoch": 2.4942069516580103,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6243
+    },
+    {
+      "epoch": 2.49460647223332,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 6244
+    },
+    {
+      "epoch": 2.49500599280863,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 6245
+    },
+    {
+      "epoch": 2.4954055133839392,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.038,
+      "step": 6246
+    },
+    {
+      "epoch": 2.495805033959249,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 6247
+    },
+    {
+      "epoch": 2.4962045545345584,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 6248
+    },
+    {
+      "epoch": 2.496604075109868,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 6249
+    },
+    {
+      "epoch": 2.497003595685178,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 6250
+    },
+    {
+      "epoch": 2.4974031162604873,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 6251
+    },
+    {
+      "epoch": 2.497802636835797,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 6252
+    },
+    {
+      "epoch": 2.498202157411107,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 6253
+    },
+    {
+      "epoch": 2.498601677986416,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 6254
+    },
+    {
+      "epoch": 2.499001198561726,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0395,
+      "step": 6255
+    },
+    {
+      "epoch": 2.4994007191370358,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 6256
+    },
+    {
+      "epoch": 2.499800239712345,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0718,
+      "step": 6257
+    },
+    {
+      "epoch": 2.500199760287655,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 6258
+    },
+    {
+      "epoch": 2.5005992808629642,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6259
+    },
+    {
+      "epoch": 2.500998801438274,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 6260
+    },
+    {
+      "epoch": 2.501398322013584,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6261
+    },
+    {
+      "epoch": 2.5017978425888936,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 6262
+    },
+    {
+      "epoch": 2.502197363164203,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 6263
+    },
+    {
+      "epoch": 2.5025968837395127,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6264
+    },
+    {
+      "epoch": 2.502996404314822,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 6265
+    },
+    {
+      "epoch": 2.503395924890132,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6266
+    },
+    {
+      "epoch": 2.5037954454654416,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 6267
+    },
+    {
+      "epoch": 2.504194966040751,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6268
+    },
+    {
+      "epoch": 2.5045944866160608,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 6269
+    },
+    {
+      "epoch": 2.50499400719137,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 6270
+    },
+    {
+      "epoch": 2.50539352776668,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6271
+    },
+    {
+      "epoch": 2.5057930483419897,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 6272
+    },
+    {
+      "epoch": 2.5061925689172995,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6273
+    },
+    {
+      "epoch": 2.506592089492609,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 6274
+    },
+    {
+      "epoch": 2.5069916100679186,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6275
+    },
+    {
+      "epoch": 2.507391130643228,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 6276
+    },
+    {
+      "epoch": 2.5077906512185377,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0725,
+      "step": 6277
+    },
+    {
+      "epoch": 2.5081901717938475,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6278
+    },
+    {
+      "epoch": 2.508589692369157,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 6279
+    },
+    {
+      "epoch": 2.5089892129444666,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6280
+    },
+    {
+      "epoch": 2.5093887335197764,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 6281
+    },
+    {
+      "epoch": 2.5097882540950858,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6282
+    },
+    {
+      "epoch": 2.5101877746703956,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 6283
+    },
+    {
+      "epoch": 2.5105872952457053,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 6284
+    },
+    {
+      "epoch": 2.5109868158210147,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6285
+    },
+    {
+      "epoch": 2.5113863363963245,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 6286
+    },
+    {
+      "epoch": 2.511785856971634,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 6287
+    },
+    {
+      "epoch": 2.5121853775469436,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 6288
+    },
+    {
+      "epoch": 2.5125848981222534,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 6289
+    },
+    {
+      "epoch": 2.512984418697563,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 6290
+    },
+    {
+      "epoch": 2.5133839392728725,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6291
+    },
+    {
+      "epoch": 2.5137834598481823,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6292
+    },
+    {
+      "epoch": 2.5141829804234916,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 6293
+    },
+    {
+      "epoch": 2.5145825009988014,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 6294
+    },
+    {
+      "epoch": 2.514982021574111,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6295
+    },
+    {
+      "epoch": 2.5153815421494206,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0749,
+      "step": 6296
+    },
+    {
+      "epoch": 2.5157810627247303,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 6297
+    },
+    {
+      "epoch": 2.5161805833000397,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 6298
+    },
+    {
+      "epoch": 2.5165801038753495,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 6299
+    },
+    {
+      "epoch": 2.5169796244506593,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 6300
+    },
+    {
+      "epoch": 2.517379145025969,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 6301
+    },
+    {
+      "epoch": 2.5177786656012784,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 6302
+    },
+    {
+      "epoch": 2.518178186176588,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 6303
+    },
+    {
+      "epoch": 2.5185777067518975,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 6304
+    },
+    {
+      "epoch": 2.5189772273272073,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6305
+    },
+    {
+      "epoch": 2.519376747902517,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 6306
+    },
+    {
+      "epoch": 2.519776268477827,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6307
+    },
+    {
+      "epoch": 2.520175789053136,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 6308
+    },
+    {
+      "epoch": 2.520575309628446,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6309
+    },
+    {
+      "epoch": 2.5209748302037553,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 6310
+    },
+    {
+      "epoch": 2.521374350779065,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 6311
+    },
+    {
+      "epoch": 2.521773871354375,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 6312
+    },
+    {
+      "epoch": 2.5221733919296843,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 6313
+    },
+    {
+      "epoch": 2.522572912504994,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 6314
+    },
+    {
+      "epoch": 2.5229724330803034,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 6315
+    },
+    {
+      "epoch": 2.523371953655613,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 6316
+    },
+    {
+      "epoch": 2.523771474230923,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 6317
+    },
+    {
+      "epoch": 2.5241709948062327,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6318
+    },
+    {
+      "epoch": 2.524570515381542,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 6319
+    },
+    {
+      "epoch": 2.524970035956852,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 6320
+    },
+    {
+      "epoch": 2.525369556532161,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 6321
+    },
+    {
+      "epoch": 2.525769077107471,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 6322
+    },
+    {
+      "epoch": 2.526168597682781,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 6323
+    },
+    {
+      "epoch": 2.52656811825809,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 6324
+    },
+    {
+      "epoch": 2.5269676388334,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 6325
+    },
+    {
+      "epoch": 2.5273671594087097,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 6326
+    },
+    {
+      "epoch": 2.527766679984019,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 6327
+    },
+    {
+      "epoch": 2.528166200559329,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 6328
+    },
+    {
+      "epoch": 2.5285657211346386,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 6329
+    },
+    {
+      "epoch": 2.528965241709948,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 6330
+    },
+    {
+      "epoch": 2.5293647622852578,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 6331
+    },
+    {
+      "epoch": 2.529764282860567,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 6332
+    },
+    {
+      "epoch": 2.530163803435877,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 6333
+    },
+    {
+      "epoch": 2.5305633240111867,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 6334
+    },
+    {
+      "epoch": 2.5309628445864965,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 6335
+    },
+    {
+      "epoch": 2.531362365161806,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6336
+    },
+    {
+      "epoch": 2.5317618857371156,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 6337
+    },
+    {
+      "epoch": 2.532161406312425,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 6338
+    },
+    {
+      "epoch": 2.5325609268877347,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6339
+    },
+    {
+      "epoch": 2.5329604474630445,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 6340
+    },
+    {
+      "epoch": 2.533359968038354,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 6341
+    },
+    {
+      "epoch": 2.5337594886136636,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 6342
+    },
+    {
+      "epoch": 2.534159009188973,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 6343
+    },
+    {
+      "epoch": 2.5345585297642828,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 6344
+    },
+    {
+      "epoch": 2.5349580503395925,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 6345
+    },
+    {
+      "epoch": 2.5353575709149023,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 6346
+    },
+    {
+      "epoch": 2.5357570914902117,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 6347
+    },
+    {
+      "epoch": 2.5361566120655215,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6348
+    },
+    {
+      "epoch": 2.536556132640831,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 6349
+    },
+    {
+      "epoch": 2.5369556532161406,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 6350
+    },
+    {
+      "epoch": 2.5373551737914504,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 6351
+    },
+    {
+      "epoch": 2.53775469436676,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6352
+    },
+    {
+      "epoch": 2.5381542149420695,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 6353
+    },
+    {
+      "epoch": 2.5385537355173793,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 6354
+    },
+    {
+      "epoch": 2.5389532560926886,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 6355
+    },
+    {
+      "epoch": 2.5393527766679984,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 6356
+    },
+    {
+      "epoch": 2.539752297243308,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 6357
+    },
+    {
+      "epoch": 2.5401518178186175,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 6358
+    },
+    {
+      "epoch": 2.5405513383939273,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0722,
+      "step": 6359
+    },
+    {
+      "epoch": 2.5409508589692367,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 6360
+    },
+    {
+      "epoch": 2.5413503795445465,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6361
+    },
+    {
+      "epoch": 2.5417499001198562,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 6362
+    },
+    {
+      "epoch": 2.542149420695166,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 6363
+    },
+    {
+      "epoch": 2.5425489412704754,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 6364
+    },
+    {
+      "epoch": 2.542948461845785,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 6365
+    },
+    {
+      "epoch": 2.5433479824210945,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 6366
+    },
+    {
+      "epoch": 2.5437475029964043,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 6367
+    },
+    {
+      "epoch": 2.544147023571714,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6368
+    },
+    {
+      "epoch": 2.5445465441470234,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 6369
+    },
+    {
+      "epoch": 2.544946064722333,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 6370
+    },
+    {
+      "epoch": 2.545345585297643,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6371
+    },
+    {
+      "epoch": 2.5457451058729523,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 6372
+    },
+    {
+      "epoch": 2.546144626448262,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 6373
+    },
+    {
+      "epoch": 2.546544147023572,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 6374
+    },
+    {
+      "epoch": 2.5469436675988812,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6375
+    },
+    {
+      "epoch": 2.547343188174191,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6376
+    },
+    {
+      "epoch": 2.5477427087495004,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 6377
+    },
+    {
+      "epoch": 2.54814222932481,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6378
+    },
+    {
+      "epoch": 2.54854174990012,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 6379
+    },
+    {
+      "epoch": 2.5489412704754297,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6380
+    },
+    {
+      "epoch": 2.549340791050739,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 6381
+    },
+    {
+      "epoch": 2.549740311626049,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6382
+    },
+    {
+      "epoch": 2.550139832201358,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 6383
+    },
+    {
+      "epoch": 2.550539352776668,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6384
+    },
+    {
+      "epoch": 2.5509388733519778,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 6385
+    },
+    {
+      "epoch": 2.551338393927287,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6386
+    },
+    {
+      "epoch": 2.551737914502597,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6387
+    },
+    {
+      "epoch": 2.5521374350779062,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 6388
+    },
+    {
+      "epoch": 2.552536955653216,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6389
+    },
+    {
+      "epoch": 2.552936476228526,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 6390
+    },
+    {
+      "epoch": 2.5533359968038356,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 6391
+    },
+    {
+      "epoch": 2.553735517379145,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 6392
+    },
+    {
+      "epoch": 2.5541350379544547,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 6393
+    },
+    {
+      "epoch": 2.554534558529764,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 6394
+    },
+    {
+      "epoch": 2.554934079105074,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 6395
+    },
+    {
+      "epoch": 2.5553335996803836,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 6396
+    },
+    {
+      "epoch": 2.5557331202556934,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 6397
+    },
+    {
+      "epoch": 2.556132640831003,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6398
+    },
+    {
+      "epoch": 2.5565321614063126,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 6399
+    },
+    {
+      "epoch": 2.556931681981622,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6400
+    },
+    {
+      "epoch": 2.5573312025569317,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6401
+    },
+    {
+      "epoch": 2.5577307231322415,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6402
+    },
+    {
+      "epoch": 2.558130243707551,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 6403
+    },
+    {
+      "epoch": 2.5585297642828606,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 6404
+    },
+    {
+      "epoch": 2.55892928485817,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0394,
+      "step": 6405
+    },
+    {
+      "epoch": 2.5593288054334797,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 6406
+    },
+    {
+      "epoch": 2.5597283260087895,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 6407
+    },
+    {
+      "epoch": 2.5601278465840993,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6408
+    },
+    {
+      "epoch": 2.5605273671594087,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6409
+    },
+    {
+      "epoch": 2.5609268877347184,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 6410
+    },
+    {
+      "epoch": 2.561326408310028,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 6411
+    },
+    {
+      "epoch": 2.5617259288853376,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 6412
+    },
+    {
+      "epoch": 2.5621254494606474,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 6413
+    },
+    {
+      "epoch": 2.5625249700359567,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0688,
+      "step": 6414
+    },
+    {
+      "epoch": 2.5629244906112665,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 6415
+    },
+    {
+      "epoch": 2.5633240111865763,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 6416
+    },
+    {
+      "epoch": 2.5637235317618856,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 6417
+    },
+    {
+      "epoch": 2.5641230523371954,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 6418
+    },
+    {
+      "epoch": 2.564522572912505,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 6419
+    },
+    {
+      "epoch": 2.5649220934878145,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 6420
+    },
+    {
+      "epoch": 2.5653216140631243,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 6421
+    },
+    {
+      "epoch": 2.5657211346384337,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 6422
+    },
+    {
+      "epoch": 2.5661206552137434,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6423
+    },
+    {
+      "epoch": 2.5665201757890532,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 6424
+    },
+    {
+      "epoch": 2.566919696364363,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 6425
+    },
+    {
+      "epoch": 2.5673192169396724,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 6426
+    },
+    {
+      "epoch": 2.567718737514982,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 6427
+    },
+    {
+      "epoch": 2.5681182580902915,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 6428
+    },
+    {
+      "epoch": 2.5685177786656013,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 6429
+    },
+    {
+      "epoch": 2.568917299240911,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6430
+    },
+    {
+      "epoch": 2.5693168198162204,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 6431
+    },
+    {
+      "epoch": 2.56971634039153,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 6432
+    },
+    {
+      "epoch": 2.5701158609668395,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 6433
+    },
+    {
+      "epoch": 2.5705153815421493,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 6434
+    },
+    {
+      "epoch": 2.570914902117459,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 6435
+    },
+    {
+      "epoch": 2.571314422692769,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 6436
+    },
+    {
+      "epoch": 2.5717139432680782,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 6437
+    },
+    {
+      "epoch": 2.572113463843388,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 6438
+    },
+    {
+      "epoch": 2.5725129844186974,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6439
+    },
+    {
+      "epoch": 2.572912504994007,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 6440
+    },
+    {
+      "epoch": 2.573312025569317,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.039,
+      "step": 6441
+    },
+    {
+      "epoch": 2.5737115461446267,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.069,
+      "step": 6442
+    },
+    {
+      "epoch": 2.574111066719936,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 6443
+    },
+    {
+      "epoch": 2.574510587295246,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 6444
+    },
+    {
+      "epoch": 2.574910107870555,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6445
+    },
+    {
+      "epoch": 2.575309628445865,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 6446
+    },
+    {
+      "epoch": 2.5757091490211748,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 6447
+    },
+    {
+      "epoch": 2.576108669596484,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 6448
+    },
+    {
+      "epoch": 2.576508190171794,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 6449
+    },
+    {
+      "epoch": 2.5769077107471032,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6450
+    },
+    {
+      "epoch": 2.577307231322413,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 6451
+    },
+    {
+      "epoch": 2.577706751897723,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 6452
+    },
+    {
+      "epoch": 2.5781062724730326,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6453
+    },
+    {
+      "epoch": 2.578505793048342,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 6454
+    },
+    {
+      "epoch": 2.5789053136236517,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 6455
+    },
+    {
+      "epoch": 2.579304834198961,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 6456
+    },
+    {
+      "epoch": 2.579704354774271,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6457
+    },
+    {
+      "epoch": 2.5801038753495806,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 6458
+    },
+    {
+      "epoch": 2.58050339592489,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 6459
+    },
+    {
+      "epoch": 2.5809029165001998,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 6460
+    },
+    {
+      "epoch": 2.5813024370755095,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 6461
+    },
+    {
+      "epoch": 2.581701957650819,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6462
+    },
+    {
+      "epoch": 2.5821014782261287,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 6463
+    },
+    {
+      "epoch": 2.5825009988014385,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 6464
+    },
+    {
+      "epoch": 2.582900519376748,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 6465
+    },
+    {
+      "epoch": 2.5833000399520576,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 6466
+    },
+    {
+      "epoch": 2.583699560527367,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 6467
+    },
+    {
+      "epoch": 2.5840990811026767,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 6468
+    },
+    {
+      "epoch": 2.5844986016779865,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 6469
+    },
+    {
+      "epoch": 2.5848981222532963,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 6470
+    },
+    {
+      "epoch": 2.5852976428286056,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 6471
+    },
+    {
+      "epoch": 2.5856971634039154,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 6472
+    },
+    {
+      "epoch": 2.5860966839792248,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 6473
+    },
+    {
+      "epoch": 2.5864962045545346,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6474
+    },
+    {
+      "epoch": 2.5868957251298443,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 6475
+    },
+    {
+      "epoch": 2.5872952457051537,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 6476
+    },
+    {
+      "epoch": 2.5876947662804635,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6477
+    },
+    {
+      "epoch": 2.588094286855773,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 6478
+    },
+    {
+      "epoch": 2.5884938074310826,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6479
+    },
+    {
+      "epoch": 2.5888933280063924,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6480
+    },
+    {
+      "epoch": 2.589292848581702,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 6481
+    },
+    {
+      "epoch": 2.5896923691570115,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 6482
+    },
+    {
+      "epoch": 2.5900918897323213,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0681,
+      "step": 6483
+    },
+    {
+      "epoch": 2.5904914103076306,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0427,
+      "step": 6484
+    },
+    {
+      "epoch": 2.5908909308829404,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6485
+    },
+    {
+      "epoch": 2.59129045145825,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6486
+    },
+    {
+      "epoch": 2.59168997203356,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 6487
+    },
+    {
+      "epoch": 2.5920894926088693,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 6488
+    },
+    {
+      "epoch": 2.592489013184179,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 6489
+    },
+    {
+      "epoch": 2.5928885337594885,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 6490
+    },
+    {
+      "epoch": 2.5932880543347983,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6491
+    },
+    {
+      "epoch": 2.593687574910108,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 6492
+    },
+    {
+      "epoch": 2.5940870954854174,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 6493
+    },
+    {
+      "epoch": 2.594486616060727,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 6494
+    },
+    {
+      "epoch": 2.5948861366360365,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 6495
+    },
+    {
+      "epoch": 2.5952856572113463,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 6496
+    },
+    {
+      "epoch": 2.595685177786656,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6497
+    },
+    {
+      "epoch": 2.596084698361966,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 6498
+    },
+    {
+      "epoch": 2.596484218937275,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 6499
+    },
+    {
+      "epoch": 2.596883739512585,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 6500
+    },
+    {
+      "epoch": 2.5972832600878943,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 6501
+    },
+    {
+      "epoch": 2.597682780663204,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 6502
+    },
+    {
+      "epoch": 2.598082301238514,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 6503
+    },
+    {
+      "epoch": 2.5984818218138233,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 6504
+    },
+    {
+      "epoch": 2.598881342389133,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 6505
+    },
+    {
+      "epoch": 2.599280862964443,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 6506
+    },
+    {
+      "epoch": 2.599680383539752,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 6507
+    },
+    {
+      "epoch": 2.600079904115062,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6508
+    },
+    {
+      "epoch": 2.6004794246903717,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6509
+    },
+    {
+      "epoch": 2.600878945265681,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 6510
+    },
+    {
+      "epoch": 2.601278465840991,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6511
+    },
+    {
+      "epoch": 2.6016779864163,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 6512
+    },
+    {
+      "epoch": 2.60207750699161,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 6513
+    },
+    {
+      "epoch": 2.60247702756692,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 6514
+    },
+    {
+      "epoch": 2.6028765481422296,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6515
+    },
+    {
+      "epoch": 2.603276068717539,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 6516
+    },
+    {
+      "epoch": 2.6036755892928487,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 6517
+    },
+    {
+      "epoch": 2.604075109868158,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6518
+    },
+    {
+      "epoch": 2.604474630443468,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 6519
+    },
+    {
+      "epoch": 2.6048741510187776,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6520
+    },
+    {
+      "epoch": 2.605273671594087,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 6521
+    },
+    {
+      "epoch": 2.6056731921693967,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6522
+    },
+    {
+      "epoch": 2.606072712744706,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 6523
+    },
+    {
+      "epoch": 2.606472233320016,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 6524
+    },
+    {
+      "epoch": 2.6068717538953257,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 6525
+    },
+    {
+      "epoch": 2.6072712744706354,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6526
+    },
+    {
+      "epoch": 2.607670795045945,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 6527
+    },
+    {
+      "epoch": 2.6080703156212546,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 6528
+    },
+    {
+      "epoch": 2.608469836196564,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 6529
+    },
+    {
+      "epoch": 2.6088693567718737,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 6530
+    },
+    {
+      "epoch": 2.6092688773471835,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6531
+    },
+    {
+      "epoch": 2.6096683979224933,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 6532
+    },
+    {
+      "epoch": 2.6100679184978026,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6533
+    },
+    {
+      "epoch": 2.6104674390731124,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 6534
+    },
+    {
+      "epoch": 2.6108669596484217,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 6535
+    },
+    {
+      "epoch": 2.6112664802237315,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6536
+    },
+    {
+      "epoch": 2.6116660007990413,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 6537
+    },
+    {
+      "epoch": 2.6120655213743507,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6538
+    },
+    {
+      "epoch": 2.6124650419496604,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6539
+    },
+    {
+      "epoch": 2.61286456252497,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 6540
+    },
+    {
+      "epoch": 2.6132640831002796,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6541
+    },
+    {
+      "epoch": 2.6136636036755894,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 6542
+    },
+    {
+      "epoch": 2.614063124250899,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6543
+    },
+    {
+      "epoch": 2.6144626448262085,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 6544
+    },
+    {
+      "epoch": 2.6148621654015183,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 6545
+    },
+    {
+      "epoch": 2.6152616859768276,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6546
+    },
+    {
+      "epoch": 2.6156612065521374,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 6547
+    },
+    {
+      "epoch": 2.616060727127447,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 6548
+    },
+    {
+      "epoch": 2.6164602477027565,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6549
+    },
+    {
+      "epoch": 2.6168597682780663,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 6550
+    },
+    {
+      "epoch": 2.617259288853376,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6551
+    },
+    {
+      "epoch": 2.6176588094286855,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6552
+    },
+    {
+      "epoch": 2.6180583300039952,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 6553
+    },
+    {
+      "epoch": 2.618457850579305,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6554
+    },
+    {
+      "epoch": 2.6188573711546144,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6555
+    },
+    {
+      "epoch": 2.619256891729924,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6556
+    },
+    {
+      "epoch": 2.6196564123052335,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 6557
+    },
+    {
+      "epoch": 2.6200559328805433,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 6558
+    },
+    {
+      "epoch": 2.620455453455853,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 6559
+    },
+    {
+      "epoch": 2.620854974031163,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 6560
+    },
+    {
+      "epoch": 2.621254494606472,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 6561
+    },
+    {
+      "epoch": 2.621654015181782,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6562
+    },
+    {
+      "epoch": 2.6220535357570913,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 6563
+    },
+    {
+      "epoch": 2.622453056332401,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 6564
+    },
+    {
+      "epoch": 2.622852576907711,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6565
+    },
+    {
+      "epoch": 2.6232520974830202,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 6566
+    },
+    {
+      "epoch": 2.62365161805833,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 6567
+    },
+    {
+      "epoch": 2.6240511386336394,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 6568
+    },
+    {
+      "epoch": 2.624450659208949,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6569
+    },
+    {
+      "epoch": 2.624850179784259,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 6570
+    },
+    {
+      "epoch": 2.6252497003595687,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 6571
+    },
+    {
+      "epoch": 2.625649220934878,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 6572
+    },
+    {
+      "epoch": 2.626048741510188,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 6573
+    },
+    {
+      "epoch": 2.626448262085497,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 6574
+    },
+    {
+      "epoch": 2.626847782660807,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 6575
+    },
+    {
+      "epoch": 2.6272473032361168,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 6576
+    },
+    {
+      "epoch": 2.6276468238114266,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6577
+    },
+    {
+      "epoch": 2.628046344386736,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 6578
+    },
+    {
+      "epoch": 2.6284458649620457,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 6579
+    },
+    {
+      "epoch": 2.628845385537355,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6580
+    },
+    {
+      "epoch": 2.629244906112665,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 6581
+    },
+    {
+      "epoch": 2.6296444266879746,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 6582
+    },
+    {
+      "epoch": 2.630043947263284,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 6583
+    },
+    {
+      "epoch": 2.6304434678385937,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6584
+    },
+    {
+      "epoch": 2.630842988413903,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 6585
+    },
+    {
+      "epoch": 2.631242508989213,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6586
+    },
+    {
+      "epoch": 2.6316420295645226,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 6587
+    },
+    {
+      "epoch": 2.6320415501398324,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 6588
+    },
+    {
+      "epoch": 2.6324410707151418,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6589
+    },
+    {
+      "epoch": 2.6328405912904516,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 6590
+    },
+    {
+      "epoch": 2.633240111865761,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 6591
+    },
+    {
+      "epoch": 2.6336396324410707,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6592
+    },
+    {
+      "epoch": 2.6340391530163805,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 6593
+    },
+    {
+      "epoch": 2.63443867359169,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 6594
+    },
+    {
+      "epoch": 2.6348381941669996,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 6595
+    },
+    {
+      "epoch": 2.6352377147423094,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 6596
+    },
+    {
+      "epoch": 2.6356372353176187,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 6597
+    },
+    {
+      "epoch": 2.6360367558929285,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 6598
+    },
+    {
+      "epoch": 2.6364362764682383,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 6599
+    },
+    {
+      "epoch": 2.6368357970435476,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 6600
+    },
+    {
+      "epoch": 2.6372353176188574,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 6601
+    },
+    {
+      "epoch": 2.6376348381941668,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 6602
+    },
+    {
+      "epoch": 2.6380343587694766,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 6603
+    },
+    {
+      "epoch": 2.6384338793447863,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 6604
+    },
+    {
+      "epoch": 2.638833399920096,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6605
+    },
+    {
+      "epoch": 2.6392329204954055,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 6606
+    },
+    {
+      "epoch": 2.6396324410707153,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 6607
+    },
+    {
+      "epoch": 2.6400319616460246,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 6608
+    },
+    {
+      "epoch": 2.6404314822213344,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 6609
+    },
+    {
+      "epoch": 2.640831002796644,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 6610
+    },
+    {
+      "epoch": 2.6412305233719535,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6611
+    },
+    {
+      "epoch": 2.6416300439472633,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 6612
+    },
+    {
+      "epoch": 2.6420295645225726,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6613
+    },
+    {
+      "epoch": 2.6424290850978824,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 6614
+    },
+    {
+      "epoch": 2.642828605673192,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 6615
+    },
+    {
+      "epoch": 2.643228126248502,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 6616
+    },
+    {
+      "epoch": 2.6436276468238114,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 6617
+    },
+    {
+      "epoch": 2.644027167399121,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 6618
+    },
+    {
+      "epoch": 2.6444266879744305,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 6619
+    },
+    {
+      "epoch": 2.6448262085497403,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 6620
+    },
+    {
+      "epoch": 2.64522572912505,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 6621
+    },
+    {
+      "epoch": 2.64562524970036,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 6622
+    },
+    {
+      "epoch": 2.646024770275669,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6623
+    },
+    {
+      "epoch": 2.646424290850979,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6624
+    },
+    {
+      "epoch": 2.6468238114262883,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6625
+    },
+    {
+      "epoch": 2.647223332001598,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6626
+    },
+    {
+      "epoch": 2.647622852576908,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6627
+    },
+    {
+      "epoch": 2.6480223731522172,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6628
+    },
+    {
+      "epoch": 2.648421893727527,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 6629
+    },
+    {
+      "epoch": 2.6488214143028364,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 6630
+    },
+    {
+      "epoch": 2.649220934878146,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 6631
+    },
+    {
+      "epoch": 2.649620455453456,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 6632
+    },
+    {
+      "epoch": 2.6500199760287657,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6633
+    },
+    {
+      "epoch": 2.650419496604075,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 6634
+    },
+    {
+      "epoch": 2.650819017179385,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 6635
+    },
+    {
+      "epoch": 2.651218537754694,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 6636
+    },
+    {
+      "epoch": 2.651618058330004,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6637
+    },
+    {
+      "epoch": 2.6520175789053138,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 6638
+    },
+    {
+      "epoch": 2.652417099480623,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 6639
+    },
+    {
+      "epoch": 2.652816620055933,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 6640
+    },
+    {
+      "epoch": 2.6532161406312427,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6641
+    },
+    {
+      "epoch": 2.653615661206552,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6642
+    },
+    {
+      "epoch": 2.654015181781862,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 6643
+    },
+    {
+      "epoch": 2.6544147023571716,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 6644
+    },
+    {
+      "epoch": 2.654814222932481,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 6645
+    },
+    {
+      "epoch": 2.6552137435077907,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 6646
+    },
+    {
+      "epoch": 2.6556132640831,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6647
+    },
+    {
+      "epoch": 2.65601278465841,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 6648
+    },
+    {
+      "epoch": 2.6564123052337196,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6649
+    },
+    {
+      "epoch": 2.6568118258090294,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 6650
+    },
+    {
+      "epoch": 2.6572113463843388,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 6651
+    },
+    {
+      "epoch": 2.6576108669596485,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 6652
+    },
+    {
+      "epoch": 2.658010387534958,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 6653
+    },
+    {
+      "epoch": 2.6584099081102677,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 6654
+    },
+    {
+      "epoch": 2.6588094286855775,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 6655
+    },
+    {
+      "epoch": 2.659208949260887,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 6656
+    },
+    {
+      "epoch": 2.6596084698361966,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0708,
+      "step": 6657
+    },
+    {
+      "epoch": 2.660007990411506,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 6658
+    },
+    {
+      "epoch": 2.6604075109868157,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 6659
+    },
+    {
+      "epoch": 2.6608070315621255,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6660
+    },
+    {
+      "epoch": 2.6612065521374353,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 6661
+    },
+    {
+      "epoch": 2.6616060727127446,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 6662
+    },
+    {
+      "epoch": 2.6620055932880544,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 6663
+    },
+    {
+      "epoch": 2.6624051138633638,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 6664
+    },
+    {
+      "epoch": 2.6628046344386735,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 6665
+    },
+    {
+      "epoch": 2.6632041550139833,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0727,
+      "step": 6666
+    },
+    {
+      "epoch": 2.663603675589293,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 6667
+    },
+    {
+      "epoch": 2.6640031961646025,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 6668
+    },
+    {
+      "epoch": 2.6644027167399122,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6669
+    },
+    {
+      "epoch": 2.6648022373152216,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 6670
+    },
+    {
+      "epoch": 2.6652017578905314,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 6671
+    },
+    {
+      "epoch": 2.665601278465841,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 6672
+    },
+    {
+      "epoch": 2.6660007990411505,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6673
+    },
+    {
+      "epoch": 2.6664003196164603,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 6674
+    },
+    {
+      "epoch": 2.6667998401917696,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6675
+    },
+    {
+      "epoch": 2.6671993607670794,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6676
+    },
+    {
+      "epoch": 2.667598881342389,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 6677
+    },
+    {
+      "epoch": 2.667998401917699,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 6678
+    },
+    {
+      "epoch": 2.6683979224930083,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 6679
+    },
+    {
+      "epoch": 2.668797443068318,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 6680
+    },
+    {
+      "epoch": 2.6691969636436275,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 6681
+    },
+    {
+      "epoch": 2.6695964842189372,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 6682
+    },
+    {
+      "epoch": 2.669996004794247,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 6683
+    },
+    {
+      "epoch": 2.6703955253695564,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6684
+    },
+    {
+      "epoch": 2.670795045944866,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 6685
+    },
+    {
+      "epoch": 2.671194566520176,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6686
+    },
+    {
+      "epoch": 2.6715940870954853,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6687
+    },
+    {
+      "epoch": 2.671993607670795,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 6688
+    },
+    {
+      "epoch": 2.672393128246105,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 6689
+    },
+    {
+      "epoch": 2.672792648821414,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6690
+    },
+    {
+      "epoch": 2.673192169396724,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 6691
+    },
+    {
+      "epoch": 2.6735916899720333,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6692
+    },
+    {
+      "epoch": 2.673991210547343,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 6693
+    },
+    {
+      "epoch": 2.674390731122653,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6694
+    },
+    {
+      "epoch": 2.6747902516979627,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 6695
+    },
+    {
+      "epoch": 2.675189772273272,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 6696
+    },
+    {
+      "epoch": 2.675589292848582,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 6697
+    },
+    {
+      "epoch": 2.675988813423891,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 6698
+    },
+    {
+      "epoch": 2.676388333999201,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 6699
+    },
+    {
+      "epoch": 2.6767878545745107,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 6700
+    },
+    {
+      "epoch": 2.67718737514982,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6701
+    },
+    {
+      "epoch": 2.67758689572513,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6702
+    },
+    {
+      "epoch": 2.677986416300439,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 6703
+    },
+    {
+      "epoch": 2.678385936875749,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 6704
+    },
+    {
+      "epoch": 2.678785457451059,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 6705
+    },
+    {
+      "epoch": 2.6791849780263686,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6706
+    },
+    {
+      "epoch": 2.679584498601678,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6707
+    },
+    {
+      "epoch": 2.6799840191769877,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 6708
+    },
+    {
+      "epoch": 2.680383539752297,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 6709
+    },
+    {
+      "epoch": 2.680783060327607,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6710
+    },
+    {
+      "epoch": 2.6811825809029166,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 6711
+    },
+    {
+      "epoch": 2.6815821014782264,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 6712
+    },
+    {
+      "epoch": 2.6819816220535357,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 6713
+    },
+    {
+      "epoch": 2.6823811426288455,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 6714
+    },
+    {
+      "epoch": 2.682780663204155,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 6715
+    },
+    {
+      "epoch": 2.6831801837794647,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 6716
+    },
+    {
+      "epoch": 2.6835797043547744,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 6717
+    },
+    {
+      "epoch": 2.683979224930084,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6718
+    },
+    {
+      "epoch": 2.6843787455053936,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 6719
+    },
+    {
+      "epoch": 2.684778266080703,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0812,
+      "step": 6720
+    },
+    {
+      "epoch": 2.6851777866560127,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 6721
+    },
+    {
+      "epoch": 2.6855773072313225,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6722
+    },
+    {
+      "epoch": 2.6859768278066323,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6723
+    },
+    {
+      "epoch": 2.6863763483819416,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 6724
+    },
+    {
+      "epoch": 2.6867758689572514,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6725
+    },
+    {
+      "epoch": 2.6871753895325607,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 6726
+    },
+    {
+      "epoch": 2.6875749101078705,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6727
+    },
+    {
+      "epoch": 2.6879744306831803,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0724,
+      "step": 6728
+    },
+    {
+      "epoch": 2.6883739512584897,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 6729
+    },
+    {
+      "epoch": 2.6887734718337994,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6730
+    },
+    {
+      "epoch": 2.6891729924091092,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6731
+    },
+    {
+      "epoch": 2.6895725129844186,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 6732
+    },
+    {
+      "epoch": 2.6899720335597284,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 6733
+    },
+    {
+      "epoch": 2.690371554135038,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 6734
+    },
+    {
+      "epoch": 2.6907710747103475,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 6735
+    },
+    {
+      "epoch": 2.6911705952856573,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 6736
+    },
+    {
+      "epoch": 2.6915701158609666,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 6737
+    },
+    {
+      "epoch": 2.6919696364362764,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6738
+    },
+    {
+      "epoch": 2.692369157011586,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 6739
+    },
+    {
+      "epoch": 2.692768677586896,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 6740
+    },
+    {
+      "epoch": 2.6931681981622053,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 6741
+    },
+    {
+      "epoch": 2.693567718737515,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6742
+    },
+    {
+      "epoch": 2.6939672393128244,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6743
+    },
+    {
+      "epoch": 2.6943667598881342,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 6744
+    },
+    {
+      "epoch": 2.694766280463444,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 6745
+    },
+    {
+      "epoch": 2.6951658010387534,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0398,
+      "step": 6746
+    },
+    {
+      "epoch": 2.695565321614063,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 6747
+    },
+    {
+      "epoch": 2.6959648421893725,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 6748
+    },
+    {
+      "epoch": 2.6963643627646823,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6749
+    },
+    {
+      "epoch": 2.696763883339992,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 6750
+    },
+    {
+      "epoch": 2.697163403915302,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 6751
+    },
+    {
+      "epoch": 2.697562924490611,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 6752
+    },
+    {
+      "epoch": 2.697962445065921,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 6753
+    },
+    {
+      "epoch": 2.6983619656412303,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 6754
+    },
+    {
+      "epoch": 2.69876148621654,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 6755
+    },
+    {
+      "epoch": 2.69916100679185,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 6756
+    },
+    {
+      "epoch": 2.6995605273671597,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 6757
+    },
+    {
+      "epoch": 2.699960047942469,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 6758
+    },
+    {
+      "epoch": 2.700359568517779,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 6759
+    },
+    {
+      "epoch": 2.700759089093088,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 6760
+    },
+    {
+      "epoch": 2.701158609668398,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 6761
+    },
+    {
+      "epoch": 2.7015581302437077,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 6762
+    },
+    {
+      "epoch": 2.701957650819017,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6763
+    },
+    {
+      "epoch": 2.702357171394327,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6764
+    },
+    {
+      "epoch": 2.702756691969636,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6765
+    },
+    {
+      "epoch": 2.703156212544946,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 6766
+    },
+    {
+      "epoch": 2.7035557331202558,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 6767
+    },
+    {
+      "epoch": 2.7039552536955656,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6768
+    },
+    {
+      "epoch": 2.704354774270875,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 6769
+    },
+    {
+      "epoch": 2.7047542948461847,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 6770
+    },
+    {
+      "epoch": 2.705153815421494,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 6771
+    },
+    {
+      "epoch": 2.705553335996804,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 6772
+    },
+    {
+      "epoch": 2.7059528565721136,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0373,
+      "step": 6773
+    },
+    {
+      "epoch": 2.706352377147423,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6774
+    },
+    {
+      "epoch": 2.7067518977227327,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 6775
+    },
+    {
+      "epoch": 2.7071514182980425,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 6776
+    },
+    {
+      "epoch": 2.707550938873352,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 6777
+    },
+    {
+      "epoch": 2.7079504594486616,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6778
+    },
+    {
+      "epoch": 2.7083499800239714,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6779
+    },
+    {
+      "epoch": 2.7087495005992808,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6780
+    },
+    {
+      "epoch": 2.7091490211745906,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 6781
+    },
+    {
+      "epoch": 2.7095485417499,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 6782
+    },
+    {
+      "epoch": 2.7099480623252097,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 6783
+    },
+    {
+      "epoch": 2.7103475829005195,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 6784
+    },
+    {
+      "epoch": 2.7107471034758293,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 6785
+    },
+    {
+      "epoch": 2.7111466240511386,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 6786
+    },
+    {
+      "epoch": 2.7115461446264484,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 6787
+    },
+    {
+      "epoch": 2.7119456652017577,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 6788
+    },
+    {
+      "epoch": 2.7123451857770675,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 6789
+    },
+    {
+      "epoch": 2.7127447063523773,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 6790
+    },
+    {
+      "epoch": 2.7131442269276866,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6791
+    },
+    {
+      "epoch": 2.7135437475029964,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 6792
+    },
+    {
+      "epoch": 2.7139432680783058,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 6793
+    },
+    {
+      "epoch": 2.7143427886536156,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 6794
+    },
+    {
+      "epoch": 2.7147423092289253,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 6795
+    },
+    {
+      "epoch": 2.715141829804235,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 6796
+    },
+    {
+      "epoch": 2.7155413503795445,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 6797
+    },
+    {
+      "epoch": 2.7159408709548543,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6798
+    },
+    {
+      "epoch": 2.7163403915301636,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 6799
+    },
+    {
+      "epoch": 2.7167399121054734,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 6800
+    },
+    {
+      "epoch": 2.717139432680783,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 6801
+    },
+    {
+      "epoch": 2.717538953256093,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6802
+    },
+    {
+      "epoch": 2.7179384738314023,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 6803
+    },
+    {
+      "epoch": 2.718337994406712,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 6804
+    },
+    {
+      "epoch": 2.7187375149820214,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 6805
+    },
+    {
+      "epoch": 2.719137035557331,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 6806
+    },
+    {
+      "epoch": 2.719536556132641,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 6807
+    },
+    {
+      "epoch": 2.7199360767079503,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 6808
+    },
+    {
+      "epoch": 2.72033559728326,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6809
+    },
+    {
+      "epoch": 2.7207351178585695,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 6810
+    },
+    {
+      "epoch": 2.7211346384338793,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 6811
+    },
+    {
+      "epoch": 2.721534159009189,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 6812
+    },
+    {
+      "epoch": 2.721933679584499,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 6813
+    },
+    {
+      "epoch": 2.722333200159808,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 6814
+    },
+    {
+      "epoch": 2.722732720735118,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 6815
+    },
+    {
+      "epoch": 2.7231322413104273,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 6816
+    },
+    {
+      "epoch": 2.723531761885737,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 6817
+    },
+    {
+      "epoch": 2.723931282461047,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 6818
+    },
+    {
+      "epoch": 2.724330803036356,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 6819
+    },
+    {
+      "epoch": 2.724730323611666,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 6820
+    },
+    {
+      "epoch": 2.725129844186976,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6821
+    },
+    {
+      "epoch": 2.725529364762285,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 6822
+    },
+    {
+      "epoch": 2.725928885337595,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 6823
+    },
+    {
+      "epoch": 2.7263284059129047,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 6824
+    },
+    {
+      "epoch": 2.726727926488214,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 6825
+    },
+    {
+      "epoch": 2.727127447063524,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 6826
+    },
+    {
+      "epoch": 2.727526967638833,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 6827
+    },
+    {
+      "epoch": 2.727926488214143,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 6828
+    },
+    {
+      "epoch": 2.7283260087894528,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 6829
+    },
+    {
+      "epoch": 2.7287255293647625,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.067,
+      "step": 6830
+    },
+    {
+      "epoch": 2.729125049940072,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 6831
+    },
+    {
+      "epoch": 2.7295245705153817,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 6832
+    },
+    {
+      "epoch": 2.729924091090691,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 6833
+    },
+    {
+      "epoch": 2.730323611666001,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 6834
+    },
+    {
+      "epoch": 2.7307231322413106,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6835
+    },
+    {
+      "epoch": 2.73112265281662,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 6836
+    },
+    {
+      "epoch": 2.7315221733919297,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 6837
+    },
+    {
+      "epoch": 2.731921693967239,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6838
+    },
+    {
+      "epoch": 2.732321214542549,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6839
+    },
+    {
+      "epoch": 2.7327207351178586,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6840
+    },
+    {
+      "epoch": 2.7331202556931684,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6841
+    },
+    {
+      "epoch": 2.7335197762684778,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 6842
+    },
+    {
+      "epoch": 2.7339192968437875,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 6843
+    },
+    {
+      "epoch": 2.734318817419097,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 6844
+    },
+    {
+      "epoch": 2.7347183379944067,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 6845
+    },
+    {
+      "epoch": 2.7351178585697165,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 6846
+    },
+    {
+      "epoch": 2.7355173791450262,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 6847
+    },
+    {
+      "epoch": 2.7359168997203356,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 6848
+    },
+    {
+      "epoch": 2.7363164202956454,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 6849
+    },
+    {
+      "epoch": 2.7367159408709547,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 6850
+    },
+    {
+      "epoch": 2.7371154614462645,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6851
+    },
+    {
+      "epoch": 2.7375149820215743,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 6852
+    },
+    {
+      "epoch": 2.7379145025968836,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 6853
+    },
+    {
+      "epoch": 2.7383140231721934,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6854
+    },
+    {
+      "epoch": 2.7387135437475028,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 6855
+    },
+    {
+      "epoch": 2.7391130643228125,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 6856
+    },
+    {
+      "epoch": 2.7395125848981223,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 6857
+    },
+    {
+      "epoch": 2.739912105473432,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 6858
+    },
+    {
+      "epoch": 2.7403116260487415,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 6859
+    },
+    {
+      "epoch": 2.7407111466240512,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 6860
+    },
+    {
+      "epoch": 2.7411106671993606,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 6861
+    },
+    {
+      "epoch": 2.7415101877746704,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 6862
+    },
+    {
+      "epoch": 2.74190970834998,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 6863
+    },
+    {
+      "epoch": 2.7423092289252895,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 6864
+    },
+    {
+      "epoch": 2.7427087495005993,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6865
+    },
+    {
+      "epoch": 2.743108270075909,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 6866
+    },
+    {
+      "epoch": 2.7435077906512184,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 6867
+    },
+    {
+      "epoch": 2.743907311226528,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 6868
+    },
+    {
+      "epoch": 2.744306831801838,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 6869
+    },
+    {
+      "epoch": 2.7447063523771473,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 6870
+    },
+    {
+      "epoch": 2.745105872952457,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 6871
+    },
+    {
+      "epoch": 2.7455053935277665,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 6872
+    },
+    {
+      "epoch": 2.7459049141030762,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 6873
+    },
+    {
+      "epoch": 2.746304434678386,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0717,
+      "step": 6874
+    },
+    {
+      "epoch": 2.746703955253696,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 6875
+    },
+    {
+      "epoch": 2.747103475829005,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 6876
+    },
+    {
+      "epoch": 2.747502996404315,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 6877
+    },
+    {
+      "epoch": 2.7479025169796243,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 6878
+    },
+    {
+      "epoch": 2.748302037554934,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 6879
+    },
+    {
+      "epoch": 2.748701558130244,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 6880
+    },
+    {
+      "epoch": 2.749101078705553,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 6881
+    },
+    {
+      "epoch": 2.749500599280863,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 6882
+    },
+    {
+      "epoch": 2.7499001198561723,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 6883
+    },
+    {
+      "epoch": 2.750299640431482,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6884
+    },
+    {
+      "epoch": 2.750699161006792,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 6885
+    },
+    {
+      "epoch": 2.7510986815821017,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6886
+    },
+    {
+      "epoch": 2.751498202157411,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 6887
+    },
+    {
+      "epoch": 2.751897722732721,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 6888
+    },
+    {
+      "epoch": 2.75229724330803,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 6889
+    },
+    {
+      "epoch": 2.75269676388334,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 6890
+    },
+    {
+      "epoch": 2.7530962844586497,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 6891
+    },
+    {
+      "epoch": 2.7534958050339595,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 6892
+    },
+    {
+      "epoch": 2.753895325609269,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 6893
+    },
+    {
+      "epoch": 2.7542948461845786,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 6894
+    },
+    {
+      "epoch": 2.754694366759888,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6895
+    },
+    {
+      "epoch": 2.7550938873351978,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 6896
+    },
+    {
+      "epoch": 2.7554934079105076,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6897
+    },
+    {
+      "epoch": 2.755892928485817,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 6898
+    },
+    {
+      "epoch": 2.7562924490611267,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 6899
+    },
+    {
+      "epoch": 2.756691969636436,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 6900
+    },
+    {
+      "epoch": 2.757091490211746,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 6901
+    },
+    {
+      "epoch": 2.7574910107870556,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6902
+    },
+    {
+      "epoch": 2.7578905313623654,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 6903
+    },
+    {
+      "epoch": 2.7582900519376747,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 6904
+    },
+    {
+      "epoch": 2.7586895725129845,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 6905
+    },
+    {
+      "epoch": 2.759089093088294,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 6906
+    },
+    {
+      "epoch": 2.7594886136636037,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 6907
+    },
+    {
+      "epoch": 2.7598881342389134,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 6908
+    },
+    {
+      "epoch": 2.760287654814223,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 6909
+    },
+    {
+      "epoch": 2.7606871753895326,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 6910
+    },
+    {
+      "epoch": 2.7610866959648424,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0411,
+      "step": 6911
+    },
+    {
+      "epoch": 2.7614862165401517,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6912
+    },
+    {
+      "epoch": 2.7618857371154615,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 6913
+    },
+    {
+      "epoch": 2.7622852576907713,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 6914
+    },
+    {
+      "epoch": 2.7626847782660806,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 6915
+    },
+    {
+      "epoch": 2.7630842988413904,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 6916
+    },
+    {
+      "epoch": 2.7634838194166997,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0699,
+      "step": 6917
+    },
+    {
+      "epoch": 2.7638833399920095,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 6918
+    },
+    {
+      "epoch": 2.7642828605673193,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 6919
+    },
+    {
+      "epoch": 2.764682381142629,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 6920
+    },
+    {
+      "epoch": 2.7650819017179384,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 6921
+    },
+    {
+      "epoch": 2.7654814222932482,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 6922
+    },
+    {
+      "epoch": 2.7658809428685576,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 6923
+    },
+    {
+      "epoch": 2.7662804634438674,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6924
+    },
+    {
+      "epoch": 2.766679984019177,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 6925
+    },
+    {
+      "epoch": 2.7670795045944865,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 6926
+    },
+    {
+      "epoch": 2.7674790251697963,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 6927
+    },
+    {
+      "epoch": 2.7678785457451056,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 6928
+    },
+    {
+      "epoch": 2.7682780663204154,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 6929
+    },
+    {
+      "epoch": 2.768677586895725,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6930
+    },
+    {
+      "epoch": 2.769077107471035,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 6931
+    },
+    {
+      "epoch": 2.7694766280463443,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 6932
+    },
+    {
+      "epoch": 2.769876148621654,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 6933
+    },
+    {
+      "epoch": 2.7702756691969634,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 6934
+    },
+    {
+      "epoch": 2.7706751897722732,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 6935
+    },
+    {
+      "epoch": 2.771074710347583,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 6936
+    },
+    {
+      "epoch": 2.771474230922893,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 6937
+    },
+    {
+      "epoch": 2.771873751498202,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 6938
+    },
+    {
+      "epoch": 2.772273272073512,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 6939
+    },
+    {
+      "epoch": 2.7726727926488213,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 6940
+    },
+    {
+      "epoch": 2.773072313224131,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 6941
+    },
+    {
+      "epoch": 2.773471833799441,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 6942
+    },
+    {
+      "epoch": 2.77387135437475,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 6943
+    },
+    {
+      "epoch": 2.77427087495006,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 6944
+    },
+    {
+      "epoch": 2.7746703955253693,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 6945
+    },
+    {
+      "epoch": 2.775069916100679,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 6946
+    },
+    {
+      "epoch": 2.775469436675989,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 6947
+    },
+    {
+      "epoch": 2.7758689572512987,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 6948
+    },
+    {
+      "epoch": 2.776268477826608,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 6949
+    },
+    {
+      "epoch": 2.776667998401918,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 6950
+    },
+    {
+      "epoch": 2.777067518977227,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 6951
+    },
+    {
+      "epoch": 2.777467039552537,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 6952
+    },
+    {
+      "epoch": 2.7778665601278467,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 6953
+    },
+    {
+      "epoch": 2.778266080703156,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 6954
+    },
+    {
+      "epoch": 2.778665601278466,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 6955
+    },
+    {
+      "epoch": 2.7790651218537756,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 6956
+    },
+    {
+      "epoch": 2.779464642429085,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 6957
+    },
+    {
+      "epoch": 2.7798641630043948,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 6958
+    },
+    {
+      "epoch": 2.7802636835797045,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 6959
+    },
+    {
+      "epoch": 2.780663204155014,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 6960
+    },
+    {
+      "epoch": 2.7810627247303237,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 6961
+    },
+    {
+      "epoch": 2.781462245305633,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 6962
+    },
+    {
+      "epoch": 2.781861765880943,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 6963
+    },
+    {
+      "epoch": 2.7822612864562526,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6964
+    },
+    {
+      "epoch": 2.7826608070315624,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 6965
+    },
+    {
+      "epoch": 2.7830603276068717,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 6966
+    },
+    {
+      "epoch": 2.7834598481821815,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 6967
+    },
+    {
+      "epoch": 2.783859368757491,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 6968
+    },
+    {
+      "epoch": 2.7842588893328006,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 6969
+    },
+    {
+      "epoch": 2.7846584099081104,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 6970
+    },
+    {
+      "epoch": 2.7850579304834198,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 6971
+    },
+    {
+      "epoch": 2.7854574510587295,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 6972
+    },
+    {
+      "epoch": 2.785856971634039,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 6973
+    },
+    {
+      "epoch": 2.7862564922093487,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0382,
+      "step": 6974
+    },
+    {
+      "epoch": 2.7866560127846585,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0364,
+      "step": 6975
+    },
+    {
+      "epoch": 2.7870555333599683,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 6976
+    },
+    {
+      "epoch": 2.7874550539352776,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 6977
+    },
+    {
+      "epoch": 2.7878545745105874,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 6978
+    },
+    {
+      "epoch": 2.7882540950858967,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 6979
+    },
+    {
+      "epoch": 2.7886536156612065,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 6980
+    },
+    {
+      "epoch": 2.7890531362365163,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 6981
+    },
+    {
+      "epoch": 2.789452656811826,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 6982
+    },
+    {
+      "epoch": 2.7898521773871354,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 6983
+    },
+    {
+      "epoch": 2.790251697962445,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 6984
+    },
+    {
+      "epoch": 2.7906512185377546,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 6985
+    },
+    {
+      "epoch": 2.7910507391130643,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 6986
+    },
+    {
+      "epoch": 2.791450259688374,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 6987
+    },
+    {
+      "epoch": 2.7918497802636835,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 6988
+    },
+    {
+      "epoch": 2.7922493008389933,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 6989
+    },
+    {
+      "epoch": 2.7926488214143026,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 6990
+    },
+    {
+      "epoch": 2.7930483419896124,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 6991
+    },
+    {
+      "epoch": 2.793447862564922,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 6992
+    },
+    {
+      "epoch": 2.793847383140232,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 6993
+    },
+    {
+      "epoch": 2.7942469037155413,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0737,
+      "step": 6994
+    },
+    {
+      "epoch": 2.794646424290851,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0734,
+      "step": 6995
+    },
+    {
+      "epoch": 2.7950459448661604,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 6996
+    },
+    {
+      "epoch": 2.79544546544147,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 6997
+    },
+    {
+      "epoch": 2.79584498601678,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 6998
+    },
+    {
+      "epoch": 2.7962445065920893,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 6999
+    },
+    {
+      "epoch": 2.796644027167399,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 7000
+    },
+    {
+      "epoch": 2.797043547742709,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7001
+    },
+    {
+      "epoch": 2.7974430683180183,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.036,
+      "step": 7002
+    },
+    {
+      "epoch": 2.797842588893328,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 7003
+    },
+    {
+      "epoch": 2.798242109468638,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7004
+    },
+    {
+      "epoch": 2.798641630043947,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 7005
+    },
+    {
+      "epoch": 2.799041150619257,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 7006
+    },
+    {
+      "epoch": 2.7994406711945663,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 7007
+    },
+    {
+      "epoch": 2.799840191769876,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 7008
+    },
+    {
+      "epoch": 2.800239712345186,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 7009
+    },
+    {
+      "epoch": 2.8006392329204957,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7010
+    },
+    {
+      "epoch": 2.801038753495805,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 7011
+    },
+    {
+      "epoch": 2.801438274071115,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 7012
+    },
+    {
+      "epoch": 2.801837794646424,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7013
+    },
+    {
+      "epoch": 2.802237315221734,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 7014
+    },
+    {
+      "epoch": 2.8026368357970437,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 7015
+    },
+    {
+      "epoch": 2.803036356372353,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 7016
+    },
+    {
+      "epoch": 2.803435876947663,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 7017
+    },
+    {
+      "epoch": 2.803835397522972,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 7018
+    },
+    {
+      "epoch": 2.804234918098282,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 7019
+    },
+    {
+      "epoch": 2.8046344386735917,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 7020
+    },
+    {
+      "epoch": 2.8050339592489015,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 7021
+    },
+    {
+      "epoch": 2.805433479824211,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 7022
+    },
+    {
+      "epoch": 2.8058330003995207,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 7023
+    },
+    {
+      "epoch": 2.80623252097483,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0763,
+      "step": 7024
+    },
+    {
+      "epoch": 2.80663204155014,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7025
+    },
+    {
+      "epoch": 2.8070315621254496,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7026
+    },
+    {
+      "epoch": 2.8074310827007594,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 7027
+    },
+    {
+      "epoch": 2.8078306032760687,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 7028
+    },
+    {
+      "epoch": 2.8082301238513785,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0427,
+      "step": 7029
+    },
+    {
+      "epoch": 2.808629644426688,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7030
+    },
+    {
+      "epoch": 2.8090291650019976,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 7031
+    },
+    {
+      "epoch": 2.8094286855773074,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 7032
+    },
+    {
+      "epoch": 2.8098282061526167,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0754,
+      "step": 7033
+    },
+    {
+      "epoch": 2.8102277267279265,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 7034
+    },
+    {
+      "epoch": 2.810627247303236,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 7035
+    },
+    {
+      "epoch": 2.8110267678785457,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 7036
+    },
+    {
+      "epoch": 2.8114262884538554,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 7037
+    },
+    {
+      "epoch": 2.8118258090291652,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 7038
+    },
+    {
+      "epoch": 2.8122253296044746,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 7039
+    },
+    {
+      "epoch": 2.8126248501797844,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 7040
+    },
+    {
+      "epoch": 2.8130243707550937,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 7041
+    },
+    {
+      "epoch": 2.8134238913304035,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 7042
+    },
+    {
+      "epoch": 2.8138234119057133,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 7043
+    },
+    {
+      "epoch": 2.8142229324810226,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 7044
+    },
+    {
+      "epoch": 2.8146224530563324,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 7045
+    },
+    {
+      "epoch": 2.815021973631642,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 7046
+    },
+    {
+      "epoch": 2.8154214942069515,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 7047
+    },
+    {
+      "epoch": 2.8158210147822613,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7048
+    },
+    {
+      "epoch": 2.816220535357571,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 7049
+    },
+    {
+      "epoch": 2.8166200559328805,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 7050
+    },
+    {
+      "epoch": 2.8170195765081902,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 7051
+    },
+    {
+      "epoch": 2.8174190970834996,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 7052
+    },
+    {
+      "epoch": 2.8178186176588094,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0696,
+      "step": 7053
+    },
+    {
+      "epoch": 2.818218138234119,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 7054
+    },
+    {
+      "epoch": 2.818617658809429,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 7055
+    },
+    {
+      "epoch": 2.8190171793847383,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 7056
+    },
+    {
+      "epoch": 2.819416699960048,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7057
+    },
+    {
+      "epoch": 2.8198162205353574,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 7058
+    },
+    {
+      "epoch": 2.820215741110667,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 7059
+    },
+    {
+      "epoch": 2.820615261685977,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 7060
+    },
+    {
+      "epoch": 2.8210147822612863,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7061
+    },
+    {
+      "epoch": 2.821414302836596,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 7062
+    },
+    {
+      "epoch": 2.8218138234119055,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 7063
+    },
+    {
+      "epoch": 2.8222133439872152,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0666,
+      "step": 7064
+    },
+    {
+      "epoch": 2.822612864562525,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 7065
+    },
+    {
+      "epoch": 2.823012385137835,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 7066
+    },
+    {
+      "epoch": 2.823411905713144,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7067
+    },
+    {
+      "epoch": 2.823811426288454,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 7068
+    },
+    {
+      "epoch": 2.8242109468637633,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7069
+    },
+    {
+      "epoch": 2.824610467439073,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7070
+    },
+    {
+      "epoch": 2.825009988014383,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7071
+    },
+    {
+      "epoch": 2.8254095085896926,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 7072
+    },
+    {
+      "epoch": 2.825809029165002,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7073
+    },
+    {
+      "epoch": 2.8262085497403118,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0672,
+      "step": 7074
+    },
+    {
+      "epoch": 2.826608070315621,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 7075
+    },
+    {
+      "epoch": 2.827007590890931,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 7076
+    },
+    {
+      "epoch": 2.8274071114662407,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 7077
+    },
+    {
+      "epoch": 2.82780663204155,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 7078
+    },
+    {
+      "epoch": 2.82820615261686,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 7079
+    },
+    {
+      "epoch": 2.828605673192169,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 7080
+    },
+    {
+      "epoch": 2.829005193767479,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 7081
+    },
+    {
+      "epoch": 2.8294047143427887,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0732,
+      "step": 7082
+    },
+    {
+      "epoch": 2.8298042349180985,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 7083
+    },
+    {
+      "epoch": 2.830203755493408,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 7084
+    },
+    {
+      "epoch": 2.8306032760687176,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7085
+    },
+    {
+      "epoch": 2.831002796644027,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 7086
+    },
+    {
+      "epoch": 2.8314023172193368,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 7087
+    },
+    {
+      "epoch": 2.8318018377946466,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 7088
+    },
+    {
+      "epoch": 2.832201358369956,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 7089
+    },
+    {
+      "epoch": 2.8326008789452657,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 7090
+    },
+    {
+      "epoch": 2.8330003995205755,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7091
+    },
+    {
+      "epoch": 2.833399920095885,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 7092
+    },
+    {
+      "epoch": 2.8337994406711946,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 7093
+    },
+    {
+      "epoch": 2.8341989612465044,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 7094
+    },
+    {
+      "epoch": 2.8345984818218137,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 7095
+    },
+    {
+      "epoch": 2.8349980023971235,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 7096
+    },
+    {
+      "epoch": 2.835397522972433,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 7097
+    },
+    {
+      "epoch": 2.8357970435477426,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 7098
+    },
+    {
+      "epoch": 2.8361965641230524,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 7099
+    },
+    {
+      "epoch": 2.836596084698362,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 7100
+    },
+    {
+      "epoch": 2.8369956052736716,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 7101
+    },
+    {
+      "epoch": 2.8373951258489813,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7102
+    },
+    {
+      "epoch": 2.8377946464242907,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 7103
+    },
+    {
+      "epoch": 2.8381941669996005,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 7104
+    },
+    {
+      "epoch": 2.8385936875749103,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 7105
+    },
+    {
+      "epoch": 2.8389932081502196,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 7106
+    },
+    {
+      "epoch": 2.8393927287255294,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 7107
+    },
+    {
+      "epoch": 2.8397922493008387,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 7108
+    },
+    {
+      "epoch": 2.8401917698761485,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 7109
+    },
+    {
+      "epoch": 2.8405912904514583,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 7110
+    },
+    {
+      "epoch": 2.840990811026768,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 7111
+    },
+    {
+      "epoch": 2.8413903316020774,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 7112
+    },
+    {
+      "epoch": 2.841789852177387,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 7113
+    },
+    {
+      "epoch": 2.8421893727526966,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 7114
+    },
+    {
+      "epoch": 2.8425888933280063,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 7115
+    },
+    {
+      "epoch": 2.842988413903316,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 7116
+    },
+    {
+      "epoch": 2.843387934478626,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 7117
+    },
+    {
+      "epoch": 2.8437874550539353,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 7118
+    },
+    {
+      "epoch": 2.844186975629245,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7119
+    },
+    {
+      "epoch": 2.8445864962045544,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 7120
+    },
+    {
+      "epoch": 2.844986016779864,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 7121
+    },
+    {
+      "epoch": 2.845385537355174,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 7122
+    },
+    {
+      "epoch": 2.8457850579304833,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 7123
+    },
+    {
+      "epoch": 2.846184578505793,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 7124
+    },
+    {
+      "epoch": 2.8465840990811024,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 7125
+    },
+    {
+      "epoch": 2.8469836196564122,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7126
+    },
+    {
+      "epoch": 2.847383140231722,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 7127
+    },
+    {
+      "epoch": 2.847782660807032,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7128
+    },
+    {
+      "epoch": 2.848182181382341,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0389,
+      "step": 7129
+    },
+    {
+      "epoch": 2.848581701957651,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7130
+    },
+    {
+      "epoch": 2.8489812225329603,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 7131
+    },
+    {
+      "epoch": 2.84938074310827,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 7132
+    },
+    {
+      "epoch": 2.84978026368358,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 7133
+    },
+    {
+      "epoch": 2.850179784258889,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 7134
+    },
+    {
+      "epoch": 2.850579304834199,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 7135
+    },
+    {
+      "epoch": 2.8509788254095088,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7136
+    },
+    {
+      "epoch": 2.851378345984818,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7137
+    },
+    {
+      "epoch": 2.851777866560128,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 7138
+    },
+    {
+      "epoch": 2.8521773871354377,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 7139
+    },
+    {
+      "epoch": 2.852576907710747,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 7140
+    },
+    {
+      "epoch": 2.852976428286057,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 7141
+    },
+    {
+      "epoch": 2.853375948861366,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7142
+    },
+    {
+      "epoch": 2.853775469436676,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7143
+    },
+    {
+      "epoch": 2.8541749900119857,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7144
+    },
+    {
+      "epoch": 2.8545745105872955,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 7145
+    },
+    {
+      "epoch": 2.854974031162605,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 7146
+    },
+    {
+      "epoch": 2.8553735517379146,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0638,
+      "step": 7147
+    },
+    {
+      "epoch": 2.855773072313224,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7148
+    },
+    {
+      "epoch": 2.8561725928885338,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 7149
+    },
+    {
+      "epoch": 2.8565721134638435,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7150
+    },
+    {
+      "epoch": 2.856971634039153,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 7151
+    },
+    {
+      "epoch": 2.8573711546144627,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 7152
+    },
+    {
+      "epoch": 2.857770675189772,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7153
+    },
+    {
+      "epoch": 2.858170195765082,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 7154
+    },
+    {
+      "epoch": 2.8585697163403916,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 7155
+    },
+    {
+      "epoch": 2.8589692369157014,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 7156
+    },
+    {
+      "epoch": 2.8593687574910107,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 7157
+    },
+    {
+      "epoch": 2.8597682780663205,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 7158
+    },
+    {
+      "epoch": 2.86016779864163,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 7159
+    },
+    {
+      "epoch": 2.8605673192169396,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 7160
+    },
+    {
+      "epoch": 2.8609668397922494,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 7161
+    },
+    {
+      "epoch": 2.861366360367559,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 7162
+    },
+    {
+      "epoch": 2.8617658809428685,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 7163
+    },
+    {
+      "epoch": 2.8621654015181783,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7164
+    },
+    {
+      "epoch": 2.8625649220934877,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 7165
+    },
+    {
+      "epoch": 2.8629644426687975,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0693,
+      "step": 7166
+    },
+    {
+      "epoch": 2.8633639632441072,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7167
+    },
+    {
+      "epoch": 2.8637634838194166,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 7168
+    },
+    {
+      "epoch": 2.8641630043947264,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7169
+    },
+    {
+      "epoch": 2.8645625249700357,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 7170
+    },
+    {
+      "epoch": 2.8649620455453455,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 7171
+    },
+    {
+      "epoch": 2.8653615661206553,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0614,
+      "step": 7172
+    },
+    {
+      "epoch": 2.865761086695965,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7173
+    },
+    {
+      "epoch": 2.8661606072712744,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 7174
+    },
+    {
+      "epoch": 2.866560127846584,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 7175
+    },
+    {
+      "epoch": 2.8669596484218935,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7176
+    },
+    {
+      "epoch": 2.8673591689972033,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7177
+    },
+    {
+      "epoch": 2.867758689572513,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 7178
+    },
+    {
+      "epoch": 2.8681582101478225,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 7179
+    },
+    {
+      "epoch": 2.8685577307231322,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 7180
+    },
+    {
+      "epoch": 2.868957251298442,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 7181
+    },
+    {
+      "epoch": 2.8693567718737514,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 7182
+    },
+    {
+      "epoch": 2.869756292449061,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7183
+    },
+    {
+      "epoch": 2.870155813024371,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 7184
+    },
+    {
+      "epoch": 2.8705553335996803,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7185
+    },
+    {
+      "epoch": 2.87095485417499,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 7186
+    },
+    {
+      "epoch": 2.8713543747502994,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7187
+    },
+    {
+      "epoch": 2.871753895325609,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0706,
+      "step": 7188
+    },
+    {
+      "epoch": 2.872153415900919,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7189
+    },
+    {
+      "epoch": 2.872552936476229,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 7190
+    },
+    {
+      "epoch": 2.872952457051538,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7191
+    },
+    {
+      "epoch": 2.873351977626848,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 7192
+    },
+    {
+      "epoch": 2.8737514982021573,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7193
+    },
+    {
+      "epoch": 2.874151018777467,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 7194
+    },
+    {
+      "epoch": 2.874550539352777,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 7195
+    },
+    {
+      "epoch": 2.874950059928086,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 7196
+    },
+    {
+      "epoch": 2.875349580503396,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 7197
+    },
+    {
+      "epoch": 2.8757491010787053,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7198
+    },
+    {
+      "epoch": 2.876148621654015,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 7199
+    },
+    {
+      "epoch": 2.876548142229325,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0425,
+      "step": 7200
+    },
+    {
+      "epoch": 2.8769476628046347,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7201
+    },
+    {
+      "epoch": 2.877347183379944,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 7202
+    },
+    {
+      "epoch": 2.877746703955254,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 7203
+    },
+    {
+      "epoch": 2.878146224530563,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7204
+    },
+    {
+      "epoch": 2.878545745105873,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7205
+    },
+    {
+      "epoch": 2.8789452656811827,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 7206
+    },
+    {
+      "epoch": 2.8793447862564925,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 7207
+    },
+    {
+      "epoch": 2.879744306831802,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0679,
+      "step": 7208
+    },
+    {
+      "epoch": 2.8801438274071116,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7209
+    },
+    {
+      "epoch": 2.880543347982421,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 7210
+    },
+    {
+      "epoch": 2.8809428685577307,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 7211
+    },
+    {
+      "epoch": 2.8813423891330405,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0405,
+      "step": 7212
+    },
+    {
+      "epoch": 2.88174190970835,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 7213
+    },
+    {
+      "epoch": 2.8821414302836597,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 7214
+    },
+    {
+      "epoch": 2.882540950858969,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7215
+    },
+    {
+      "epoch": 2.882940471434279,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 7216
+    },
+    {
+      "epoch": 2.8833399920095886,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 7217
+    },
+    {
+      "epoch": 2.8837395125848984,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0408,
+      "step": 7218
+    },
+    {
+      "epoch": 2.8841390331602077,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 7219
+    },
+    {
+      "epoch": 2.8845385537355175,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7220
+    },
+    {
+      "epoch": 2.884938074310827,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 7221
+    },
+    {
+      "epoch": 2.8853375948861366,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 7222
+    },
+    {
+      "epoch": 2.8857371154614464,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 7223
+    },
+    {
+      "epoch": 2.8861366360367557,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 7224
+    },
+    {
+      "epoch": 2.8865361566120655,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 7225
+    },
+    {
+      "epoch": 2.8869356771873753,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 7226
+    },
+    {
+      "epoch": 2.8873351977626847,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7227
+    },
+    {
+      "epoch": 2.8877347183379944,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 7228
+    },
+    {
+      "epoch": 2.8881342389133042,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 7229
+    },
+    {
+      "epoch": 2.8885337594886136,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 7230
+    },
+    {
+      "epoch": 2.8889332800639234,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0408,
+      "step": 7231
+    },
+    {
+      "epoch": 2.8893328006392327,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7232
+    },
+    {
+      "epoch": 2.8897323212145425,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0704,
+      "step": 7233
+    },
+    {
+      "epoch": 2.8901318417898523,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 7234
+    },
+    {
+      "epoch": 2.890531362365162,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7235
+    },
+    {
+      "epoch": 2.8909308829404714,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 7236
+    },
+    {
+      "epoch": 2.891330403515781,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 7237
+    },
+    {
+      "epoch": 2.8917299240910905,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 7238
+    },
+    {
+      "epoch": 2.8921294446664003,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7239
+    },
+    {
+      "epoch": 2.89252896524171,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 7240
+    },
+    {
+      "epoch": 2.8929284858170194,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 7241
+    },
+    {
+      "epoch": 2.8933280063923292,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7242
+    },
+    {
+      "epoch": 2.8937275269676386,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 7243
+    },
+    {
+      "epoch": 2.8941270475429484,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 7244
+    },
+    {
+      "epoch": 2.894526568118258,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7245
+    },
+    {
+      "epoch": 2.894926088693568,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 7246
+    },
+    {
+      "epoch": 2.8953256092688773,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 7247
+    },
+    {
+      "epoch": 2.895725129844187,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 7248
+    },
+    {
+      "epoch": 2.8961246504194964,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7249
+    },
+    {
+      "epoch": 2.896524170994806,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7250
+    },
+    {
+      "epoch": 2.896923691570116,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 7251
+    },
+    {
+      "epoch": 2.8973232121454258,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7252
+    },
+    {
+      "epoch": 2.897722732720735,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 7253
+    },
+    {
+      "epoch": 2.898122253296045,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7254
+    },
+    {
+      "epoch": 2.8985217738713542,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 7255
+    },
+    {
+      "epoch": 2.898921294446664,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 7256
+    },
+    {
+      "epoch": 2.899320815021974,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7257
+    },
+    {
+      "epoch": 2.899720335597283,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 7258
+    },
+    {
+      "epoch": 2.900119856172593,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7259
+    },
+    {
+      "epoch": 2.9005193767479023,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7260
+    },
+    {
+      "epoch": 2.900918897323212,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 7261
+    },
+    {
+      "epoch": 2.901318417898522,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7262
+    },
+    {
+      "epoch": 2.9017179384738316,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 7263
+    },
+    {
+      "epoch": 2.902117459049141,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 7264
+    },
+    {
+      "epoch": 2.9025169796244508,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 7265
+    },
+    {
+      "epoch": 2.90291650019976,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 7266
+    },
+    {
+      "epoch": 2.90331602077507,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0368,
+      "step": 7267
+    },
+    {
+      "epoch": 2.9037155413503797,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 7268
+    },
+    {
+      "epoch": 2.904115061925689,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 7269
+    },
+    {
+      "epoch": 2.904514582500999,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7270
+    },
+    {
+      "epoch": 2.9049141030763086,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 7271
+    },
+    {
+      "epoch": 2.905313623651618,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 7272
+    },
+    {
+      "epoch": 2.9057131442269277,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 7273
+    },
+    {
+      "epoch": 2.9061126648022375,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 7274
+    },
+    {
+      "epoch": 2.906512185377547,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 7275
+    },
+    {
+      "epoch": 2.9069117059528566,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 7276
+    },
+    {
+      "epoch": 2.907311226528166,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7277
+    },
+    {
+      "epoch": 2.9077107471034758,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 7278
+    },
+    {
+      "epoch": 2.9081102676787856,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 7279
+    },
+    {
+      "epoch": 2.9085097882540953,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 7280
+    },
+    {
+      "epoch": 2.9089093088294047,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 7281
+    },
+    {
+      "epoch": 2.9093088294047145,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7282
+    },
+    {
+      "epoch": 2.909708349980024,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7283
+    },
+    {
+      "epoch": 2.9101078705553336,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 7284
+    },
+    {
+      "epoch": 2.9105073911306434,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 7285
+    },
+    {
+      "epoch": 2.9109069117059527,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7286
+    },
+    {
+      "epoch": 2.9113064322812625,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 7287
+    },
+    {
+      "epoch": 2.911705952856572,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 7288
+    },
+    {
+      "epoch": 2.9121054734318816,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 7289
+    },
+    {
+      "epoch": 2.9125049940071914,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 7290
+    },
+    {
+      "epoch": 2.912904514582501,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 7291
+    },
+    {
+      "epoch": 2.9133040351578106,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 7292
+    },
+    {
+      "epoch": 2.9137035557331203,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 7293
+    },
+    {
+      "epoch": 2.9141030763084297,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 7294
+    },
+    {
+      "epoch": 2.9145025968837395,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 7295
+    },
+    {
+      "epoch": 2.9149021174590493,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7296
+    },
+    {
+      "epoch": 2.915301638034359,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7297
+    },
+    {
+      "epoch": 2.9157011586096684,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 7298
+    },
+    {
+      "epoch": 2.916100679184978,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 7299
+    },
+    {
+      "epoch": 2.9165001997602875,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 7300
+    },
+    {
+      "epoch": 2.9168997203355973,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 7301
+    },
+    {
+      "epoch": 2.917299240910907,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7302
+    },
+    {
+      "epoch": 2.9176987614862164,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 7303
+    },
+    {
+      "epoch": 2.918098282061526,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 7304
+    },
+    {
+      "epoch": 2.9184978026368356,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7305
+    },
+    {
+      "epoch": 2.9188973232121453,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 7306
+    },
+    {
+      "epoch": 2.919296843787455,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 7307
+    },
+    {
+      "epoch": 2.919696364362765,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 7308
+    },
+    {
+      "epoch": 2.9200958849380743,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0687,
+      "step": 7309
+    },
+    {
+      "epoch": 2.920495405513384,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 7310
+    },
+    {
+      "epoch": 2.9208949260886934,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0445,
+      "step": 7311
+    },
+    {
+      "epoch": 2.921294446664003,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 7312
+    },
+    {
+      "epoch": 2.921693967239313,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 7313
+    },
+    {
+      "epoch": 2.9220934878146223,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 7314
+    },
+    {
+      "epoch": 2.922493008389932,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7315
+    },
+    {
+      "epoch": 2.922892528965242,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 7316
+    },
+    {
+      "epoch": 2.923292049540551,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 7317
+    },
+    {
+      "epoch": 2.923691570115861,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 7318
+    },
+    {
+      "epoch": 2.924091090691171,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 7319
+    },
+    {
+      "epoch": 2.92449061126648,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 7320
+    },
+    {
+      "epoch": 2.92489013184179,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 7321
+    },
+    {
+      "epoch": 2.9252896524170993,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7322
+    },
+    {
+      "epoch": 2.925689172992409,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 7323
+    },
+    {
+      "epoch": 2.926088693567719,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 7324
+    },
+    {
+      "epoch": 2.9264882141430286,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 7325
+    },
+    {
+      "epoch": 2.926887734718338,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 7326
+    },
+    {
+      "epoch": 2.9272872552936477,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7327
+    },
+    {
+      "epoch": 2.927686775868957,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 7328
+    },
+    {
+      "epoch": 2.928086296444267,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 7329
+    },
+    {
+      "epoch": 2.9284858170195767,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 7330
+    },
+    {
+      "epoch": 2.928885337594886,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 7331
+    },
+    {
+      "epoch": 2.929284858170196,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 7332
+    },
+    {
+      "epoch": 2.929684378745505,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 7333
+    },
+    {
+      "epoch": 2.930083899320815,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7334
+    },
+    {
+      "epoch": 2.9304834198961247,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7335
+    },
+    {
+      "epoch": 2.9308829404714345,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 7336
+    },
+    {
+      "epoch": 2.931282461046744,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 7337
+    },
+    {
+      "epoch": 2.9316819816220536,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 7338
+    },
+    {
+      "epoch": 2.932081502197363,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 7339
+    },
+    {
+      "epoch": 2.9324810227726728,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 7340
+    },
+    {
+      "epoch": 2.9328805433479825,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 7341
+    },
+    {
+      "epoch": 2.9332800639232923,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 7342
+    },
+    {
+      "epoch": 2.9336795844986017,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7343
+    },
+    {
+      "epoch": 2.9340791050739115,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 7344
+    },
+    {
+      "epoch": 2.934478625649221,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7345
+    },
+    {
+      "epoch": 2.9348781462245306,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7346
+    },
+    {
+      "epoch": 2.9352776667998404,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 7347
+    },
+    {
+      "epoch": 2.9356771873751497,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7348
+    },
+    {
+      "epoch": 2.9360767079504595,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 7349
+    },
+    {
+      "epoch": 2.936476228525769,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 7350
+    },
+    {
+      "epoch": 2.9368757491010786,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 7351
+    },
+    {
+      "epoch": 2.9372752696763884,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 7352
+    },
+    {
+      "epoch": 2.937674790251698,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 7353
+    },
+    {
+      "epoch": 2.9380743108270075,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 7354
+    },
+    {
+      "epoch": 2.9384738314023173,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 7355
+    },
+    {
+      "epoch": 2.9388733519776267,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 7356
+    },
+    {
+      "epoch": 2.9392728725529365,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 7357
+    },
+    {
+      "epoch": 2.9396723931282462,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7358
+    },
+    {
+      "epoch": 2.9400719137035556,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 7359
+    },
+    {
+      "epoch": 2.9404714342788654,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7360
+    },
+    {
+      "epoch": 2.9408709548541747,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 7361
+    },
+    {
+      "epoch": 2.9412704754294845,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0707,
+      "step": 7362
+    },
+    {
+      "epoch": 2.9416699960047943,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 7363
+    },
+    {
+      "epoch": 2.942069516580104,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 7364
+    },
+    {
+      "epoch": 2.9424690371554134,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7365
+    },
+    {
+      "epoch": 2.942868557730723,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7366
+    },
+    {
+      "epoch": 2.9432680783060325,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7367
+    },
+    {
+      "epoch": 2.9436675988813423,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7368
+    },
+    {
+      "epoch": 2.944067119456652,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 7369
+    },
+    {
+      "epoch": 2.944466640031962,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 7370
+    },
+    {
+      "epoch": 2.9448661606072712,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 7371
+    },
+    {
+      "epoch": 2.945265681182581,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 7372
+    },
+    {
+      "epoch": 2.9456652017578904,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 7373
+    },
+    {
+      "epoch": 2.9460647223332,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 7374
+    },
+    {
+      "epoch": 2.94646424290851,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 7375
+    },
+    {
+      "epoch": 2.9468637634838193,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 7376
+    },
+    {
+      "epoch": 2.947263284059129,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 7377
+    },
+    {
+      "epoch": 2.9476628046344384,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 7378
+    },
+    {
+      "epoch": 2.948062325209748,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 7379
+    },
+    {
+      "epoch": 2.948461845785058,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7380
+    },
+    {
+      "epoch": 2.9488613663603678,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7381
+    },
+    {
+      "epoch": 2.949260886935677,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 7382
+    },
+    {
+      "epoch": 2.949660407510987,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 7383
+    },
+    {
+      "epoch": 2.9500599280862962,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 7384
+    },
+    {
+      "epoch": 2.950459448661606,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7385
+    },
+    {
+      "epoch": 2.950858969236916,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 7386
+    },
+    {
+      "epoch": 2.9512584898122256,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 7387
+    },
+    {
+      "epoch": 2.951658010387535,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7388
+    },
+    {
+      "epoch": 2.9520575309628447,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 7389
+    },
+    {
+      "epoch": 2.952457051538154,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 7390
+    },
+    {
+      "epoch": 2.952856572113464,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7391
+    },
+    {
+      "epoch": 2.9532560926887736,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 7392
+    },
+    {
+      "epoch": 2.953655613264083,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7393
+    },
+    {
+      "epoch": 2.9540551338393928,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 7394
+    },
+    {
+      "epoch": 2.954454654414702,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 7395
+    },
+    {
+      "epoch": 2.954854174990012,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7396
+    },
+    {
+      "epoch": 2.9552536955653217,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 7397
+    },
+    {
+      "epoch": 2.9556532161406315,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 7398
+    },
+    {
+      "epoch": 2.956052736715941,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 7399
+    },
+    {
+      "epoch": 2.9564522572912506,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 7400
+    },
+    {
+      "epoch": 2.95685177786656,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 7401
+    },
+    {
+      "epoch": 2.9572512984418697,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 7402
+    },
+    {
+      "epoch": 2.9576508190171795,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 7403
+    },
+    {
+      "epoch": 2.958050339592489,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 7404
+    },
+    {
+      "epoch": 2.9584498601677987,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7405
+    },
+    {
+      "epoch": 2.958849380743108,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 7406
+    },
+    {
+      "epoch": 2.959248901318418,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 7407
+    },
+    {
+      "epoch": 2.9596484218937276,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7408
+    },
+    {
+      "epoch": 2.9600479424690374,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 7409
+    },
+    {
+      "epoch": 2.9604474630443467,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7410
+    },
+    {
+      "epoch": 2.9608469836196565,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 7411
+    },
+    {
+      "epoch": 2.961246504194966,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 7412
+    },
+    {
+      "epoch": 2.9616460247702756,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 7413
+    },
+    {
+      "epoch": 2.9620455453455854,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7414
+    },
+    {
+      "epoch": 2.962445065920895,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 7415
+    },
+    {
+      "epoch": 2.9628445864962045,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 7416
+    },
+    {
+      "epoch": 2.9632441070715143,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0686,
+      "step": 7417
+    },
+    {
+      "epoch": 2.9636436276468237,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 7418
+    },
+    {
+      "epoch": 2.9640431482221334,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 7419
+    },
+    {
+      "epoch": 2.9644426687974432,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7420
+    },
+    {
+      "epoch": 2.9648421893727526,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7421
+    },
+    {
+      "epoch": 2.9652417099480624,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7422
+    },
+    {
+      "epoch": 2.9656412305233717,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 7423
+    },
+    {
+      "epoch": 2.9660407510986815,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0663,
+      "step": 7424
+    },
+    {
+      "epoch": 2.9664402716739913,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7425
+    },
+    {
+      "epoch": 2.966839792249301,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 7426
+    },
+    {
+      "epoch": 2.9672393128246104,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 7427
+    },
+    {
+      "epoch": 2.96763883339992,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 7428
+    },
+    {
+      "epoch": 2.9680383539752295,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 7429
+    },
+    {
+      "epoch": 2.9684378745505393,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7430
+    },
+    {
+      "epoch": 2.968837395125849,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7431
+    },
+    {
+      "epoch": 2.969236915701159,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 7432
+    },
+    {
+      "epoch": 2.9696364362764682,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 7433
+    },
+    {
+      "epoch": 2.970035956851778,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 7434
+    },
+    {
+      "epoch": 2.9704354774270874,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7435
+    },
+    {
+      "epoch": 2.970834998002397,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 7436
+    },
+    {
+      "epoch": 2.971234518577707,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 7437
+    },
+    {
+      "epoch": 2.9716340391530163,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 7438
+    },
+    {
+      "epoch": 2.972033559728326,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 7439
+    },
+    {
+      "epoch": 2.9724330803036354,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7440
+    },
+    {
+      "epoch": 2.972832600878945,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 7441
+    },
+    {
+      "epoch": 2.973232121454255,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 7442
+    },
+    {
+      "epoch": 2.9736316420295648,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7443
+    },
+    {
+      "epoch": 2.974031162604874,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 7444
+    },
+    {
+      "epoch": 2.974430683180184,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7445
+    },
+    {
+      "epoch": 2.9748302037554932,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 7446
+    },
+    {
+      "epoch": 2.975229724330803,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0657,
+      "step": 7447
+    },
+    {
+      "epoch": 2.975629244906113,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0674,
+      "step": 7448
+    },
+    {
+      "epoch": 2.976028765481422,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 7449
+    },
+    {
+      "epoch": 2.976428286056732,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 7450
+    },
+    {
+      "epoch": 2.9768278066320413,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7451
+    },
+    {
+      "epoch": 2.977227327207351,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7452
+    },
+    {
+      "epoch": 2.977626847782661,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7453
+    },
+    {
+      "epoch": 2.9780263683579706,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 7454
+    },
+    {
+      "epoch": 2.97842588893328,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0747,
+      "step": 7455
+    },
+    {
+      "epoch": 2.9788254095085898,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7456
+    },
+    {
+      "epoch": 2.979224930083899,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 7457
+    },
+    {
+      "epoch": 2.979624450659209,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 7458
+    },
+    {
+      "epoch": 2.9800239712345187,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7459
+    },
+    {
+      "epoch": 2.9804234918098285,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 7460
+    },
+    {
+      "epoch": 2.980823012385138,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 7461
+    },
+    {
+      "epoch": 2.9812225329604476,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 7462
+    },
+    {
+      "epoch": 2.981622053535757,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 7463
+    },
+    {
+      "epoch": 2.9820215741110667,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7464
+    },
+    {
+      "epoch": 2.9824210946863765,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 7465
+    },
+    {
+      "epoch": 2.982820615261686,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 7466
+    },
+    {
+      "epoch": 2.9832201358369956,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.063,
+      "step": 7467
+    },
+    {
+      "epoch": 2.983619656412305,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7468
+    },
+    {
+      "epoch": 2.9840191769876148,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7469
+    },
+    {
+      "epoch": 2.9844186975629245,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7470
+    },
+    {
+      "epoch": 2.9848182181382343,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 7471
+    },
+    {
+      "epoch": 2.9852177387135437,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7472
+    },
+    {
+      "epoch": 2.9856172592888535,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 7473
+    },
+    {
+      "epoch": 2.986016779864163,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7474
+    },
+    {
+      "epoch": 2.9864163004394726,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 7475
+    },
+    {
+      "epoch": 2.9868158210147824,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0367,
+      "step": 7476
+    },
+    {
+      "epoch": 2.987215341590092,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 7477
+    },
+    {
+      "epoch": 2.9876148621654015,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7478
+    },
+    {
+      "epoch": 2.9880143827407113,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 7479
+    },
+    {
+      "epoch": 2.9884139033160206,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0384,
+      "step": 7480
+    },
+    {
+      "epoch": 2.9888134238913304,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 7481
+    },
+    {
+      "epoch": 2.98921294446664,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 7482
+    },
+    {
+      "epoch": 2.9896124650419496,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7483
+    },
+    {
+      "epoch": 2.9900119856172593,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 7484
+    },
+    {
+      "epoch": 2.9904115061925687,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 7485
+    },
+    {
+      "epoch": 2.9908110267678785,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7486
+    },
+    {
+      "epoch": 2.9912105473431883,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 7487
+    },
+    {
+      "epoch": 2.991610067918498,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7488
+    },
+    {
+      "epoch": 2.9920095884938074,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 7489
+    },
+    {
+      "epoch": 2.992409109069117,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 7490
+    },
+    {
+      "epoch": 2.9928086296444265,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 7491
+    },
+    {
+      "epoch": 2.9932081502197363,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 7492
+    },
+    {
+      "epoch": 2.993607670795046,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 7493
+    },
+    {
+      "epoch": 2.9940071913703554,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 7494
+    },
+    {
+      "epoch": 2.994406711945665,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 7495
+    },
+    {
+      "epoch": 2.9948062325209746,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7496
+    },
+    {
+      "epoch": 2.9952057530962843,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7497
+    },
+    {
+      "epoch": 2.995605273671594,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 7498
+    },
+    {
+      "epoch": 2.996004794246904,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 7499
+    },
+    {
+      "epoch": 2.9964043148222133,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 7500
+    },
+    {
+      "epoch": 2.996803835397523,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 7501
+    },
+    {
+      "epoch": 2.9972033559728324,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 7502
+    },
+    {
+      "epoch": 2.997602876548142,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 7503
+    },
+    {
+      "epoch": 2.998002397123452,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 7504
+    },
+    {
+      "epoch": 2.9984019176987617,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 7505
+    },
+    {
+      "epoch": 2.998801438274071,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 7506
+    },
+    {
+      "epoch": 2.999200958849381,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0713,
+      "step": 7507
+    },
+    {
+      "epoch": 2.99960047942469,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 7508
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 7509
+    },
+    {
+      "epoch": 3.00039952057531,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 7510
+    },
+    {
+      "epoch": 3.000799041150619,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7511
+    },
+    {
+      "epoch": 3.001198561725929,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 7512
+    },
+    {
+      "epoch": 3.0015980823012387,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 7513
+    },
+    {
+      "epoch": 3.001997602876548,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 7514
+    },
+    {
+      "epoch": 3.002397123451858,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 7515
+    },
+    {
+      "epoch": 3.0027966440271676,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 7516
+    },
+    {
+      "epoch": 3.003196164602477,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 7517
+    },
+    {
+      "epoch": 3.0035956851777867,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 7518
+    },
+    {
+      "epoch": 3.003995205753096,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 7519
+    },
+    {
+      "epoch": 3.004394726328406,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 7520
+    },
+    {
+      "epoch": 3.0047942469037157,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7521
+    },
+    {
+      "epoch": 3.005193767479025,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 7522
+    },
+    {
+      "epoch": 3.005593288054335,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 7523
+    },
+    {
+      "epoch": 3.0059928086296446,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 7524
+    },
+    {
+      "epoch": 3.006392329204954,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 7525
+    },
+    {
+      "epoch": 3.0067918497802637,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 7526
+    },
+    {
+      "epoch": 3.0071913703555735,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 7527
+    },
+    {
+      "epoch": 3.007590890930883,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 7528
+    },
+    {
+      "epoch": 3.0079904115061926,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7529
+    },
+    {
+      "epoch": 3.0083899320815024,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 7530
+    },
+    {
+      "epoch": 3.0087894526568117,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 7531
+    },
+    {
+      "epoch": 3.0091889732321215,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 7532
+    },
+    {
+      "epoch": 3.009588493807431,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 7533
+    },
+    {
+      "epoch": 3.0099880143827407,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 7534
+    },
+    {
+      "epoch": 3.0103875349580504,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 7535
+    },
+    {
+      "epoch": 3.01078705553336,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 7536
+    },
+    {
+      "epoch": 3.0111865761086696,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 7537
+    },
+    {
+      "epoch": 3.0115860966839794,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 7538
+    },
+    {
+      "epoch": 3.0119856172592887,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 7539
+    },
+    {
+      "epoch": 3.0123851378345985,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7540
+    },
+    {
+      "epoch": 3.0127846584099083,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 7541
+    },
+    {
+      "epoch": 3.0131841789852176,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7542
+    },
+    {
+      "epoch": 3.0135836995605274,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 7543
+    },
+    {
+      "epoch": 3.013983220135837,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 7544
+    },
+    {
+      "epoch": 3.0143827407111465,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7545
+    },
+    {
+      "epoch": 3.0147822612864563,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0423,
+      "step": 7546
+    },
+    {
+      "epoch": 3.0151817818617657,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 7547
+    },
+    {
+      "epoch": 3.0155813024370755,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7548
+    },
+    {
+      "epoch": 3.0159808230123852,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 7549
+    },
+    {
+      "epoch": 3.0163803435876946,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 7550
+    },
+    {
+      "epoch": 3.0167798641630044,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 7551
+    },
+    {
+      "epoch": 3.017179384738314,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 7552
+    },
+    {
+      "epoch": 3.0175789053136235,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 7553
+    },
+    {
+      "epoch": 3.0179784258889333,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 7554
+    },
+    {
+      "epoch": 3.018377946464243,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 7555
+    },
+    {
+      "epoch": 3.0187774670395524,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 7556
+    },
+    {
+      "epoch": 3.019176987614862,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 7557
+    },
+    {
+      "epoch": 3.019576508190172,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 7558
+    },
+    {
+      "epoch": 3.0199760287654813,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 7559
+    },
+    {
+      "epoch": 3.020375549340791,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7560
+    },
+    {
+      "epoch": 3.020775069916101,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 7561
+    },
+    {
+      "epoch": 3.0211745904914102,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7562
+    },
+    {
+      "epoch": 3.02157411106672,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0397,
+      "step": 7563
+    },
+    {
+      "epoch": 3.0219736316420294,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0408,
+      "step": 7564
+    },
+    {
+      "epoch": 3.022373152217339,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7565
+    },
+    {
+      "epoch": 3.022772672792649,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 7566
+    },
+    {
+      "epoch": 3.0231721933679583,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7567
+    },
+    {
+      "epoch": 3.023571713943268,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 7568
+    },
+    {
+      "epoch": 3.023971234518578,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0645,
+      "step": 7569
+    },
+    {
+      "epoch": 3.024370755093887,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 7570
+    },
+    {
+      "epoch": 3.024770275669197,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 7571
+    },
+    {
+      "epoch": 3.0251697962445068,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 7572
+    },
+    {
+      "epoch": 3.025569316819816,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 7573
+    },
+    {
+      "epoch": 3.025968837395126,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7574
+    },
+    {
+      "epoch": 3.0263683579704357,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 7575
+    },
+    {
+      "epoch": 3.026767878545745,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 7576
+    },
+    {
+      "epoch": 3.027167399121055,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 7577
+    },
+    {
+      "epoch": 3.027566919696364,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7578
+    },
+    {
+      "epoch": 3.027966440271674,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 7579
+    },
+    {
+      "epoch": 3.0283659608469837,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 7580
+    },
+    {
+      "epoch": 3.028765481422293,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 7581
+    },
+    {
+      "epoch": 3.029165001997603,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7582
+    },
+    {
+      "epoch": 3.0295645225729126,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 7583
+    },
+    {
+      "epoch": 3.029964043148222,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 7584
+    },
+    {
+      "epoch": 3.0303635637235318,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 7585
+    },
+    {
+      "epoch": 3.0307630842988416,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0392,
+      "step": 7586
+    },
+    {
+      "epoch": 3.031162604874151,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 7587
+    },
+    {
+      "epoch": 3.0315621254494607,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 7588
+    },
+    {
+      "epoch": 3.0319616460247705,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 7589
+    },
+    {
+      "epoch": 3.03236116660008,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 7590
+    },
+    {
+      "epoch": 3.0327606871753896,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0697,
+      "step": 7591
+    },
+    {
+      "epoch": 3.033160207750699,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 7592
+    },
+    {
+      "epoch": 3.0335597283260087,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 7593
+    },
+    {
+      "epoch": 3.0339592489013185,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7594
+    },
+    {
+      "epoch": 3.034358769476628,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 7595
+    },
+    {
+      "epoch": 3.0347582900519376,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7596
+    },
+    {
+      "epoch": 3.0351578106272474,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 7597
+    },
+    {
+      "epoch": 3.0355573312025568,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 7598
+    },
+    {
+      "epoch": 3.0359568517778666,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7599
+    },
+    {
+      "epoch": 3.0363563723531763,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 7600
+    },
+    {
+      "epoch": 3.0367558929284857,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0418,
+      "step": 7601
+    },
+    {
+      "epoch": 3.0371554135037955,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7602
+    },
+    {
+      "epoch": 3.0375549340791053,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 7603
+    },
+    {
+      "epoch": 3.0379544546544146,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 7604
+    },
+    {
+      "epoch": 3.0383539752297244,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 7605
+    },
+    {
+      "epoch": 3.038753495805034,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 7606
+    },
+    {
+      "epoch": 3.0391530163803435,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 7607
+    },
+    {
+      "epoch": 3.0395525369556533,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 7608
+    },
+    {
+      "epoch": 3.0399520575309626,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 7609
+    },
+    {
+      "epoch": 3.0403515781062724,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 7610
+    },
+    {
+      "epoch": 3.040751098681582,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 7611
+    },
+    {
+      "epoch": 3.0411506192568916,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 7612
+    },
+    {
+      "epoch": 3.0415501398322013,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7613
+    },
+    {
+      "epoch": 3.041949660407511,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 7614
+    },
+    {
+      "epoch": 3.0423491809828205,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 7615
+    },
+    {
+      "epoch": 3.0427487015581303,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 7616
+    },
+    {
+      "epoch": 3.04314822213344,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 7617
+    },
+    {
+      "epoch": 3.0435477427087494,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 7618
+    },
+    {
+      "epoch": 3.043947263284059,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 7619
+    },
+    {
+      "epoch": 3.044346783859369,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 7620
+    },
+    {
+      "epoch": 3.0447463044346783,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 7621
+    },
+    {
+      "epoch": 3.045145825009988,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 7622
+    },
+    {
+      "epoch": 3.0455453455852974,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7623
+    },
+    {
+      "epoch": 3.0459448661606072,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 7624
+    },
+    {
+      "epoch": 3.046344386735917,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 7625
+    },
+    {
+      "epoch": 3.0467439073112264,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 7626
+    },
+    {
+      "epoch": 3.047143427886536,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7627
+    },
+    {
+      "epoch": 3.047542948461846,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7628
+    },
+    {
+      "epoch": 3.0479424690371553,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 7629
+    },
+    {
+      "epoch": 3.048341989612465,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.03,
+      "step": 7630
+    },
+    {
+      "epoch": 3.048741510187775,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7631
+    },
+    {
+      "epoch": 3.049141030763084,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 7632
+    },
+    {
+      "epoch": 3.049540551338394,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7633
+    },
+    {
+      "epoch": 3.0499400719137038,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 7634
+    },
+    {
+      "epoch": 3.050339592489013,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 7635
+    },
+    {
+      "epoch": 3.050739113064323,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 7636
+    },
+    {
+      "epoch": 3.0511386336396322,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 7637
+    },
+    {
+      "epoch": 3.051538154214942,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 7638
+    },
+    {
+      "epoch": 3.051937674790252,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7639
+    },
+    {
+      "epoch": 3.052337195365561,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 7640
+    },
+    {
+      "epoch": 3.052736715940871,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7641
+    },
+    {
+      "epoch": 3.0531362365161807,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 7642
+    },
+    {
+      "epoch": 3.05353575709149,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 7643
+    },
+    {
+      "epoch": 3.0539352776668,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 7644
+    },
+    {
+      "epoch": 3.0543347982421096,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7645
+    },
+    {
+      "epoch": 3.054734318817419,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 7646
+    },
+    {
+      "epoch": 3.0551338393927288,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 7647
+    },
+    {
+      "epoch": 3.0555333599680385,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 7648
+    },
+    {
+      "epoch": 3.055932880543348,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 7649
+    },
+    {
+      "epoch": 3.0563324011186577,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 7650
+    },
+    {
+      "epoch": 3.0567319216939675,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 7651
+    },
+    {
+      "epoch": 3.057131442269277,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 7652
+    },
+    {
+      "epoch": 3.0575309628445866,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 7653
+    },
+    {
+      "epoch": 3.057930483419896,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 7654
+    },
+    {
+      "epoch": 3.0583300039952057,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7655
+    },
+    {
+      "epoch": 3.0587295245705155,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7656
+    },
+    {
+      "epoch": 3.059129045145825,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7657
+    },
+    {
+      "epoch": 3.0595285657211346,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7658
+    },
+    {
+      "epoch": 3.0599280862964444,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 7659
+    },
+    {
+      "epoch": 3.0603276068717538,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 7660
+    },
+    {
+      "epoch": 3.0607271274470635,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 7661
+    },
+    {
+      "epoch": 3.0611266480223733,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7662
+    },
+    {
+      "epoch": 3.0615261685976827,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 7663
+    },
+    {
+      "epoch": 3.0619256891729925,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 7664
+    },
+    {
+      "epoch": 3.0623252097483022,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 7665
+    },
+    {
+      "epoch": 3.0627247303236116,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7666
+    },
+    {
+      "epoch": 3.0631242508989214,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7667
+    },
+    {
+      "epoch": 3.0635237714742307,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 7668
+    },
+    {
+      "epoch": 3.0639232920495405,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 7669
+    },
+    {
+      "epoch": 3.0643228126248503,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 7670
+    },
+    {
+      "epoch": 3.0647223332001596,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 7671
+    },
+    {
+      "epoch": 3.0651218537754694,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 7672
+    },
+    {
+      "epoch": 3.065521374350779,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 7673
+    },
+    {
+      "epoch": 3.0659208949260885,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 7674
+    },
+    {
+      "epoch": 3.0663204155013983,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 7675
+    },
+    {
+      "epoch": 3.066719936076708,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 7676
+    },
+    {
+      "epoch": 3.0671194566520175,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 7677
+    },
+    {
+      "epoch": 3.0675189772273272,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0682,
+      "step": 7678
+    },
+    {
+      "epoch": 3.067918497802637,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7679
+    },
+    {
+      "epoch": 3.0683180183779464,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 7680
+    },
+    {
+      "epoch": 3.068717538953256,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 7681
+    },
+    {
+      "epoch": 3.0691170595285655,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7682
+    },
+    {
+      "epoch": 3.0695165801038753,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 7683
+    },
+    {
+      "epoch": 3.069916100679185,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7684
+    },
+    {
+      "epoch": 3.0703156212544944,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 7685
+    },
+    {
+      "epoch": 3.070715141829804,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 7686
+    },
+    {
+      "epoch": 3.071114662405114,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7687
+    },
+    {
+      "epoch": 3.0715141829804233,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 7688
+    },
+    {
+      "epoch": 3.071913703555733,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0358,
+      "step": 7689
+    },
+    {
+      "epoch": 3.072313224131043,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7690
+    },
+    {
+      "epoch": 3.0727127447063523,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 7691
+    },
+    {
+      "epoch": 3.073112265281662,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 7692
+    },
+    {
+      "epoch": 3.073511785856972,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 7693
+    },
+    {
+      "epoch": 3.073911306432281,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7694
+    },
+    {
+      "epoch": 3.074310827007591,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 7695
+    },
+    {
+      "epoch": 3.0747103475829007,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0662,
+      "step": 7696
+    },
+    {
+      "epoch": 3.07510986815821,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 7697
+    },
+    {
+      "epoch": 3.07550938873352,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 7698
+    },
+    {
+      "epoch": 3.075908909308829,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 7699
+    },
+    {
+      "epoch": 3.076308429884139,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7700
+    },
+    {
+      "epoch": 3.076707950459449,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7701
+    },
+    {
+      "epoch": 3.077107471034758,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 7702
+    },
+    {
+      "epoch": 3.077506991610068,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 7703
+    },
+    {
+      "epoch": 3.0779065121853777,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0413,
+      "step": 7704
+    },
+    {
+      "epoch": 3.078306032760687,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 7705
+    },
+    {
+      "epoch": 3.078705553335997,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7706
+    },
+    {
+      "epoch": 3.0791050739113066,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 7707
+    },
+    {
+      "epoch": 3.079504594486616,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 7708
+    },
+    {
+      "epoch": 3.0799041150619257,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7709
+    },
+    {
+      "epoch": 3.0803036356372355,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 7710
+    },
+    {
+      "epoch": 3.080703156212545,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 7711
+    },
+    {
+      "epoch": 3.0811026767878547,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7712
+    },
+    {
+      "epoch": 3.081502197363164,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 7713
+    },
+    {
+      "epoch": 3.081901717938474,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7714
+    },
+    {
+      "epoch": 3.0823012385137836,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0433,
+      "step": 7715
+    },
+    {
+      "epoch": 3.082700759089093,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7716
+    },
+    {
+      "epoch": 3.0831002796644027,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 7717
+    },
+    {
+      "epoch": 3.0834998002397125,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 7718
+    },
+    {
+      "epoch": 3.083899320815022,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 7719
+    },
+    {
+      "epoch": 3.0842988413903316,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 7720
+    },
+    {
+      "epoch": 3.0846983619656414,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 7721
+    },
+    {
+      "epoch": 3.0850978825409507,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 7722
+    },
+    {
+      "epoch": 3.0854974031162605,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7723
+    },
+    {
+      "epoch": 3.0858969236915703,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0683,
+      "step": 7724
+    },
+    {
+      "epoch": 3.0862964442668797,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 7725
+    },
+    {
+      "epoch": 3.0866959648421894,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7726
+    },
+    {
+      "epoch": 3.087095485417499,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 7727
+    },
+    {
+      "epoch": 3.0874950059928086,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 7728
+    },
+    {
+      "epoch": 3.0878945265681184,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 7729
+    },
+    {
+      "epoch": 3.0882940471434277,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7730
+    },
+    {
+      "epoch": 3.0886935677187375,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 7731
+    },
+    {
+      "epoch": 3.0890930882940473,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 7732
+    },
+    {
+      "epoch": 3.0894926088693566,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7733
+    },
+    {
+      "epoch": 3.0898921294446664,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 7734
+    },
+    {
+      "epoch": 3.090291650019976,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 7735
+    },
+    {
+      "epoch": 3.0906911705952855,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 7736
+    },
+    {
+      "epoch": 3.0910906911705953,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7737
+    },
+    {
+      "epoch": 3.091490211745905,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 7738
+    },
+    {
+      "epoch": 3.0918897323212144,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 7739
+    },
+    {
+      "epoch": 3.0922892528965242,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 7740
+    },
+    {
+      "epoch": 3.092688773471834,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7741
+    },
+    {
+      "epoch": 3.0930882940471434,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7742
+    },
+    {
+      "epoch": 3.093487814622453,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 7743
+    },
+    {
+      "epoch": 3.0938873351977625,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0659,
+      "step": 7744
+    },
+    {
+      "epoch": 3.0942868557730723,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 7745
+    },
+    {
+      "epoch": 3.094686376348382,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 7746
+    },
+    {
+      "epoch": 3.0950858969236914,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7747
+    },
+    {
+      "epoch": 3.095485417499001,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 7748
+    },
+    {
+      "epoch": 3.095884938074311,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 7749
+    },
+    {
+      "epoch": 3.0962844586496203,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 7750
+    },
+    {
+      "epoch": 3.09668397922493,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 7751
+    },
+    {
+      "epoch": 3.09708349980024,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 7752
+    },
+    {
+      "epoch": 3.0974830203755492,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0723,
+      "step": 7753
+    },
+    {
+      "epoch": 3.097882540950859,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 7754
+    },
+    {
+      "epoch": 3.098282061526169,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 7755
+    },
+    {
+      "epoch": 3.098681582101478,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7756
+    },
+    {
+      "epoch": 3.099081102676788,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 7757
+    },
+    {
+      "epoch": 3.0994806232520973,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 7758
+    },
+    {
+      "epoch": 3.099880143827407,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7759
+    },
+    {
+      "epoch": 3.100279664402717,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7760
+    },
+    {
+      "epoch": 3.100679184978026,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 7761
+    },
+    {
+      "epoch": 3.101078705553336,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 7762
+    },
+    {
+      "epoch": 3.1014782261286458,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 7763
+    },
+    {
+      "epoch": 3.101877746703955,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 7764
+    },
+    {
+      "epoch": 3.102277267279265,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 7765
+    },
+    {
+      "epoch": 3.1026767878545747,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0366,
+      "step": 7766
+    },
+    {
+      "epoch": 3.103076308429884,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 7767
+    },
+    {
+      "epoch": 3.103475829005194,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 7768
+    },
+    {
+      "epoch": 3.1038753495805036,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 7769
+    },
+    {
+      "epoch": 3.104274870155813,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 7770
+    },
+    {
+      "epoch": 3.1046743907311227,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7771
+    },
+    {
+      "epoch": 3.105073911306432,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0394,
+      "step": 7772
+    },
+    {
+      "epoch": 3.105473431881742,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 7773
+    },
+    {
+      "epoch": 3.1058729524570516,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 7774
+    },
+    {
+      "epoch": 3.106272473032361,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7775
+    },
+    {
+      "epoch": 3.1066719936076708,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 7776
+    },
+    {
+      "epoch": 3.1070715141829806,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 7777
+    },
+    {
+      "epoch": 3.10747103475829,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 7778
+    },
+    {
+      "epoch": 3.1078705553335997,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7779
+    },
+    {
+      "epoch": 3.1082700759089095,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 7780
+    },
+    {
+      "epoch": 3.108669596484219,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 7781
+    },
+    {
+      "epoch": 3.1090691170595286,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0702,
+      "step": 7782
+    },
+    {
+      "epoch": 3.1094686376348384,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0721,
+      "step": 7783
+    },
+    {
+      "epoch": 3.1098681582101477,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 7784
+    },
+    {
+      "epoch": 3.1102676787854575,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 7785
+    },
+    {
+      "epoch": 3.1106671993607673,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 7786
+    },
+    {
+      "epoch": 3.1110667199360766,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 7787
+    },
+    {
+      "epoch": 3.1114662405113864,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 7788
+    },
+    {
+      "epoch": 3.1118657610866958,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 7789
+    },
+    {
+      "epoch": 3.1122652816620056,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 7790
+    },
+    {
+      "epoch": 3.1126648022373153,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7791
+    },
+    {
+      "epoch": 3.1130643228126247,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 7792
+    },
+    {
+      "epoch": 3.1134638433879345,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 7793
+    },
+    {
+      "epoch": 3.1138633639632443,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 7794
+    },
+    {
+      "epoch": 3.1142628845385536,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 7795
+    },
+    {
+      "epoch": 3.1146624051138634,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 7796
+    },
+    {
+      "epoch": 3.115061925689173,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7797
+    },
+    {
+      "epoch": 3.1154614462644825,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 7798
+    },
+    {
+      "epoch": 3.1158609668397923,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7799
+    },
+    {
+      "epoch": 3.116260487415102,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0408,
+      "step": 7800
+    },
+    {
+      "epoch": 3.1166600079904114,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7801
+    },
+    {
+      "epoch": 3.117059528565721,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7802
+    },
+    {
+      "epoch": 3.1174590491410306,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 7803
+    },
+    {
+      "epoch": 3.1178585697163403,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7804
+    },
+    {
+      "epoch": 3.11825809029165,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 7805
+    },
+    {
+      "epoch": 3.1186576108669595,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0403,
+      "step": 7806
+    },
+    {
+      "epoch": 3.1190571314422693,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 7807
+    },
+    {
+      "epoch": 3.119456652017579,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 7808
+    },
+    {
+      "epoch": 3.1198561725928884,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 7809
+    },
+    {
+      "epoch": 3.120255693168198,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7810
+    },
+    {
+      "epoch": 3.120655213743508,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 7811
+    },
+    {
+      "epoch": 3.1210547343188173,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 7812
+    },
+    {
+      "epoch": 3.121454254894127,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 7813
+    },
+    {
+      "epoch": 3.121853775469437,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 7814
+    },
+    {
+      "epoch": 3.122253296044746,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 7815
+    },
+    {
+      "epoch": 3.122652816620056,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 7816
+    },
+    {
+      "epoch": 3.1230523371953653,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 7817
+    },
+    {
+      "epoch": 3.123451857770675,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 7818
+    },
+    {
+      "epoch": 3.123851378345985,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.04,
+      "step": 7819
+    },
+    {
+      "epoch": 3.1242508989212943,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 7820
+    },
+    {
+      "epoch": 3.124650419496604,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7821
+    },
+    {
+      "epoch": 3.125049940071914,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0643,
+      "step": 7822
+    },
+    {
+      "epoch": 3.125449460647223,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 7823
+    },
+    {
+      "epoch": 3.125848981222533,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 7824
+    },
+    {
+      "epoch": 3.1262485017978427,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7825
+    },
+    {
+      "epoch": 3.126648022373152,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 7826
+    },
+    {
+      "epoch": 3.127047542948462,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 7827
+    },
+    {
+      "epoch": 3.1274470635237717,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 7828
+    },
+    {
+      "epoch": 3.127846584099081,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 7829
+    },
+    {
+      "epoch": 3.128246104674391,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 7830
+    },
+    {
+      "epoch": 3.1286456252497006,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 7831
+    },
+    {
+      "epoch": 3.12904514582501,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 7832
+    },
+    {
+      "epoch": 3.1294446664003197,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0376,
+      "step": 7833
+    },
+    {
+      "epoch": 3.129844186975629,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 7834
+    },
+    {
+      "epoch": 3.130243707550939,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7835
+    },
+    {
+      "epoch": 3.1306432281262486,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 7836
+    },
+    {
+      "epoch": 3.131042748701558,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7837
+    },
+    {
+      "epoch": 3.1314422692768678,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7838
+    },
+    {
+      "epoch": 3.1318417898521775,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 7839
+    },
+    {
+      "epoch": 3.132241310427487,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7840
+    },
+    {
+      "epoch": 3.1326408310027967,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 7841
+    },
+    {
+      "epoch": 3.1330403515781065,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 7842
+    },
+    {
+      "epoch": 3.133439872153416,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7843
+    },
+    {
+      "epoch": 3.1338393927287256,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 7844
+    },
+    {
+      "epoch": 3.134238913304035,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 7845
+    },
+    {
+      "epoch": 3.1346384338793447,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 7846
+    },
+    {
+      "epoch": 3.1350379544546545,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 7847
+    },
+    {
+      "epoch": 3.135437475029964,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 7848
+    },
+    {
+      "epoch": 3.1358369956052736,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 7849
+    },
+    {
+      "epoch": 3.1362365161805834,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7850
+    },
+    {
+      "epoch": 3.1366360367558928,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 7851
+    },
+    {
+      "epoch": 3.1370355573312025,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 7852
+    },
+    {
+      "epoch": 3.1374350779065123,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 7853
+    },
+    {
+      "epoch": 3.1378345984818217,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 7854
+    },
+    {
+      "epoch": 3.1382341190571315,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7855
+    },
+    {
+      "epoch": 3.1386336396324412,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 7856
+    },
+    {
+      "epoch": 3.1390331602077506,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 7857
+    },
+    {
+      "epoch": 3.1394326807830604,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7858
+    },
+    {
+      "epoch": 3.13983220135837,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7859
+    },
+    {
+      "epoch": 3.1402317219336795,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7860
+    },
+    {
+      "epoch": 3.1406312425089893,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 7861
+    },
+    {
+      "epoch": 3.1410307630842986,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 7862
+    },
+    {
+      "epoch": 3.1414302836596084,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 7863
+    },
+    {
+      "epoch": 3.141829804234918,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 7864
+    },
+    {
+      "epoch": 3.1422293248102275,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7865
+    },
+    {
+      "epoch": 3.1426288453855373,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7866
+    },
+    {
+      "epoch": 3.143028365960847,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 7867
+    },
+    {
+      "epoch": 3.1434278865361565,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 7868
+    },
+    {
+      "epoch": 3.1438274071114662,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 7869
+    },
+    {
+      "epoch": 3.144226927686776,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 7870
+    },
+    {
+      "epoch": 3.1446264482620854,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 7871
+    },
+    {
+      "epoch": 3.145025968837395,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 7872
+    },
+    {
+      "epoch": 3.145425489412705,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 7873
+    },
+    {
+      "epoch": 3.1458250099880143,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 7874
+    },
+    {
+      "epoch": 3.146224530563324,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 7875
+    },
+    {
+      "epoch": 3.146624051138634,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 7876
+    },
+    {
+      "epoch": 3.147023571713943,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7877
+    },
+    {
+      "epoch": 3.147423092289253,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 7878
+    },
+    {
+      "epoch": 3.1478226128645623,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7879
+    },
+    {
+      "epoch": 3.148222133439872,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7880
+    },
+    {
+      "epoch": 3.148621654015182,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 7881
+    },
+    {
+      "epoch": 3.1490211745904912,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 7882
+    },
+    {
+      "epoch": 3.149420695165801,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 7883
+    },
+    {
+      "epoch": 3.149820215741111,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 7884
+    },
+    {
+      "epoch": 3.15021973631642,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 7885
+    },
+    {
+      "epoch": 3.15061925689173,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 7886
+    },
+    {
+      "epoch": 3.1510187774670397,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 7887
+    },
+    {
+      "epoch": 3.151418298042349,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 7888
+    },
+    {
+      "epoch": 3.151817818617659,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 7889
+    },
+    {
+      "epoch": 3.152217339192968,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 7890
+    },
+    {
+      "epoch": 3.152616859768278,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 7891
+    },
+    {
+      "epoch": 3.1530163803435878,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 7892
+    },
+    {
+      "epoch": 3.153415900918897,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 7893
+    },
+    {
+      "epoch": 3.153815421494207,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 7894
+    },
+    {
+      "epoch": 3.1542149420695167,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 7895
+    },
+    {
+      "epoch": 3.154614462644826,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 7896
+    },
+    {
+      "epoch": 3.155013983220136,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 7897
+    },
+    {
+      "epoch": 3.1554135037954456,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 7898
+    },
+    {
+      "epoch": 3.155813024370755,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7899
+    },
+    {
+      "epoch": 3.1562125449460647,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 7900
+    },
+    {
+      "epoch": 3.1566120655213745,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 7901
+    },
+    {
+      "epoch": 3.157011586096684,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 7902
+    },
+    {
+      "epoch": 3.1574111066719937,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 7903
+    },
+    {
+      "epoch": 3.1578106272473034,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 7904
+    },
+    {
+      "epoch": 3.158210147822613,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 7905
+    },
+    {
+      "epoch": 3.1586096683979226,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 7906
+    },
+    {
+      "epoch": 3.159009188973232,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 7907
+    },
+    {
+      "epoch": 3.1594087095485417,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 7908
+    },
+    {
+      "epoch": 3.1598082301238515,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0705,
+      "step": 7909
+    },
+    {
+      "epoch": 3.160207750699161,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 7910
+    },
+    {
+      "epoch": 3.1606072712744706,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 7911
+    },
+    {
+      "epoch": 3.1610067918497804,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 7912
+    },
+    {
+      "epoch": 3.1614063124250897,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 7913
+    },
+    {
+      "epoch": 3.1618058330003995,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0368,
+      "step": 7914
+    },
+    {
+      "epoch": 3.1622053535757093,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 7915
+    },
+    {
+      "epoch": 3.1626048741510187,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 7916
+    },
+    {
+      "epoch": 3.1630043947263284,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 7917
+    },
+    {
+      "epoch": 3.1634039153016382,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 7918
+    },
+    {
+      "epoch": 3.1638034358769476,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7919
+    },
+    {
+      "epoch": 3.1642029564522574,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 7920
+    },
+    {
+      "epoch": 3.164602477027567,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0694,
+      "step": 7921
+    },
+    {
+      "epoch": 3.1650019976028765,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0399,
+      "step": 7922
+    },
+    {
+      "epoch": 3.1654015181781863,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 7923
+    },
+    {
+      "epoch": 3.1658010387534956,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 7924
+    },
+    {
+      "epoch": 3.1662005593288054,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 7925
+    },
+    {
+      "epoch": 3.166600079904115,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 7926
+    },
+    {
+      "epoch": 3.1669996004794245,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 7927
+    },
+    {
+      "epoch": 3.1673991210547343,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 7928
+    },
+    {
+      "epoch": 3.167798641630044,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 7929
+    },
+    {
+      "epoch": 3.1681981622053534,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 7930
+    },
+    {
+      "epoch": 3.1685976827806632,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 7931
+    },
+    {
+      "epoch": 3.168997203355973,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 7932
+    },
+    {
+      "epoch": 3.1693967239312824,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 7933
+    },
+    {
+      "epoch": 3.169796244506592,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 7934
+    },
+    {
+      "epoch": 3.1701957650819015,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 7935
+    },
+    {
+      "epoch": 3.1705952856572113,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 7936
+    },
+    {
+      "epoch": 3.170994806232521,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 7937
+    },
+    {
+      "epoch": 3.1713943268078304,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 7938
+    },
+    {
+      "epoch": 3.17179384738314,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 7939
+    },
+    {
+      "epoch": 3.17219336795845,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 7940
+    },
+    {
+      "epoch": 3.1725928885337593,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 7941
+    },
+    {
+      "epoch": 3.172992409109069,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7942
+    },
+    {
+      "epoch": 3.173391929684379,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 7943
+    },
+    {
+      "epoch": 3.1737914502596882,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 7944
+    },
+    {
+      "epoch": 3.174190970834998,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 7945
+    },
+    {
+      "epoch": 3.174590491410308,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 7946
+    },
+    {
+      "epoch": 3.174990011985617,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7947
+    },
+    {
+      "epoch": 3.175389532560927,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 7948
+    },
+    {
+      "epoch": 3.1757890531362367,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 7949
+    },
+    {
+      "epoch": 3.176188573711546,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 7950
+    },
+    {
+      "epoch": 3.176588094286856,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 7951
+    },
+    {
+      "epoch": 3.176987614862165,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0433,
+      "step": 7952
+    },
+    {
+      "epoch": 3.177387135437475,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 7953
+    },
+    {
+      "epoch": 3.1777866560127848,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 7954
+    },
+    {
+      "epoch": 3.178186176588094,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0415,
+      "step": 7955
+    },
+    {
+      "epoch": 3.178585697163404,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 7956
+    },
+    {
+      "epoch": 3.1789852177387137,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 7957
+    },
+    {
+      "epoch": 3.179384738314023,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 7958
+    },
+    {
+      "epoch": 3.179784258889333,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 7959
+    },
+    {
+      "epoch": 3.1801837794646426,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0677,
+      "step": 7960
+    },
+    {
+      "epoch": 3.180583300039952,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 7961
+    },
+    {
+      "epoch": 3.1809828206152617,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 7962
+    },
+    {
+      "epoch": 3.1813823411905715,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 7963
+    },
+    {
+      "epoch": 3.181781861765881,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 7964
+    },
+    {
+      "epoch": 3.1821813823411906,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 7965
+    },
+    {
+      "epoch": 3.1825809029165004,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 7966
+    },
+    {
+      "epoch": 3.1829804234918098,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 7967
+    },
+    {
+      "epoch": 3.1833799440671195,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 7968
+    },
+    {
+      "epoch": 3.183779464642429,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 7969
+    },
+    {
+      "epoch": 3.1841789852177387,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7970
+    },
+    {
+      "epoch": 3.1845785057930485,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 7971
+    },
+    {
+      "epoch": 3.184978026368358,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 7972
+    },
+    {
+      "epoch": 3.1853775469436676,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 7973
+    },
+    {
+      "epoch": 3.1857770675189774,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 7974
+    },
+    {
+      "epoch": 3.1861765880942867,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 7975
+    },
+    {
+      "epoch": 3.1865761086695965,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 7976
+    },
+    {
+      "epoch": 3.1869756292449063,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 7977
+    },
+    {
+      "epoch": 3.1873751498202156,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 7978
+    },
+    {
+      "epoch": 3.1877746703955254,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 7979
+    },
+    {
+      "epoch": 3.1881741909708348,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 7980
+    },
+    {
+      "epoch": 3.1885737115461446,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7981
+    },
+    {
+      "epoch": 3.1889732321214543,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 7982
+    },
+    {
+      "epoch": 3.1893727526967637,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 7983
+    },
+    {
+      "epoch": 3.1897722732720735,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 7984
+    },
+    {
+      "epoch": 3.1901717938473833,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 7985
+    },
+    {
+      "epoch": 3.1905713144226926,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 7986
+    },
+    {
+      "epoch": 3.1909708349980024,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0366,
+      "step": 7987
+    },
+    {
+      "epoch": 3.191370355573312,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 7988
+    },
+    {
+      "epoch": 3.1917698761486215,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 7989
+    },
+    {
+      "epoch": 3.1921693967239313,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 7990
+    },
+    {
+      "epoch": 3.192568917299241,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 7991
+    },
+    {
+      "epoch": 3.1929684378745504,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 7992
+    },
+    {
+      "epoch": 3.19336795844986,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 7993
+    },
+    {
+      "epoch": 3.19376747902517,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 7994
+    },
+    {
+      "epoch": 3.1941669996004793,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 7995
+    },
+    {
+      "epoch": 3.194566520175789,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 7996
+    },
+    {
+      "epoch": 3.1949660407510985,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 7997
+    },
+    {
+      "epoch": 3.1953655613264083,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 7998
+    },
+    {
+      "epoch": 3.195765081901718,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 7999
+    },
+    {
+      "epoch": 3.1961646024770274,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 8000
+    },
+    {
+      "epoch": 3.196564123052337,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8001
+    },
+    {
+      "epoch": 3.196963643627647,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 8002
+    },
+    {
+      "epoch": 3.1973631642029563,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 8003
+    },
+    {
+      "epoch": 3.197762684778266,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0655,
+      "step": 8004
+    },
+    {
+      "epoch": 3.198162205353576,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 8005
+    },
+    {
+      "epoch": 3.198561725928885,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0399,
+      "step": 8006
+    },
+    {
+      "epoch": 3.198961246504195,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0364,
+      "step": 8007
+    },
+    {
+      "epoch": 3.199360767079505,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 8008
+    },
+    {
+      "epoch": 3.199760287654814,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8009
+    },
+    {
+      "epoch": 3.200159808230124,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 8010
+    },
+    {
+      "epoch": 3.2005593288054337,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 8011
+    },
+    {
+      "epoch": 3.200958849380743,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 8012
+    },
+    {
+      "epoch": 3.201358369956053,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 8013
+    },
+    {
+      "epoch": 3.201757890531362,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 8014
+    },
+    {
+      "epoch": 3.202157411106672,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8015
+    },
+    {
+      "epoch": 3.2025569316819817,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 8016
+    },
+    {
+      "epoch": 3.202956452257291,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 8017
+    },
+    {
+      "epoch": 3.203355972832601,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 8018
+    },
+    {
+      "epoch": 3.2037554934079107,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 8019
+    },
+    {
+      "epoch": 3.20415501398322,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 8020
+    },
+    {
+      "epoch": 3.20455453455853,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 8021
+    },
+    {
+      "epoch": 3.2049540551338396,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0676,
+      "step": 8022
+    },
+    {
+      "epoch": 3.205353575709149,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 8023
+    },
+    {
+      "epoch": 3.2057530962844587,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 8024
+    },
+    {
+      "epoch": 3.206152616859768,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 8025
+    },
+    {
+      "epoch": 3.206552137435078,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 8026
+    },
+    {
+      "epoch": 3.2069516580103876,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8027
+    },
+    {
+      "epoch": 3.207351178585697,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 8028
+    },
+    {
+      "epoch": 3.2077506991610067,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8029
+    },
+    {
+      "epoch": 3.2081502197363165,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8030
+    },
+    {
+      "epoch": 3.208549740311626,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 8031
+    },
+    {
+      "epoch": 3.2089492608869357,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 8032
+    },
+    {
+      "epoch": 3.2093487814622454,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8033
+    },
+    {
+      "epoch": 3.209748302037555,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 8034
+    },
+    {
+      "epoch": 3.2101478226128646,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 8035
+    },
+    {
+      "epoch": 3.2105473431881744,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8036
+    },
+    {
+      "epoch": 3.2109468637634837,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 8037
+    },
+    {
+      "epoch": 3.2113463843387935,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 8038
+    },
+    {
+      "epoch": 3.2117459049141033,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 8039
+    },
+    {
+      "epoch": 3.2121454254894126,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 8040
+    },
+    {
+      "epoch": 3.2125449460647224,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 8041
+    },
+    {
+      "epoch": 3.2129444666400317,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 8042
+    },
+    {
+      "epoch": 3.2133439872153415,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 8043
+    },
+    {
+      "epoch": 3.2137435077906513,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 8044
+    },
+    {
+      "epoch": 3.2141430283659607,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0648,
+      "step": 8045
+    },
+    {
+      "epoch": 3.2145425489412705,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 8046
+    },
+    {
+      "epoch": 3.2149420695165802,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 8047
+    },
+    {
+      "epoch": 3.2153415900918896,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8048
+    },
+    {
+      "epoch": 3.2157411106671994,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 8049
+    },
+    {
+      "epoch": 3.216140631242509,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 8050
+    },
+    {
+      "epoch": 3.2165401518178185,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 8051
+    },
+    {
+      "epoch": 3.2169396723931283,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8052
+    },
+    {
+      "epoch": 3.217339192968438,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 8053
+    },
+    {
+      "epoch": 3.2177387135437474,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 8054
+    },
+    {
+      "epoch": 3.218138234119057,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 8055
+    },
+    {
+      "epoch": 3.218537754694367,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8056
+    },
+    {
+      "epoch": 3.2189372752696763,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 8057
+    },
+    {
+      "epoch": 3.219336795844986,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8058
+    },
+    {
+      "epoch": 3.2197363164202955,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 8059
+    },
+    {
+      "epoch": 3.2201358369956052,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 8060
+    },
+    {
+      "epoch": 3.220535357570915,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 8061
+    },
+    {
+      "epoch": 3.2209348781462244,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8062
+    },
+    {
+      "epoch": 3.221334398721534,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 8063
+    },
+    {
+      "epoch": 3.221733919296844,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8064
+    },
+    {
+      "epoch": 3.2221334398721533,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8065
+    },
+    {
+      "epoch": 3.222532960447463,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 8066
+    },
+    {
+      "epoch": 3.222932481022773,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8067
+    },
+    {
+      "epoch": 3.223332001598082,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.071,
+      "step": 8068
+    },
+    {
+      "epoch": 3.223731522173392,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 8069
+    },
+    {
+      "epoch": 3.2241310427487013,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8070
+    },
+    {
+      "epoch": 3.224530563324011,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0415,
+      "step": 8071
+    },
+    {
+      "epoch": 3.224930083899321,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8072
+    },
+    {
+      "epoch": 3.2253296044746302,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 8073
+    },
+    {
+      "epoch": 3.22572912504994,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8074
+    },
+    {
+      "epoch": 3.22612864562525,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 8075
+    },
+    {
+      "epoch": 3.226528166200559,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 8076
+    },
+    {
+      "epoch": 3.226927686775869,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8077
+    },
+    {
+      "epoch": 3.2273272073511787,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 8078
+    },
+    {
+      "epoch": 3.227726727926488,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 8079
+    },
+    {
+      "epoch": 3.228126248501798,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 8080
+    },
+    {
+      "epoch": 3.2285257690771076,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 8081
+    },
+    {
+      "epoch": 3.228925289652417,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8082
+    },
+    {
+      "epoch": 3.2293248102277268,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8083
+    },
+    {
+      "epoch": 3.2297243308030366,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8084
+    },
+    {
+      "epoch": 3.230123851378346,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8085
+    },
+    {
+      "epoch": 3.2305233719536557,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 8086
+    },
+    {
+      "epoch": 3.230922892528965,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 8087
+    },
+    {
+      "epoch": 3.231322413104275,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 8088
+    },
+    {
+      "epoch": 3.2317219336795846,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8089
+    },
+    {
+      "epoch": 3.232121454254894,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 8090
+    },
+    {
+      "epoch": 3.2325209748302037,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 8091
+    },
+    {
+      "epoch": 3.2329204954055135,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 8092
+    },
+    {
+      "epoch": 3.233320015980823,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8093
+    },
+    {
+      "epoch": 3.2337195365561326,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0669,
+      "step": 8094
+    },
+    {
+      "epoch": 3.2341190571314424,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 8095
+    },
+    {
+      "epoch": 3.2345185777067518,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8096
+    },
+    {
+      "epoch": 3.2349180982820616,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 8097
+    },
+    {
+      "epoch": 3.2353176188573713,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 8098
+    },
+    {
+      "epoch": 3.2357171394326807,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8099
+    },
+    {
+      "epoch": 3.2361166600079905,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 8100
+    },
+    {
+      "epoch": 3.2365161805833003,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 8101
+    },
+    {
+      "epoch": 3.2369157011586096,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0251,
+      "step": 8102
+    },
+    {
+      "epoch": 3.2373152217339194,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 8103
+    },
+    {
+      "epoch": 3.2377147423092287,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 8104
+    },
+    {
+      "epoch": 3.2381142628845385,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 8105
+    },
+    {
+      "epoch": 3.2385137834598483,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 8106
+    },
+    {
+      "epoch": 3.2389133040351576,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 8107
+    },
+    {
+      "epoch": 3.2393128246104674,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0411,
+      "step": 8108
+    },
+    {
+      "epoch": 3.239712345185777,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8109
+    },
+    {
+      "epoch": 3.2401118657610866,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 8110
+    },
+    {
+      "epoch": 3.2405113863363963,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8111
+    },
+    {
+      "epoch": 3.240910906911706,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 8112
+    },
+    {
+      "epoch": 3.2413104274870155,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8113
+    },
+    {
+      "epoch": 3.2417099480623253,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8114
+    },
+    {
+      "epoch": 3.2421094686376346,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 8115
+    },
+    {
+      "epoch": 3.2425089892129444,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 8116
+    },
+    {
+      "epoch": 3.242908509788254,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8117
+    },
+    {
+      "epoch": 3.2433080303635635,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 8118
+    },
+    {
+      "epoch": 3.2437075509388733,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 8119
+    },
+    {
+      "epoch": 3.244107071514183,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 8120
+    },
+    {
+      "epoch": 3.2445065920894924,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 8121
+    },
+    {
+      "epoch": 3.2449061126648022,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 8122
+    },
+    {
+      "epoch": 3.245305633240112,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 8123
+    },
+    {
+      "epoch": 3.2457051538154214,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 8124
+    },
+    {
+      "epoch": 3.246104674390731,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 8125
+    },
+    {
+      "epoch": 3.246504194966041,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 8126
+    },
+    {
+      "epoch": 3.2469037155413503,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8127
+    },
+    {
+      "epoch": 3.24730323611666,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8128
+    },
+    {
+      "epoch": 3.24770275669197,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 8129
+    },
+    {
+      "epoch": 3.248102277267279,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 8130
+    },
+    {
+      "epoch": 3.248501797842589,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8131
+    },
+    {
+      "epoch": 3.2489013184178983,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8132
+    },
+    {
+      "epoch": 3.249300838993208,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8133
+    },
+    {
+      "epoch": 3.249700359568518,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 8134
+    },
+    {
+      "epoch": 3.2500998801438272,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 8135
+    },
+    {
+      "epoch": 3.250499400719137,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8136
+    },
+    {
+      "epoch": 3.250898921294447,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 8137
+    },
+    {
+      "epoch": 3.251298441869756,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8138
+    },
+    {
+      "epoch": 3.251697962445066,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0313,
+      "step": 8139
+    },
+    {
+      "epoch": 3.2520974830203757,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 8140
+    },
+    {
+      "epoch": 3.252497003595685,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8141
+    },
+    {
+      "epoch": 3.252896524170995,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 8142
+    },
+    {
+      "epoch": 3.2532960447463046,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8143
+    },
+    {
+      "epoch": 3.253695565321614,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 8144
+    },
+    {
+      "epoch": 3.2540950858969238,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 8145
+    },
+    {
+      "epoch": 3.2544946064722335,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 8146
+    },
+    {
+      "epoch": 3.254894127047543,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 8147
+    },
+    {
+      "epoch": 3.2552936476228527,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8148
+    },
+    {
+      "epoch": 3.255693168198162,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8149
+    },
+    {
+      "epoch": 3.256092688773472,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 8150
+    },
+    {
+      "epoch": 3.2564922093487816,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 8151
+    },
+    {
+      "epoch": 3.256891729924091,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8152
+    },
+    {
+      "epoch": 3.2572912504994007,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8153
+    },
+    {
+      "epoch": 3.2576907710747105,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 8154
+    },
+    {
+      "epoch": 3.25809029165002,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8155
+    },
+    {
+      "epoch": 3.2584898122253296,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 8156
+    },
+    {
+      "epoch": 3.2588893328006394,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 8157
+    },
+    {
+      "epoch": 3.2592888533759488,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 8158
+    },
+    {
+      "epoch": 3.2596883739512585,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 8159
+    },
+    {
+      "epoch": 3.260087894526568,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0415,
+      "step": 8160
+    },
+    {
+      "epoch": 3.2604874151018777,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 8161
+    },
+    {
+      "epoch": 3.2608869356771875,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 8162
+    },
+    {
+      "epoch": 3.2612864562524972,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8163
+    },
+    {
+      "epoch": 3.2616859768278066,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 8164
+    },
+    {
+      "epoch": 3.2620854974031164,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 8165
+    },
+    {
+      "epoch": 3.2624850179784257,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 8166
+    },
+    {
+      "epoch": 3.2628845385537355,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 8167
+    },
+    {
+      "epoch": 3.2632840591290453,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8168
+    },
+    {
+      "epoch": 3.2636835797043546,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 8169
+    },
+    {
+      "epoch": 3.2640831002796644,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8170
+    },
+    {
+      "epoch": 3.264482620854974,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 8171
+    },
+    {
+      "epoch": 3.2648821414302835,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 8172
+    },
+    {
+      "epoch": 3.2652816620055933,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 8173
+    },
+    {
+      "epoch": 3.265681182580903,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8174
+    },
+    {
+      "epoch": 3.2660807031562125,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 8175
+    },
+    {
+      "epoch": 3.2664802237315222,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 8176
+    },
+    {
+      "epoch": 3.2668797443068316,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 8177
+    },
+    {
+      "epoch": 3.2672792648821414,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 8178
+    },
+    {
+      "epoch": 3.267678785457451,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 8179
+    },
+    {
+      "epoch": 3.2680783060327605,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 8180
+    },
+    {
+      "epoch": 3.2684778266080703,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 8181
+    },
+    {
+      "epoch": 3.26887734718338,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 8182
+    },
+    {
+      "epoch": 3.2692768677586894,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 8183
+    },
+    {
+      "epoch": 3.269676388333999,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8184
+    },
+    {
+      "epoch": 3.270075908909309,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 8185
+    },
+    {
+      "epoch": 3.2704754294846183,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 8186
+    },
+    {
+      "epoch": 3.270874950059928,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0678,
+      "step": 8187
+    },
+    {
+      "epoch": 3.271274470635238,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0675,
+      "step": 8188
+    },
+    {
+      "epoch": 3.2716739912105472,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 8189
+    },
+    {
+      "epoch": 3.272073511785857,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 8190
+    },
+    {
+      "epoch": 3.272473032361167,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 8191
+    },
+    {
+      "epoch": 3.272872552936476,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0375,
+      "step": 8192
+    },
+    {
+      "epoch": 3.273272073511786,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 8193
+    },
+    {
+      "epoch": 3.2736715940870953,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8194
+    },
+    {
+      "epoch": 3.274071114662405,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 8195
+    },
+    {
+      "epoch": 3.274470635237715,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 8196
+    },
+    {
+      "epoch": 3.274870155813024,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 8197
+    },
+    {
+      "epoch": 3.275269676388334,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 8198
+    },
+    {
+      "epoch": 3.275669196963644,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 8199
+    },
+    {
+      "epoch": 3.276068717538953,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 8200
+    },
+    {
+      "epoch": 3.276468238114263,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 8201
+    },
+    {
+      "epoch": 3.2768677586895727,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8202
+    },
+    {
+      "epoch": 3.277267279264882,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 8203
+    },
+    {
+      "epoch": 3.277666799840192,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 8204
+    },
+    {
+      "epoch": 3.278066320415501,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8205
+    },
+    {
+      "epoch": 3.278465840990811,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 8206
+    },
+    {
+      "epoch": 3.2788653615661207,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.035,
+      "step": 8207
+    },
+    {
+      "epoch": 3.2792648821414305,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 8208
+    },
+    {
+      "epoch": 3.27966440271674,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8209
+    },
+    {
+      "epoch": 3.2800639232920497,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 8210
+    },
+    {
+      "epoch": 3.280463443867359,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 8211
+    },
+    {
+      "epoch": 3.280862964442669,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 8212
+    },
+    {
+      "epoch": 3.2812624850179786,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 8213
+    },
+    {
+      "epoch": 3.281662005593288,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 8214
+    },
+    {
+      "epoch": 3.2820615261685977,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 8215
+    },
+    {
+      "epoch": 3.2824610467439075,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 8216
+    },
+    {
+      "epoch": 3.282860567319217,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 8217
+    },
+    {
+      "epoch": 3.2832600878945266,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8218
+    },
+    {
+      "epoch": 3.2836596084698364,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8219
+    },
+    {
+      "epoch": 3.2840591290451457,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 8220
+    },
+    {
+      "epoch": 3.2844586496204555,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 8221
+    },
+    {
+      "epoch": 3.284858170195765,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.036,
+      "step": 8222
+    },
+    {
+      "epoch": 3.2852576907710747,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 8223
+    },
+    {
+      "epoch": 3.2856572113463844,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8224
+    },
+    {
+      "epoch": 3.286056731921694,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 8225
+    },
+    {
+      "epoch": 3.2864562524970036,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 8226
+    },
+    {
+      "epoch": 3.2868557730723134,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0402,
+      "step": 8227
+    },
+    {
+      "epoch": 3.2872552936476227,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 8228
+    },
+    {
+      "epoch": 3.2876548142229325,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 8229
+    },
+    {
+      "epoch": 3.2880543347982423,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8230
+    },
+    {
+      "epoch": 3.2884538553735516,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 8231
+    },
+    {
+      "epoch": 3.2888533759488614,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 8232
+    },
+    {
+      "epoch": 3.289252896524171,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8233
+    },
+    {
+      "epoch": 3.2896524170994805,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8234
+    },
+    {
+      "epoch": 3.2900519376747903,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8235
+    },
+    {
+      "epoch": 3.2904514582501,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 8236
+    },
+    {
+      "epoch": 3.2908509788254094,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0405,
+      "step": 8237
+    },
+    {
+      "epoch": 3.2912504994007192,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 8238
+    },
+    {
+      "epoch": 3.2916500199760286,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 8239
+    },
+    {
+      "epoch": 3.2920495405513384,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8240
+    },
+    {
+      "epoch": 3.292449061126648,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 8241
+    },
+    {
+      "epoch": 3.2928485817019575,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8242
+    },
+    {
+      "epoch": 3.2932481022772673,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 8243
+    },
+    {
+      "epoch": 3.293647622852577,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 8244
+    },
+    {
+      "epoch": 3.2940471434278864,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 8245
+    },
+    {
+      "epoch": 3.294446664003196,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8246
+    },
+    {
+      "epoch": 3.294846184578506,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 8247
+    },
+    {
+      "epoch": 3.2952457051538153,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 8248
+    },
+    {
+      "epoch": 3.295645225729125,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 8249
+    },
+    {
+      "epoch": 3.2960447463044344,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 8250
+    },
+    {
+      "epoch": 3.2964442668797442,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8251
+    },
+    {
+      "epoch": 3.296843787455054,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 8252
+    },
+    {
+      "epoch": 3.297243308030364,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 8253
+    },
+    {
+      "epoch": 3.297642828605673,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8254
+    },
+    {
+      "epoch": 3.298042349180983,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 8255
+    },
+    {
+      "epoch": 3.2984418697562923,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0605,
+      "step": 8256
+    },
+    {
+      "epoch": 3.298841390331602,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8257
+    },
+    {
+      "epoch": 3.299240910906912,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 8258
+    },
+    {
+      "epoch": 3.299640431482221,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 8259
+    },
+    {
+      "epoch": 3.300039952057531,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 8260
+    },
+    {
+      "epoch": 3.3004394726328408,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0701,
+      "step": 8261
+    },
+    {
+      "epoch": 3.30083899320815,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8262
+    },
+    {
+      "epoch": 3.30123851378346,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 8263
+    },
+    {
+      "epoch": 3.3016380343587697,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 8264
+    },
+    {
+      "epoch": 3.302037554934079,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.039,
+      "step": 8265
+    },
+    {
+      "epoch": 3.302437075509389,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 8266
+    },
+    {
+      "epoch": 3.302836596084698,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 8267
+    },
+    {
+      "epoch": 3.303236116660008,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 8268
+    },
+    {
+      "epoch": 3.3036356372353177,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 8269
+    },
+    {
+      "epoch": 3.304035157810627,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 8270
+    },
+    {
+      "epoch": 3.304434678385937,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 8271
+    },
+    {
+      "epoch": 3.3048341989612466,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8272
+    },
+    {
+      "epoch": 3.305233719536556,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8273
+    },
+    {
+      "epoch": 3.3056332401118658,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 8274
+    },
+    {
+      "epoch": 3.3060327606871756,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 8275
+    },
+    {
+      "epoch": 3.306432281262485,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8276
+    },
+    {
+      "epoch": 3.3068318018377947,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 8277
+    },
+    {
+      "epoch": 3.3072313224131045,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8278
+    },
+    {
+      "epoch": 3.307630842988414,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 8279
+    },
+    {
+      "epoch": 3.3080303635637236,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 8280
+    },
+    {
+      "epoch": 3.3084298841390334,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 8281
+    },
+    {
+      "epoch": 3.3088294047143427,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 8282
+    },
+    {
+      "epoch": 3.3092289252896525,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 8283
+    },
+    {
+      "epoch": 3.309628445864962,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 8284
+    },
+    {
+      "epoch": 3.3100279664402716,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 8285
+    },
+    {
+      "epoch": 3.3104274870155814,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 8286
+    },
+    {
+      "epoch": 3.3108270075908908,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 8287
+    },
+    {
+      "epoch": 3.3112265281662006,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 8288
+    },
+    {
+      "epoch": 3.3116260487415103,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 8289
+    },
+    {
+      "epoch": 3.3120255693168197,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 8290
+    },
+    {
+      "epoch": 3.3124250898921295,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 8291
+    },
+    {
+      "epoch": 3.3128246104674393,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 8292
+    },
+    {
+      "epoch": 3.3132241310427486,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 8293
+    },
+    {
+      "epoch": 3.3136236516180584,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0602,
+      "step": 8294
+    },
+    {
+      "epoch": 3.3140231721933677,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 8295
+    },
+    {
+      "epoch": 3.3144226927686775,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8296
+    },
+    {
+      "epoch": 3.3148222133439873,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 8297
+    },
+    {
+      "epoch": 3.315221733919297,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 8298
+    },
+    {
+      "epoch": 3.3156212544946064,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8299
+    },
+    {
+      "epoch": 3.316020775069916,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 8300
+    },
+    {
+      "epoch": 3.3164202956452256,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 8301
+    },
+    {
+      "epoch": 3.3168198162205353,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8302
+    },
+    {
+      "epoch": 3.317219336795845,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 8303
+    },
+    {
+      "epoch": 3.3176188573711545,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 8304
+    },
+    {
+      "epoch": 3.3180183779464643,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 8305
+    },
+    {
+      "epoch": 3.318417898521774,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8306
+    },
+    {
+      "epoch": 3.3188174190970834,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 8307
+    },
+    {
+      "epoch": 3.319216939672393,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 8308
+    },
+    {
+      "epoch": 3.319616460247703,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 8309
+    },
+    {
+      "epoch": 3.3200159808230123,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0418,
+      "step": 8310
+    },
+    {
+      "epoch": 3.320415501398322,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 8311
+    },
+    {
+      "epoch": 3.3208150219736314,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 8312
+    },
+    {
+      "epoch": 3.321214542548941,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8313
+    },
+    {
+      "epoch": 3.321614063124251,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 8314
+    },
+    {
+      "epoch": 3.3220135836995603,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8315
+    },
+    {
+      "epoch": 3.32241310427487,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 8316
+    },
+    {
+      "epoch": 3.32281262485018,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8317
+    },
+    {
+      "epoch": 3.3232121454254893,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 8318
+    },
+    {
+      "epoch": 3.323611666000799,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 8319
+    },
+    {
+      "epoch": 3.324011186576109,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 8320
+    },
+    {
+      "epoch": 3.324410707151418,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 8321
+    },
+    {
+      "epoch": 3.324810227726728,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8322
+    },
+    {
+      "epoch": 3.3252097483020377,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8323
+    },
+    {
+      "epoch": 3.325609268877347,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 8324
+    },
+    {
+      "epoch": 3.326008789452657,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8325
+    },
+    {
+      "epoch": 3.3264083100279667,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8326
+    },
+    {
+      "epoch": 3.326807830603276,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 8327
+    },
+    {
+      "epoch": 3.327207351178586,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8328
+    },
+    {
+      "epoch": 3.327606871753895,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8329
+    },
+    {
+      "epoch": 3.328006392329205,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 8330
+    },
+    {
+      "epoch": 3.3284059129045147,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0397,
+      "step": 8331
+    },
+    {
+      "epoch": 3.328805433479824,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 8332
+    },
+    {
+      "epoch": 3.329204954055134,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 8333
+    },
+    {
+      "epoch": 3.3296044746304436,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 8334
+    },
+    {
+      "epoch": 3.330003995205753,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 8335
+    },
+    {
+      "epoch": 3.3304035157810628,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8336
+    },
+    {
+      "epoch": 3.3308030363563725,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 8337
+    },
+    {
+      "epoch": 3.331202556931682,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.06,
+      "step": 8338
+    },
+    {
+      "epoch": 3.3316020775069917,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8339
+    },
+    {
+      "epoch": 3.332001598082301,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 8340
+    },
+    {
+      "epoch": 3.332401118657611,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 8341
+    },
+    {
+      "epoch": 3.3328006392329206,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 8342
+    },
+    {
+      "epoch": 3.3332001598082304,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 8343
+    },
+    {
+      "epoch": 3.3335996803835397,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 8344
+    },
+    {
+      "epoch": 3.3339992009588495,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8345
+    },
+    {
+      "epoch": 3.334398721534159,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8346
+    },
+    {
+      "epoch": 3.3347982421094686,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 8347
+    },
+    {
+      "epoch": 3.3351977626847784,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 8348
+    },
+    {
+      "epoch": 3.3355972832600878,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 8349
+    },
+    {
+      "epoch": 3.3359968038353975,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 8350
+    },
+    {
+      "epoch": 3.3363963244107073,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 8351
+    },
+    {
+      "epoch": 3.3367958449860167,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 8352
+    },
+    {
+      "epoch": 3.3371953655613265,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8353
+    },
+    {
+      "epoch": 3.3375948861366362,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 8354
+    },
+    {
+      "epoch": 3.3379944067119456,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 8355
+    },
+    {
+      "epoch": 3.3383939272872554,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 8356
+    },
+    {
+      "epoch": 3.3387934478625647,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 8357
+    },
+    {
+      "epoch": 3.3391929684378745,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8358
+    },
+    {
+      "epoch": 3.3395924890131843,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 8359
+    },
+    {
+      "epoch": 3.3399920095884936,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8360
+    },
+    {
+      "epoch": 3.3403915301638034,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 8361
+    },
+    {
+      "epoch": 3.340791050739113,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 8362
+    },
+    {
+      "epoch": 3.3411905713144225,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0354,
+      "step": 8363
+    },
+    {
+      "epoch": 3.3415900918897323,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 8364
+    },
+    {
+      "epoch": 3.341989612465042,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 8365
+    },
+    {
+      "epoch": 3.3423891330403515,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8366
+    },
+    {
+      "epoch": 3.3427886536156612,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8367
+    },
+    {
+      "epoch": 3.343188174190971,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 8368
+    },
+    {
+      "epoch": 3.3435876947662804,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 8369
+    },
+    {
+      "epoch": 3.34398721534159,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8370
+    },
+    {
+      "epoch": 3.3443867359169,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 8371
+    },
+    {
+      "epoch": 3.3447862564922093,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0668,
+      "step": 8372
+    },
+    {
+      "epoch": 3.345185777067519,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 8373
+    },
+    {
+      "epoch": 3.3455852976428284,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 8374
+    },
+    {
+      "epoch": 3.345984818218138,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 8375
+    },
+    {
+      "epoch": 3.346384338793448,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 8376
+    },
+    {
+      "epoch": 3.3467838593687573,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 8377
+    },
+    {
+      "epoch": 3.347183379944067,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 8378
+    },
+    {
+      "epoch": 3.347582900519377,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 8379
+    },
+    {
+      "epoch": 3.3479824210946862,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 8380
+    },
+    {
+      "epoch": 3.348381941669996,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 8381
+    },
+    {
+      "epoch": 3.348781462245306,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 8382
+    },
+    {
+      "epoch": 3.349180982820615,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 8383
+    },
+    {
+      "epoch": 3.349580503395925,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8384
+    },
+    {
+      "epoch": 3.3499800239712343,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8385
+    },
+    {
+      "epoch": 3.350379544546544,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0413,
+      "step": 8386
+    },
+    {
+      "epoch": 3.350779065121854,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 8387
+    },
+    {
+      "epoch": 3.3511785856971636,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8388
+    },
+    {
+      "epoch": 3.351578106272473,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 8389
+    },
+    {
+      "epoch": 3.3519776268477828,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 8390
+    },
+    {
+      "epoch": 3.352377147423092,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 8391
+    },
+    {
+      "epoch": 3.352776667998402,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 8392
+    },
+    {
+      "epoch": 3.3531761885737117,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8393
+    },
+    {
+      "epoch": 3.353575709149021,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 8394
+    },
+    {
+      "epoch": 3.353975229724331,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 8395
+    },
+    {
+      "epoch": 3.3543747502996406,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8396
+    },
+    {
+      "epoch": 3.35477427087495,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 8397
+    },
+    {
+      "epoch": 3.3551737914502597,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 8398
+    },
+    {
+      "epoch": 3.3555733120255695,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8399
+    },
+    {
+      "epoch": 3.355972832600879,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8400
+    },
+    {
+      "epoch": 3.3563723531761886,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8401
+    },
+    {
+      "epoch": 3.356771873751498,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 8402
+    },
+    {
+      "epoch": 3.357171394326808,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8403
+    },
+    {
+      "epoch": 3.3575709149021176,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 8404
+    },
+    {
+      "epoch": 3.357970435477427,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8405
+    },
+    {
+      "epoch": 3.3583699560527367,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 8406
+    },
+    {
+      "epoch": 3.3587694766280465,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 8407
+    },
+    {
+      "epoch": 3.359168997203356,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 8408
+    },
+    {
+      "epoch": 3.3595685177786656,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 8409
+    },
+    {
+      "epoch": 3.3599680383539754,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 8410
+    },
+    {
+      "epoch": 3.3603675589292847,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8411
+    },
+    {
+      "epoch": 3.3607670795045945,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8412
+    },
+    {
+      "epoch": 3.361166600079904,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0363,
+      "step": 8413
+    },
+    {
+      "epoch": 3.3615661206552137,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 8414
+    },
+    {
+      "epoch": 3.3619656412305234,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 8415
+    },
+    {
+      "epoch": 3.3623651618058332,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 8416
+    },
+    {
+      "epoch": 3.3627646823811426,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 8417
+    },
+    {
+      "epoch": 3.3631642029564524,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 8418
+    },
+    {
+      "epoch": 3.3635637235317617,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 8419
+    },
+    {
+      "epoch": 3.3639632441070715,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 8420
+    },
+    {
+      "epoch": 3.3643627646823813,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 8421
+    },
+    {
+      "epoch": 3.3647622852576906,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8422
+    },
+    {
+      "epoch": 3.3651618058330004,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 8423
+    },
+    {
+      "epoch": 3.36556132640831,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8424
+    },
+    {
+      "epoch": 3.3659608469836195,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 8425
+    },
+    {
+      "epoch": 3.3663603675589293,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 8426
+    },
+    {
+      "epoch": 3.366759888134239,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 8427
+    },
+    {
+      "epoch": 3.3671594087095484,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 8428
+    },
+    {
+      "epoch": 3.3675589292848582,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 8429
+    },
+    {
+      "epoch": 3.3679584498601676,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 8430
+    },
+    {
+      "epoch": 3.3683579704354774,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 8431
+    },
+    {
+      "epoch": 3.368757491010787,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8432
+    },
+    {
+      "epoch": 3.369157011586097,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 8433
+    },
+    {
+      "epoch": 3.3695565321614063,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 8434
+    },
+    {
+      "epoch": 3.369956052736716,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 8435
+    },
+    {
+      "epoch": 3.3703555733120254,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.038,
+      "step": 8436
+    },
+    {
+      "epoch": 3.370755093887335,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 8437
+    },
+    {
+      "epoch": 3.371154614462645,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 8438
+    },
+    {
+      "epoch": 3.3715541350379543,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0322,
+      "step": 8439
+    },
+    {
+      "epoch": 3.371953655613264,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8440
+    },
+    {
+      "epoch": 3.372353176188574,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8441
+    },
+    {
+      "epoch": 3.3727526967638832,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 8442
+    },
+    {
+      "epoch": 3.373152217339193,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8443
+    },
+    {
+      "epoch": 3.373551737914503,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 8444
+    },
+    {
+      "epoch": 3.373951258489812,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 8445
+    },
+    {
+      "epoch": 3.374350779065122,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 8446
+    },
+    {
+      "epoch": 3.3747502996404313,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 8447
+    },
+    {
+      "epoch": 3.375149820215741,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8448
+    },
+    {
+      "epoch": 3.375549340791051,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 8449
+    },
+    {
+      "epoch": 3.37594886136636,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 8450
+    },
+    {
+      "epoch": 3.37634838194167,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8451
+    },
+    {
+      "epoch": 3.3767479025169798,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 8452
+    },
+    {
+      "epoch": 3.377147423092289,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 8453
+    },
+    {
+      "epoch": 3.377546943667599,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 8454
+    },
+    {
+      "epoch": 3.3779464642429087,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 8455
+    },
+    {
+      "epoch": 3.378345984818218,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 8456
+    },
+    {
+      "epoch": 3.378745505393528,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 8457
+    },
+    {
+      "epoch": 3.379145025968837,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 8458
+    },
+    {
+      "epoch": 3.379544546544147,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8459
+    },
+    {
+      "epoch": 3.3799440671194567,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 8460
+    },
+    {
+      "epoch": 3.3803435876947665,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8461
+    },
+    {
+      "epoch": 3.380743108270076,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 8462
+    },
+    {
+      "epoch": 3.3811426288453856,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0698,
+      "step": 8463
+    },
+    {
+      "epoch": 3.381542149420695,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 8464
+    },
+    {
+      "epoch": 3.3819416699960048,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8465
+    },
+    {
+      "epoch": 3.3823411905713145,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 8466
+    },
+    {
+      "epoch": 3.382740711146624,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 8467
+    },
+    {
+      "epoch": 3.3831402317219337,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8468
+    },
+    {
+      "epoch": 3.3835397522972435,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 8469
+    },
+    {
+      "epoch": 3.383939272872553,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 8470
+    },
+    {
+      "epoch": 3.3843387934478626,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8471
+    },
+    {
+      "epoch": 3.3847383140231724,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 8472
+    },
+    {
+      "epoch": 3.3851378345984817,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 8473
+    },
+    {
+      "epoch": 3.3855373551737915,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0377,
+      "step": 8474
+    },
+    {
+      "epoch": 3.385936875749101,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 8475
+    },
+    {
+      "epoch": 3.3863363963244106,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 8476
+    },
+    {
+      "epoch": 3.3867359168997204,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 8477
+    },
+    {
+      "epoch": 3.38713543747503,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8478
+    },
+    {
+      "epoch": 3.3875349580503396,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 8479
+    },
+    {
+      "epoch": 3.3879344786256493,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 8480
+    },
+    {
+      "epoch": 3.3883339992009587,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 8481
+    },
+    {
+      "epoch": 3.3887335197762685,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 8482
+    },
+    {
+      "epoch": 3.3891330403515783,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 8483
+    },
+    {
+      "epoch": 3.3895325609268876,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8484
+    },
+    {
+      "epoch": 3.3899320815021974,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8485
+    },
+    {
+      "epoch": 3.390331602077507,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 8486
+    },
+    {
+      "epoch": 3.3907311226528165,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.043,
+      "step": 8487
+    },
+    {
+      "epoch": 3.3911306432281263,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8488
+    },
+    {
+      "epoch": 3.391530163803436,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 8489
+    },
+    {
+      "epoch": 3.3919296843787454,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 8490
+    },
+    {
+      "epoch": 3.392329204954055,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 8491
+    },
+    {
+      "epoch": 3.3927287255293646,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 8492
+    },
+    {
+      "epoch": 3.3931282461046743,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 8493
+    },
+    {
+      "epoch": 3.393527766679984,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 8494
+    },
+    {
+      "epoch": 3.3939272872552935,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 8495
+    },
+    {
+      "epoch": 3.3943268078306033,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 8496
+    },
+    {
+      "epoch": 3.394726328405913,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8497
+    },
+    {
+      "epoch": 3.3951258489812224,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 8498
+    },
+    {
+      "epoch": 3.395525369556532,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8499
+    },
+    {
+      "epoch": 3.395924890131842,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8500
+    },
+    {
+      "epoch": 3.3963244107071513,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 8501
+    },
+    {
+      "epoch": 3.396723931282461,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0404,
+      "step": 8502
+    },
+    {
+      "epoch": 3.3971234518577704,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0408,
+      "step": 8503
+    },
+    {
+      "epoch": 3.39752297243308,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8504
+    },
+    {
+      "epoch": 3.39792249300839,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0373,
+      "step": 8505
+    },
+    {
+      "epoch": 3.3983220135837,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 8506
+    },
+    {
+      "epoch": 3.398721534159009,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8507
+    },
+    {
+      "epoch": 3.399121054734319,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 8508
+    },
+    {
+      "epoch": 3.3995205753096283,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 8509
+    },
+    {
+      "epoch": 3.399920095884938,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0408,
+      "step": 8510
+    },
+    {
+      "epoch": 3.400319616460248,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 8511
+    },
+    {
+      "epoch": 3.400719137035557,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 8512
+    },
+    {
+      "epoch": 3.401118657610867,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 8513
+    },
+    {
+      "epoch": 3.4015181781861767,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 8514
+    },
+    {
+      "epoch": 3.401917698761486,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 8515
+    },
+    {
+      "epoch": 3.402317219336796,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 8516
+    },
+    {
+      "epoch": 3.4027167399121057,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 8517
+    },
+    {
+      "epoch": 3.403116260487415,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 8518
+    },
+    {
+      "epoch": 3.403515781062725,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0366,
+      "step": 8519
+    },
+    {
+      "epoch": 3.403915301638034,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 8520
+    },
+    {
+      "epoch": 3.404314822213344,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8521
+    },
+    {
+      "epoch": 3.4047143427886537,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 8522
+    },
+    {
+      "epoch": 3.4051138633639635,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8523
+    },
+    {
+      "epoch": 3.405513383939273,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 8524
+    },
+    {
+      "epoch": 3.4059129045145826,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 8525
+    },
+    {
+      "epoch": 3.406312425089892,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8526
+    },
+    {
+      "epoch": 3.4067119456652017,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8527
+    },
+    {
+      "epoch": 3.4071114662405115,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 8528
+    },
+    {
+      "epoch": 3.407510986815821,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 8529
+    },
+    {
+      "epoch": 3.4079105073911307,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 8530
+    },
+    {
+      "epoch": 3.4083100279664404,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8531
+    },
+    {
+      "epoch": 3.40870954854175,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0398,
+      "step": 8532
+    },
+    {
+      "epoch": 3.4091090691170596,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 8533
+    },
+    {
+      "epoch": 3.4095085896923694,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0403,
+      "step": 8534
+    },
+    {
+      "epoch": 3.4099081102676787,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 8535
+    },
+    {
+      "epoch": 3.4103076308429885,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 8536
+    },
+    {
+      "epoch": 3.410707151418298,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 8537
+    },
+    {
+      "epoch": 3.4111066719936076,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8538
+    },
+    {
+      "epoch": 3.4115061925689174,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 8539
+    },
+    {
+      "epoch": 3.4119057131442267,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 8540
+    },
+    {
+      "epoch": 3.4123052337195365,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 8541
+    },
+    {
+      "epoch": 3.4127047542948463,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 8542
+    },
+    {
+      "epoch": 3.4131042748701557,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 8543
+    },
+    {
+      "epoch": 3.4135037954454654,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8544
+    },
+    {
+      "epoch": 3.4139033160207752,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 8545
+    },
+    {
+      "epoch": 3.4143028365960846,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 8546
+    },
+    {
+      "epoch": 3.4147023571713944,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8547
+    },
+    {
+      "epoch": 3.4151018777467037,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 8548
+    },
+    {
+      "epoch": 3.4155013983220135,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 8549
+    },
+    {
+      "epoch": 3.4159009188973233,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 8550
+    },
+    {
+      "epoch": 3.416300439472633,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 8551
+    },
+    {
+      "epoch": 3.4166999600479424,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 8552
+    },
+    {
+      "epoch": 3.417099480623252,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8553
+    },
+    {
+      "epoch": 3.4174990011985615,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8554
+    },
+    {
+      "epoch": 3.4178985217738713,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8555
+    },
+    {
+      "epoch": 3.418298042349181,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 8556
+    },
+    {
+      "epoch": 3.4186975629244905,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 8557
+    },
+    {
+      "epoch": 3.4190970834998002,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 8558
+    },
+    {
+      "epoch": 3.41949660407511,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 8559
+    },
+    {
+      "epoch": 3.4198961246504194,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8560
+    },
+    {
+      "epoch": 3.420295645225729,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8561
+    },
+    {
+      "epoch": 3.420695165801039,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 8562
+    },
+    {
+      "epoch": 3.4210946863763483,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 8563
+    },
+    {
+      "epoch": 3.421494206951658,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8564
+    },
+    {
+      "epoch": 3.4218937275269674,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8565
+    },
+    {
+      "epoch": 3.422293248102277,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 8566
+    },
+    {
+      "epoch": 3.422692768677587,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 8567
+    },
+    {
+      "epoch": 3.4230922892528968,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 8568
+    },
+    {
+      "epoch": 3.423491809828206,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 8569
+    },
+    {
+      "epoch": 3.423891330403516,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 8570
+    },
+    {
+      "epoch": 3.4242908509788252,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8571
+    },
+    {
+      "epoch": 3.424690371554135,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 8572
+    },
+    {
+      "epoch": 3.425089892129445,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 8573
+    },
+    {
+      "epoch": 3.425489412704754,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 8574
+    },
+    {
+      "epoch": 3.425888933280064,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 8575
+    },
+    {
+      "epoch": 3.4262884538553737,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 8576
+    },
+    {
+      "epoch": 3.426687974430683,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 8577
+    },
+    {
+      "epoch": 3.427087495005993,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 8578
+    },
+    {
+      "epoch": 3.4274870155813026,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 8579
+    },
+    {
+      "epoch": 3.427886536156612,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8580
+    },
+    {
+      "epoch": 3.4282860567319218,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 8581
+    },
+    {
+      "epoch": 3.428685577307231,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 8582
+    },
+    {
+      "epoch": 3.429085097882541,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8583
+    },
+    {
+      "epoch": 3.4294846184578507,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8584
+    },
+    {
+      "epoch": 3.42988413903316,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 8585
+    },
+    {
+      "epoch": 3.43028365960847,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8586
+    },
+    {
+      "epoch": 3.4306831801837796,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 8587
+    },
+    {
+      "epoch": 3.431082700759089,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8588
+    },
+    {
+      "epoch": 3.4314822213343987,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 8589
+    },
+    {
+      "epoch": 3.4318817419097085,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 8590
+    },
+    {
+      "epoch": 3.432281262485018,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 8591
+    },
+    {
+      "epoch": 3.4326807830603276,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0709,
+      "step": 8592
+    },
+    {
+      "epoch": 3.433080303635637,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8593
+    },
+    {
+      "epoch": 3.4334798242109468,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 8594
+    },
+    {
+      "epoch": 3.4338793447862566,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8595
+    },
+    {
+      "epoch": 3.4342788653615663,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 8596
+    },
+    {
+      "epoch": 3.4346783859368757,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 8597
+    },
+    {
+      "epoch": 3.4350779065121855,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 8598
+    },
+    {
+      "epoch": 3.435477427087495,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 8599
+    },
+    {
+      "epoch": 3.4358769476628046,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 8600
+    },
+    {
+      "epoch": 3.4362764682381144,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 8601
+    },
+    {
+      "epoch": 3.4366759888134237,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 8602
+    },
+    {
+      "epoch": 3.4370755093887335,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 8603
+    },
+    {
+      "epoch": 3.4374750299640433,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 8604
+    },
+    {
+      "epoch": 3.4378745505393526,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0691,
+      "step": 8605
+    },
+    {
+      "epoch": 3.4382740711146624,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 8606
+    },
+    {
+      "epoch": 3.438673591689972,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 8607
+    },
+    {
+      "epoch": 3.4390731122652816,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8608
+    },
+    {
+      "epoch": 3.4394726328405913,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 8609
+    },
+    {
+      "epoch": 3.4398721534159007,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8610
+    },
+    {
+      "epoch": 3.4402716739912105,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 8611
+    },
+    {
+      "epoch": 3.4406711945665203,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 8612
+    },
+    {
+      "epoch": 3.44107071514183,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 8613
+    },
+    {
+      "epoch": 3.4414702357171394,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 8614
+    },
+    {
+      "epoch": 3.441869756292449,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8615
+    },
+    {
+      "epoch": 3.4422692768677585,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8616
+    },
+    {
+      "epoch": 3.4426687974430683,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8617
+    },
+    {
+      "epoch": 3.443068318018378,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 8618
+    },
+    {
+      "epoch": 3.4434678385936874,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 8619
+    },
+    {
+      "epoch": 3.443867359168997,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 8620
+    },
+    {
+      "epoch": 3.444266879744307,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8621
+    },
+    {
+      "epoch": 3.4446664003196164,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8622
+    },
+    {
+      "epoch": 3.445065920894926,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0378,
+      "step": 8623
+    },
+    {
+      "epoch": 3.445465441470236,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 8624
+    },
+    {
+      "epoch": 3.4458649620455453,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8625
+    },
+    {
+      "epoch": 3.446264482620855,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 8626
+    },
+    {
+      "epoch": 3.4466640031961644,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.059,
+      "step": 8627
+    },
+    {
+      "epoch": 3.447063523771474,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0354,
+      "step": 8628
+    },
+    {
+      "epoch": 3.447463044346784,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 8629
+    },
+    {
+      "epoch": 3.4478625649220933,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 8630
+    },
+    {
+      "epoch": 3.448262085497403,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 8631
+    },
+    {
+      "epoch": 3.448661606072713,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8632
+    },
+    {
+      "epoch": 3.4490611266480222,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 8633
+    },
+    {
+      "epoch": 3.449460647223332,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 8634
+    },
+    {
+      "epoch": 3.449860167798642,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 8635
+    },
+    {
+      "epoch": 3.450259688373951,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 8636
+    },
+    {
+      "epoch": 3.450659208949261,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 8637
+    },
+    {
+      "epoch": 3.4510587295245703,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0413,
+      "step": 8638
+    },
+    {
+      "epoch": 3.45145825009988,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 8639
+    },
+    {
+      "epoch": 3.45185777067519,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 8640
+    },
+    {
+      "epoch": 3.4522572912504996,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 8641
+    },
+    {
+      "epoch": 3.452656811825809,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 8642
+    },
+    {
+      "epoch": 3.4530563324011188,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0421,
+      "step": 8643
+    },
+    {
+      "epoch": 3.453455852976428,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 8644
+    },
+    {
+      "epoch": 3.453855373551738,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8645
+    },
+    {
+      "epoch": 3.4542548941270477,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.04,
+      "step": 8646
+    },
+    {
+      "epoch": 3.454654414702357,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 8647
+    },
+    {
+      "epoch": 3.455053935277667,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 8648
+    },
+    {
+      "epoch": 3.4554534558529766,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.068,
+      "step": 8649
+    },
+    {
+      "epoch": 3.455852976428286,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 8650
+    },
+    {
+      "epoch": 3.4562524970035957,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0667,
+      "step": 8651
+    },
+    {
+      "epoch": 3.4566520175789055,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8652
+    },
+    {
+      "epoch": 3.457051538154215,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 8653
+    },
+    {
+      "epoch": 3.4574510587295246,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 8654
+    },
+    {
+      "epoch": 3.457850579304834,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 8655
+    },
+    {
+      "epoch": 3.4582500998801438,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 8656
+    },
+    {
+      "epoch": 3.4586496204554535,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 8657
+    },
+    {
+      "epoch": 3.4590491410307633,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 8658
+    },
+    {
+      "epoch": 3.4594486616060727,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 8659
+    },
+    {
+      "epoch": 3.4598481821813825,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8660
+    },
+    {
+      "epoch": 3.460247702756692,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 8661
+    },
+    {
+      "epoch": 3.4606472233320016,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 8662
+    },
+    {
+      "epoch": 3.4610467439073114,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0646,
+      "step": 8663
+    },
+    {
+      "epoch": 3.4614462644826207,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8664
+    },
+    {
+      "epoch": 3.4618457850579305,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8665
+    },
+    {
+      "epoch": 3.4622453056332403,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 8666
+    },
+    {
+      "epoch": 3.4626448262085496,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 8667
+    },
+    {
+      "epoch": 3.4630443467838594,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 8668
+    },
+    {
+      "epoch": 3.463443867359169,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 8669
+    },
+    {
+      "epoch": 3.4638433879344785,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 8670
+    },
+    {
+      "epoch": 3.4642429085097883,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 8671
+    },
+    {
+      "epoch": 3.4646424290850977,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 8672
+    },
+    {
+      "epoch": 3.4650419496604075,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0395,
+      "step": 8673
+    },
+    {
+      "epoch": 3.4654414702357172,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 8674
+    },
+    {
+      "epoch": 3.4658409908110266,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0418,
+      "step": 8675
+    },
+    {
+      "epoch": 3.4662405113863364,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 8676
+    },
+    {
+      "epoch": 3.466640031961646,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 8677
+    },
+    {
+      "epoch": 3.4670395525369555,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 8678
+    },
+    {
+      "epoch": 3.4674390731122653,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8679
+    },
+    {
+      "epoch": 3.467838593687575,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8680
+    },
+    {
+      "epoch": 3.4682381142628844,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8681
+    },
+    {
+      "epoch": 3.468637634838194,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 8682
+    },
+    {
+      "epoch": 3.4690371554135035,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0357,
+      "step": 8683
+    },
+    {
+      "epoch": 3.4694366759888133,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8684
+    },
+    {
+      "epoch": 3.469836196564123,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.039,
+      "step": 8685
+    },
+    {
+      "epoch": 3.470235717139433,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 8686
+    },
+    {
+      "epoch": 3.4706352377147422,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8687
+    },
+    {
+      "epoch": 3.471034758290052,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 8688
+    },
+    {
+      "epoch": 3.4714342788653614,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 8689
+    },
+    {
+      "epoch": 3.471833799440671,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 8690
+    },
+    {
+      "epoch": 3.472233320015981,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 8691
+    },
+    {
+      "epoch": 3.4726328405912903,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 8692
+    },
+    {
+      "epoch": 3.4730323611666,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8693
+    },
+    {
+      "epoch": 3.47343188174191,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 8694
+    },
+    {
+      "epoch": 3.473831402317219,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0348,
+      "step": 8695
+    },
+    {
+      "epoch": 3.474230922892529,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 8696
+    },
+    {
+      "epoch": 3.474630443467839,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8697
+    },
+    {
+      "epoch": 3.475029964043148,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 8698
+    },
+    {
+      "epoch": 3.475429484618458,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 8699
+    },
+    {
+      "epoch": 3.4758290051937673,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 8700
+    },
+    {
+      "epoch": 3.476228525769077,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 8701
+    },
+    {
+      "epoch": 3.476628046344387,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 8702
+    },
+    {
+      "epoch": 3.4770275669196966,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 8703
+    },
+    {
+      "epoch": 3.477427087495006,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 8704
+    },
+    {
+      "epoch": 3.4778266080703157,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8705
+    },
+    {
+      "epoch": 3.478226128645625,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 8706
+    },
+    {
+      "epoch": 3.478625649220935,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 8707
+    },
+    {
+      "epoch": 3.4790251697962447,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0418,
+      "step": 8708
+    },
+    {
+      "epoch": 3.479424690371554,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 8709
+    },
+    {
+      "epoch": 3.479824210946864,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8710
+    },
+    {
+      "epoch": 3.4802237315221736,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 8711
+    },
+    {
+      "epoch": 3.480623252097483,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 8712
+    },
+    {
+      "epoch": 3.4810227726727927,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0607,
+      "step": 8713
+    },
+    {
+      "epoch": 3.4814222932481025,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 8714
+    },
+    {
+      "epoch": 3.481821813823412,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 8715
+    },
+    {
+      "epoch": 3.4822213343987216,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 8716
+    },
+    {
+      "epoch": 3.482620854974031,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0593,
+      "step": 8717
+    },
+    {
+      "epoch": 3.4830203755493407,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8718
+    },
+    {
+      "epoch": 3.4834198961246505,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 8719
+    },
+    {
+      "epoch": 3.48381941669996,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.072,
+      "step": 8720
+    },
+    {
+      "epoch": 3.4842189372752697,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 8721
+    },
+    {
+      "epoch": 3.4846184578505794,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 8722
+    },
+    {
+      "epoch": 3.485017978425889,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 8723
+    },
+    {
+      "epoch": 3.4854174990011986,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 8724
+    },
+    {
+      "epoch": 3.4858170195765084,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.04,
+      "step": 8725
+    },
+    {
+      "epoch": 3.4862165401518177,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 8726
+    },
+    {
+      "epoch": 3.4866160607271275,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 8727
+    },
+    {
+      "epoch": 3.487015581302437,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 8728
+    },
+    {
+      "epoch": 3.4874151018777466,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8729
+    },
+    {
+      "epoch": 3.4878146224530564,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 8730
+    },
+    {
+      "epoch": 3.488214143028366,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0415,
+      "step": 8731
+    },
+    {
+      "epoch": 3.4886136636036755,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 8732
+    },
+    {
+      "epoch": 3.4890131841789853,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 8733
+    },
+    {
+      "epoch": 3.4894127047542947,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 8734
+    },
+    {
+      "epoch": 3.4898122253296044,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 8735
+    },
+    {
+      "epoch": 3.4902117459049142,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 8736
+    },
+    {
+      "epoch": 3.4906112664802236,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 8737
+    },
+    {
+      "epoch": 3.4910107870555334,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 8738
+    },
+    {
+      "epoch": 3.491410307630843,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8739
+    },
+    {
+      "epoch": 3.4918098282061525,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 8740
+    },
+    {
+      "epoch": 3.4922093487814623,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8741
+    },
+    {
+      "epoch": 3.492608869356772,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 8742
+    },
+    {
+      "epoch": 3.4930083899320814,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0658,
+      "step": 8743
+    },
+    {
+      "epoch": 3.493407910507391,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 8744
+    },
+    {
+      "epoch": 3.4938074310827005,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8745
+    },
+    {
+      "epoch": 3.4942069516580103,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 8746
+    },
+    {
+      "epoch": 3.49460647223332,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 8747
+    },
+    {
+      "epoch": 3.49500599280863,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 8748
+    },
+    {
+      "epoch": 3.4954055133839392,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8749
+    },
+    {
+      "epoch": 3.495805033959249,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8750
+    },
+    {
+      "epoch": 3.4962045545345584,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8751
+    },
+    {
+      "epoch": 3.496604075109868,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 8752
+    },
+    {
+      "epoch": 3.497003595685178,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 8753
+    },
+    {
+      "epoch": 3.4974031162604873,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 8754
+    },
+    {
+      "epoch": 3.497802636835797,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0392,
+      "step": 8755
+    },
+    {
+      "epoch": 3.498202157411107,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 8756
+    },
+    {
+      "epoch": 3.498601677986416,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 8757
+    },
+    {
+      "epoch": 3.499001198561726,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 8758
+    },
+    {
+      "epoch": 3.4994007191370358,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 8759
+    },
+    {
+      "epoch": 3.499800239712345,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 8760
+    },
+    {
+      "epoch": 3.500199760287655,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 8761
+    },
+    {
+      "epoch": 3.5005992808629642,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8762
+    },
+    {
+      "epoch": 3.500998801438274,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 8763
+    },
+    {
+      "epoch": 3.501398322013584,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 8764
+    },
+    {
+      "epoch": 3.5017978425888936,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8765
+    },
+    {
+      "epoch": 3.502197363164203,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8766
+    },
+    {
+      "epoch": 3.5025968837395127,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0617,
+      "step": 8767
+    },
+    {
+      "epoch": 3.502996404314822,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 8768
+    },
+    {
+      "epoch": 3.503395924890132,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 8769
+    },
+    {
+      "epoch": 3.5037954454654416,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 8770
+    },
+    {
+      "epoch": 3.504194966040751,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 8771
+    },
+    {
+      "epoch": 3.5045944866160608,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8772
+    },
+    {
+      "epoch": 3.50499400719137,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 8773
+    },
+    {
+      "epoch": 3.50539352776668,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0411,
+      "step": 8774
+    },
+    {
+      "epoch": 3.5057930483419897,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 8775
+    },
+    {
+      "epoch": 3.5061925689172995,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8776
+    },
+    {
+      "epoch": 3.506592089492609,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8777
+    },
+    {
+      "epoch": 3.5069916100679186,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 8778
+    },
+    {
+      "epoch": 3.507391130643228,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0299,
+      "step": 8779
+    },
+    {
+      "epoch": 3.5077906512185377,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 8780
+    },
+    {
+      "epoch": 3.5081901717938475,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 8781
+    },
+    {
+      "epoch": 3.508589692369157,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0427,
+      "step": 8782
+    },
+    {
+      "epoch": 3.5089892129444666,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 8783
+    },
+    {
+      "epoch": 3.5093887335197764,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8784
+    },
+    {
+      "epoch": 3.5097882540950858,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 8785
+    },
+    {
+      "epoch": 3.5101877746703956,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0652,
+      "step": 8786
+    },
+    {
+      "epoch": 3.5105872952457053,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 8787
+    },
+    {
+      "epoch": 3.5109868158210147,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 8788
+    },
+    {
+      "epoch": 3.5113863363963245,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 8789
+    },
+    {
+      "epoch": 3.511785856971634,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 8790
+    },
+    {
+      "epoch": 3.5121853775469436,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 8791
+    },
+    {
+      "epoch": 3.5125848981222534,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 8792
+    },
+    {
+      "epoch": 3.512984418697563,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0375,
+      "step": 8793
+    },
+    {
+      "epoch": 3.5133839392728725,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 8794
+    },
+    {
+      "epoch": 3.5137834598481823,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 8795
+    },
+    {
+      "epoch": 3.5141829804234916,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 8796
+    },
+    {
+      "epoch": 3.5145825009988014,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 8797
+    },
+    {
+      "epoch": 3.514982021574111,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0376,
+      "step": 8798
+    },
+    {
+      "epoch": 3.5153815421494206,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 8799
+    },
+    {
+      "epoch": 3.5157810627247303,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 8800
+    },
+    {
+      "epoch": 3.5161805833000397,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 8801
+    },
+    {
+      "epoch": 3.5165801038753495,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 8802
+    },
+    {
+      "epoch": 3.5169796244506593,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 8803
+    },
+    {
+      "epoch": 3.517379145025969,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 8804
+    },
+    {
+      "epoch": 3.5177786656012784,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8805
+    },
+    {
+      "epoch": 3.518178186176588,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 8806
+    },
+    {
+      "epoch": 3.5185777067518975,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 8807
+    },
+    {
+      "epoch": 3.5189772273272073,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 8808
+    },
+    {
+      "epoch": 3.519376747902517,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8809
+    },
+    {
+      "epoch": 3.519776268477827,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8810
+    },
+    {
+      "epoch": 3.520175789053136,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8811
+    },
+    {
+      "epoch": 3.520575309628446,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 8812
+    },
+    {
+      "epoch": 3.5209748302037553,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 8813
+    },
+    {
+      "epoch": 3.521374350779065,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 8814
+    },
+    {
+      "epoch": 3.521773871354375,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 8815
+    },
+    {
+      "epoch": 3.5221733919296843,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0327,
+      "step": 8816
+    },
+    {
+      "epoch": 3.522572912504994,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 8817
+    },
+    {
+      "epoch": 3.5229724330803034,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 8818
+    },
+    {
+      "epoch": 3.523371953655613,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8819
+    },
+    {
+      "epoch": 3.523771474230923,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 8820
+    },
+    {
+      "epoch": 3.5241709948062327,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 8821
+    },
+    {
+      "epoch": 3.524570515381542,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 8822
+    },
+    {
+      "epoch": 3.524970035956852,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 8823
+    },
+    {
+      "epoch": 3.525369556532161,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 8824
+    },
+    {
+      "epoch": 3.525769077107471,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 8825
+    },
+    {
+      "epoch": 3.526168597682781,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 8826
+    },
+    {
+      "epoch": 3.52656811825809,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 8827
+    },
+    {
+      "epoch": 3.5269676388334,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 8828
+    },
+    {
+      "epoch": 3.5273671594087097,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 8829
+    },
+    {
+      "epoch": 3.527766679984019,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 8830
+    },
+    {
+      "epoch": 3.528166200559329,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 8831
+    },
+    {
+      "epoch": 3.5285657211346386,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 8832
+    },
+    {
+      "epoch": 3.528965241709948,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 8833
+    },
+    {
+      "epoch": 3.5293647622852578,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 8834
+    },
+    {
+      "epoch": 3.529764282860567,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8835
+    },
+    {
+      "epoch": 3.530163803435877,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 8836
+    },
+    {
+      "epoch": 3.5305633240111867,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8837
+    },
+    {
+      "epoch": 3.5309628445864965,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 8838
+    },
+    {
+      "epoch": 3.531362365161806,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.04,
+      "step": 8839
+    },
+    {
+      "epoch": 3.5317618857371156,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 8840
+    },
+    {
+      "epoch": 3.532161406312425,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 8841
+    },
+    {
+      "epoch": 3.5325609268877347,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8842
+    },
+    {
+      "epoch": 3.5329604474630445,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 8843
+    },
+    {
+      "epoch": 3.533359968038354,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 8844
+    },
+    {
+      "epoch": 3.5337594886136636,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8845
+    },
+    {
+      "epoch": 3.534159009188973,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 8846
+    },
+    {
+      "epoch": 3.5345585297642828,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8847
+    },
+    {
+      "epoch": 3.5349580503395925,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8848
+    },
+    {
+      "epoch": 3.5353575709149023,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 8849
+    },
+    {
+      "epoch": 3.5357570914902117,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0445,
+      "step": 8850
+    },
+    {
+      "epoch": 3.5361566120655215,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 8851
+    },
+    {
+      "epoch": 3.536556132640831,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 8852
+    },
+    {
+      "epoch": 3.5369556532161406,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 8853
+    },
+    {
+      "epoch": 3.5373551737914504,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 8854
+    },
+    {
+      "epoch": 3.53775469436676,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 8855
+    },
+    {
+      "epoch": 3.5381542149420695,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8856
+    },
+    {
+      "epoch": 3.5385537355173793,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 8857
+    },
+    {
+      "epoch": 3.5389532560926886,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 8858
+    },
+    {
+      "epoch": 3.5393527766679984,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 8859
+    },
+    {
+      "epoch": 3.539752297243308,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 8860
+    },
+    {
+      "epoch": 3.5401518178186175,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 8861
+    },
+    {
+      "epoch": 3.5405513383939273,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 8862
+    },
+    {
+      "epoch": 3.5409508589692367,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 8863
+    },
+    {
+      "epoch": 3.5413503795445465,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 8864
+    },
+    {
+      "epoch": 3.5417499001198562,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 8865
+    },
+    {
+      "epoch": 3.542149420695166,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 8866
+    },
+    {
+      "epoch": 3.5425489412704754,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 8867
+    },
+    {
+      "epoch": 3.542948461845785,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 8868
+    },
+    {
+      "epoch": 3.5433479824210945,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 8869
+    },
+    {
+      "epoch": 3.5437475029964043,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.064,
+      "step": 8870
+    },
+    {
+      "epoch": 3.544147023571714,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 8871
+    },
+    {
+      "epoch": 3.5445465441470234,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0421,
+      "step": 8872
+    },
+    {
+      "epoch": 3.544946064722333,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 8873
+    },
+    {
+      "epoch": 3.545345585297643,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 8874
+    },
+    {
+      "epoch": 3.5457451058729523,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 8875
+    },
+    {
+      "epoch": 3.546144626448262,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 8876
+    },
+    {
+      "epoch": 3.546544147023572,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8877
+    },
+    {
+      "epoch": 3.5469436675988812,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 8878
+    },
+    {
+      "epoch": 3.547343188174191,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 8879
+    },
+    {
+      "epoch": 3.5477427087495004,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 8880
+    },
+    {
+      "epoch": 3.54814222932481,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 8881
+    },
+    {
+      "epoch": 3.54854174990012,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0596,
+      "step": 8882
+    },
+    {
+      "epoch": 3.5489412704754297,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 8883
+    },
+    {
+      "epoch": 3.549340791050739,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 8884
+    },
+    {
+      "epoch": 3.549740311626049,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 8885
+    },
+    {
+      "epoch": 3.550139832201358,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 8886
+    },
+    {
+      "epoch": 3.550539352776668,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 8887
+    },
+    {
+      "epoch": 3.5509388733519778,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0639,
+      "step": 8888
+    },
+    {
+      "epoch": 3.551338393927287,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 8889
+    },
+    {
+      "epoch": 3.551737914502597,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 8890
+    },
+    {
+      "epoch": 3.5521374350779062,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 8891
+    },
+    {
+      "epoch": 3.552536955653216,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8892
+    },
+    {
+      "epoch": 3.552936476228526,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 8893
+    },
+    {
+      "epoch": 3.5533359968038356,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 8894
+    },
+    {
+      "epoch": 3.553735517379145,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 8895
+    },
+    {
+      "epoch": 3.5541350379544547,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 8896
+    },
+    {
+      "epoch": 3.554534558529764,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 8897
+    },
+    {
+      "epoch": 3.554934079105074,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 8898
+    },
+    {
+      "epoch": 3.5553335996803836,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 8899
+    },
+    {
+      "epoch": 3.5557331202556934,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 8900
+    },
+    {
+      "epoch": 3.556132640831003,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 8901
+    },
+    {
+      "epoch": 3.5565321614063126,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 8902
+    },
+    {
+      "epoch": 3.556931681981622,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 8903
+    },
+    {
+      "epoch": 3.5573312025569317,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 8904
+    },
+    {
+      "epoch": 3.5577307231322415,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 8905
+    },
+    {
+      "epoch": 3.558130243707551,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 8906
+    },
+    {
+      "epoch": 3.5585297642828606,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8907
+    },
+    {
+      "epoch": 3.55892928485817,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8908
+    },
+    {
+      "epoch": 3.5593288054334797,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0425,
+      "step": 8909
+    },
+    {
+      "epoch": 3.5597283260087895,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 8910
+    },
+    {
+      "epoch": 3.5601278465840993,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 8911
+    },
+    {
+      "epoch": 3.5605273671594087,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8912
+    },
+    {
+      "epoch": 3.5609268877347184,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 8913
+    },
+    {
+      "epoch": 3.561326408310028,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 8914
+    },
+    {
+      "epoch": 3.5617259288853376,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 8915
+    },
+    {
+      "epoch": 3.5621254494606474,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 8916
+    },
+    {
+      "epoch": 3.5625249700359567,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 8917
+    },
+    {
+      "epoch": 3.5629244906112665,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 8918
+    },
+    {
+      "epoch": 3.5633240111865763,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 8919
+    },
+    {
+      "epoch": 3.5637235317618856,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 8920
+    },
+    {
+      "epoch": 3.5641230523371954,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0393,
+      "step": 8921
+    },
+    {
+      "epoch": 3.564522572912505,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 8922
+    },
+    {
+      "epoch": 3.5649220934878145,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.061,
+      "step": 8923
+    },
+    {
+      "epoch": 3.5653216140631243,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 8924
+    },
+    {
+      "epoch": 3.5657211346384337,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0684,
+      "step": 8925
+    },
+    {
+      "epoch": 3.5661206552137434,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 8926
+    },
+    {
+      "epoch": 3.5665201757890532,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8927
+    },
+    {
+      "epoch": 3.566919696364363,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0421,
+      "step": 8928
+    },
+    {
+      "epoch": 3.5673192169396724,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 8929
+    },
+    {
+      "epoch": 3.567718737514982,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 8930
+    },
+    {
+      "epoch": 3.5681182580902915,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 8931
+    },
+    {
+      "epoch": 3.5685177786656013,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 8932
+    },
+    {
+      "epoch": 3.568917299240911,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 8933
+    },
+    {
+      "epoch": 3.5693168198162204,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 8934
+    },
+    {
+      "epoch": 3.56971634039153,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 8935
+    },
+    {
+      "epoch": 3.5701158609668395,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0433,
+      "step": 8936
+    },
+    {
+      "epoch": 3.5705153815421493,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 8937
+    },
+    {
+      "epoch": 3.570914902117459,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 8938
+    },
+    {
+      "epoch": 3.571314422692769,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 8939
+    },
+    {
+      "epoch": 3.5717139432680782,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 8940
+    },
+    {
+      "epoch": 3.572113463843388,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8941
+    },
+    {
+      "epoch": 3.5725129844186974,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 8942
+    },
+    {
+      "epoch": 3.572912504994007,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 8943
+    },
+    {
+      "epoch": 3.573312025569317,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0375,
+      "step": 8944
+    },
+    {
+      "epoch": 3.5737115461446267,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8945
+    },
+    {
+      "epoch": 3.574111066719936,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 8946
+    },
+    {
+      "epoch": 3.574510587295246,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 8947
+    },
+    {
+      "epoch": 3.574910107870555,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 8948
+    },
+    {
+      "epoch": 3.575309628445865,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0303,
+      "step": 8949
+    },
+    {
+      "epoch": 3.5757091490211748,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 8950
+    },
+    {
+      "epoch": 3.576108669596484,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 8951
+    },
+    {
+      "epoch": 3.576508190171794,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 8952
+    },
+    {
+      "epoch": 3.5769077107471032,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 8953
+    },
+    {
+      "epoch": 3.577307231322413,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 8954
+    },
+    {
+      "epoch": 3.577706751897723,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 8955
+    },
+    {
+      "epoch": 3.5781062724730326,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 8956
+    },
+    {
+      "epoch": 3.578505793048342,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 8957
+    },
+    {
+      "epoch": 3.5789053136236517,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 8958
+    },
+    {
+      "epoch": 3.579304834198961,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 8959
+    },
+    {
+      "epoch": 3.579704354774271,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 8960
+    },
+    {
+      "epoch": 3.5801038753495806,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8961
+    },
+    {
+      "epoch": 3.58050339592489,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 8962
+    },
+    {
+      "epoch": 3.5809029165001998,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 8963
+    },
+    {
+      "epoch": 3.5813024370755095,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 8964
+    },
+    {
+      "epoch": 3.581701957650819,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 8965
+    },
+    {
+      "epoch": 3.5821014782261287,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 8966
+    },
+    {
+      "epoch": 3.5825009988014385,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 8967
+    },
+    {
+      "epoch": 3.582900519376748,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 8968
+    },
+    {
+      "epoch": 3.5833000399520576,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 8969
+    },
+    {
+      "epoch": 3.583699560527367,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 8970
+    },
+    {
+      "epoch": 3.5840990811026767,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 8971
+    },
+    {
+      "epoch": 3.5844986016779865,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 8972
+    },
+    {
+      "epoch": 3.5848981222532963,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 8973
+    },
+    {
+      "epoch": 3.5852976428286056,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 8974
+    },
+    {
+      "epoch": 3.5856971634039154,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 8975
+    },
+    {
+      "epoch": 3.5860966839792248,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 8976
+    },
+    {
+      "epoch": 3.5864962045545346,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 8977
+    },
+    {
+      "epoch": 3.5868957251298443,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 8978
+    },
+    {
+      "epoch": 3.5872952457051537,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 8979
+    },
+    {
+      "epoch": 3.5876947662804635,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 8980
+    },
+    {
+      "epoch": 3.588094286855773,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0365,
+      "step": 8981
+    },
+    {
+      "epoch": 3.5884938074310826,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0362,
+      "step": 8982
+    },
+    {
+      "epoch": 3.5888933280063924,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 8983
+    },
+    {
+      "epoch": 3.589292848581702,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 8984
+    },
+    {
+      "epoch": 3.5896923691570115,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0349,
+      "step": 8985
+    },
+    {
+      "epoch": 3.5900918897323213,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 8986
+    },
+    {
+      "epoch": 3.5904914103076306,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0671,
+      "step": 8987
+    },
+    {
+      "epoch": 3.5908909308829404,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 8988
+    },
+    {
+      "epoch": 3.59129045145825,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 8989
+    },
+    {
+      "epoch": 3.59168997203356,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 8990
+    },
+    {
+      "epoch": 3.5920894926088693,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 8991
+    },
+    {
+      "epoch": 3.592489013184179,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0397,
+      "step": 8992
+    },
+    {
+      "epoch": 3.5928885337594885,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 8993
+    },
+    {
+      "epoch": 3.5932880543347983,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 8994
+    },
+    {
+      "epoch": 3.593687574910108,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 8995
+    },
+    {
+      "epoch": 3.5940870954854174,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 8996
+    },
+    {
+      "epoch": 3.594486616060727,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 8997
+    },
+    {
+      "epoch": 3.5948861366360365,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 8998
+    },
+    {
+      "epoch": 3.5952856572113463,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 8999
+    },
+    {
+      "epoch": 3.595685177786656,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 9000
+    },
+    {
+      "epoch": 3.596084698361966,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 9001
+    },
+    {
+      "epoch": 3.596484218937275,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 9002
+    },
+    {
+      "epoch": 3.596883739512585,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0413,
+      "step": 9003
+    },
+    {
+      "epoch": 3.5972832600878943,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9004
+    },
+    {
+      "epoch": 3.597682780663204,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9005
+    },
+    {
+      "epoch": 3.598082301238514,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 9006
+    },
+    {
+      "epoch": 3.5984818218138233,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 9007
+    },
+    {
+      "epoch": 3.598881342389133,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 9008
+    },
+    {
+      "epoch": 3.599280862964443,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0382,
+      "step": 9009
+    },
+    {
+      "epoch": 3.599680383539752,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 9010
+    },
+    {
+      "epoch": 3.600079904115062,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.037,
+      "step": 9011
+    },
+    {
+      "epoch": 3.6004794246903717,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 9012
+    },
+    {
+      "epoch": 3.600878945265681,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 9013
+    },
+    {
+      "epoch": 3.601278465840991,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9014
+    },
+    {
+      "epoch": 3.6016779864163,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9015
+    },
+    {
+      "epoch": 3.60207750699161,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9016
+    },
+    {
+      "epoch": 3.60247702756692,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 9017
+    },
+    {
+      "epoch": 3.6028765481422296,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 9018
+    },
+    {
+      "epoch": 3.603276068717539,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 9019
+    },
+    {
+      "epoch": 3.6036755892928487,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 9020
+    },
+    {
+      "epoch": 3.604075109868158,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 9021
+    },
+    {
+      "epoch": 3.604474630443468,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9022
+    },
+    {
+      "epoch": 3.6048741510187776,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 9023
+    },
+    {
+      "epoch": 3.605273671594087,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 9024
+    },
+    {
+      "epoch": 3.6056731921693967,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0415,
+      "step": 9025
+    },
+    {
+      "epoch": 3.606072712744706,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 9026
+    },
+    {
+      "epoch": 3.606472233320016,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 9027
+    },
+    {
+      "epoch": 3.6068717538953257,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 9028
+    },
+    {
+      "epoch": 3.6072712744706354,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9029
+    },
+    {
+      "epoch": 3.607670795045945,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0628,
+      "step": 9030
+    },
+    {
+      "epoch": 3.6080703156212546,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9031
+    },
+    {
+      "epoch": 3.608469836196564,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9032
+    },
+    {
+      "epoch": 3.6088693567718737,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 9033
+    },
+    {
+      "epoch": 3.6092688773471835,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0434,
+      "step": 9034
+    },
+    {
+      "epoch": 3.6096683979224933,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9035
+    },
+    {
+      "epoch": 3.6100679184978026,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 9036
+    },
+    {
+      "epoch": 3.6104674390731124,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9037
+    },
+    {
+      "epoch": 3.6108669596484217,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 9038
+    },
+    {
+      "epoch": 3.6112664802237315,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 9039
+    },
+    {
+      "epoch": 3.6116660007990413,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 9040
+    },
+    {
+      "epoch": 3.6120655213743507,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 9041
+    },
+    {
+      "epoch": 3.6124650419496604,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0374,
+      "step": 9042
+    },
+    {
+      "epoch": 3.61286456252497,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 9043
+    },
+    {
+      "epoch": 3.6132640831002796,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 9044
+    },
+    {
+      "epoch": 3.6136636036755894,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005,
+      "loss": 1.052,
+      "step": 9045
+    },
+    {
+      "epoch": 3.614063124250899,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 9046
+    },
+    {
+      "epoch": 3.6144626448262085,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 9047
+    },
+    {
+      "epoch": 3.6148621654015183,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9048
+    },
+    {
+      "epoch": 3.6152616859768276,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 9049
+    },
+    {
+      "epoch": 3.6156612065521374,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 9050
+    },
+    {
+      "epoch": 3.616060727127447,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 9051
+    },
+    {
+      "epoch": 3.6164602477027565,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 9052
+    },
+    {
+      "epoch": 3.6168597682780663,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 9053
+    },
+    {
+      "epoch": 3.617259288853376,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 9054
+    },
+    {
+      "epoch": 3.6176588094286855,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9055
+    },
+    {
+      "epoch": 3.6180583300039952,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9056
+    },
+    {
+      "epoch": 3.618457850579305,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 9057
+    },
+    {
+      "epoch": 3.6188573711546144,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 9058
+    },
+    {
+      "epoch": 3.619256891729924,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9059
+    },
+    {
+      "epoch": 3.6196564123052335,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 9060
+    },
+    {
+      "epoch": 3.6200559328805433,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 9061
+    },
+    {
+      "epoch": 3.620455453455853,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 9062
+    },
+    {
+      "epoch": 3.620854974031163,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 9063
+    },
+    {
+      "epoch": 3.621254494606472,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9064
+    },
+    {
+      "epoch": 3.621654015181782,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 9065
+    },
+    {
+      "epoch": 3.6220535357570913,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 9066
+    },
+    {
+      "epoch": 3.622453056332401,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9067
+    },
+    {
+      "epoch": 3.622852576907711,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9068
+    },
+    {
+      "epoch": 3.6232520974830202,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9069
+    },
+    {
+      "epoch": 3.62365161805833,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 9070
+    },
+    {
+      "epoch": 3.6240511386336394,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 9071
+    },
+    {
+      "epoch": 3.624450659208949,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 9072
+    },
+    {
+      "epoch": 3.624850179784259,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 9073
+    },
+    {
+      "epoch": 3.6252497003595687,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 9074
+    },
+    {
+      "epoch": 3.625649220934878,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0434,
+      "step": 9075
+    },
+    {
+      "epoch": 3.626048741510188,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0551,
+      "step": 9076
+    },
+    {
+      "epoch": 3.626448262085497,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 9077
+    },
+    {
+      "epoch": 3.626847782660807,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 9078
+    },
+    {
+      "epoch": 3.6272473032361168,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 9079
+    },
+    {
+      "epoch": 3.6276468238114266,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0473,
+      "step": 9080
+    },
+    {
+      "epoch": 3.628046344386736,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 9081
+    },
+    {
+      "epoch": 3.6284458649620457,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 9082
+    },
+    {
+      "epoch": 3.628845385537355,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0612,
+      "step": 9083
+    },
+    {
+      "epoch": 3.629244906112665,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 9084
+    },
+    {
+      "epoch": 3.6296444266879746,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 9085
+    },
+    {
+      "epoch": 3.630043947263284,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 9086
+    },
+    {
+      "epoch": 3.6304434678385937,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 9087
+    },
+    {
+      "epoch": 3.630842988413903,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 9088
+    },
+    {
+      "epoch": 3.631242508989213,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 9089
+    },
+    {
+      "epoch": 3.6316420295645226,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9090
+    },
+    {
+      "epoch": 3.6320415501398324,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9091
+    },
+    {
+      "epoch": 3.6324410707151418,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0644,
+      "step": 9092
+    },
+    {
+      "epoch": 3.6328405912904516,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 9093
+    },
+    {
+      "epoch": 3.633240111865761,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 9094
+    },
+    {
+      "epoch": 3.6336396324410707,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9095
+    },
+    {
+      "epoch": 3.6340391530163805,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 9096
+    },
+    {
+      "epoch": 3.63443867359169,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0387,
+      "step": 9097
+    },
+    {
+      "epoch": 3.6348381941669996,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 9098
+    },
+    {
+      "epoch": 3.6352377147423094,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 9099
+    },
+    {
+      "epoch": 3.6356372353176187,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 9100
+    },
+    {
+      "epoch": 3.6360367558929285,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9101
+    },
+    {
+      "epoch": 3.6364362764682383,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 9102
+    },
+    {
+      "epoch": 3.6368357970435476,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 9103
+    },
+    {
+      "epoch": 3.6372353176188574,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 9104
+    },
+    {
+      "epoch": 3.6376348381941668,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 9105
+    },
+    {
+      "epoch": 3.6380343587694766,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9106
+    },
+    {
+      "epoch": 3.6384338793447863,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 9107
+    },
+    {
+      "epoch": 3.638833399920096,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 9108
+    },
+    {
+      "epoch": 3.6392329204954055,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9109
+    },
+    {
+      "epoch": 3.6396324410707153,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0399,
+      "step": 9110
+    },
+    {
+      "epoch": 3.6400319616460246,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9111
+    },
+    {
+      "epoch": 3.6404314822213344,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 9112
+    },
+    {
+      "epoch": 3.640831002796644,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9113
+    },
+    {
+      "epoch": 3.6412305233719535,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 9114
+    },
+    {
+      "epoch": 3.6416300439472633,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 9115
+    },
+    {
+      "epoch": 3.6420295645225726,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 9116
+    },
+    {
+      "epoch": 3.6424290850978824,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9117
+    },
+    {
+      "epoch": 3.642828605673192,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0384,
+      "step": 9118
+    },
+    {
+      "epoch": 3.643228126248502,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 9119
+    },
+    {
+      "epoch": 3.6436276468238114,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 9120
+    },
+    {
+      "epoch": 3.644027167399121,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9121
+    },
+    {
+      "epoch": 3.6444266879744305,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 9122
+    },
+    {
+      "epoch": 3.6448262085497403,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 9123
+    },
+    {
+      "epoch": 3.64522572912505,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 9124
+    },
+    {
+      "epoch": 3.64562524970036,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9125
+    },
+    {
+      "epoch": 3.646024770275669,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9126
+    },
+    {
+      "epoch": 3.646424290850979,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9127
+    },
+    {
+      "epoch": 3.6468238114262883,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 9128
+    },
+    {
+      "epoch": 3.647223332001598,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 9129
+    },
+    {
+      "epoch": 3.647622852576908,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 9130
+    },
+    {
+      "epoch": 3.6480223731522172,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9131
+    },
+    {
+      "epoch": 3.648421893727527,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9132
+    },
+    {
+      "epoch": 3.6488214143028364,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0403,
+      "step": 9133
+    },
+    {
+      "epoch": 3.649220934878146,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9134
+    },
+    {
+      "epoch": 3.649620455453456,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0611,
+      "step": 9135
+    },
+    {
+      "epoch": 3.6500199760287657,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9136
+    },
+    {
+      "epoch": 3.650419496604075,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 9137
+    },
+    {
+      "epoch": 3.650819017179385,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9138
+    },
+    {
+      "epoch": 3.651218537754694,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9139
+    },
+    {
+      "epoch": 3.651618058330004,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0355,
+      "step": 9140
+    },
+    {
+      "epoch": 3.6520175789053138,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 9141
+    },
+    {
+      "epoch": 3.652417099480623,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 9142
+    },
+    {
+      "epoch": 3.652816620055933,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9143
+    },
+    {
+      "epoch": 3.6532161406312427,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 9144
+    },
+    {
+      "epoch": 3.653615661206552,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0689,
+      "step": 9145
+    },
+    {
+      "epoch": 3.654015181781862,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 9146
+    },
+    {
+      "epoch": 3.6544147023571716,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0369,
+      "step": 9147
+    },
+    {
+      "epoch": 3.654814222932481,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 9148
+    },
+    {
+      "epoch": 3.6552137435077907,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 9149
+    },
+    {
+      "epoch": 3.6556132640831,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 9150
+    },
+    {
+      "epoch": 3.65601278465841,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 9151
+    },
+    {
+      "epoch": 3.6564123052337196,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9152
+    },
+    {
+      "epoch": 3.6568118258090294,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9153
+    },
+    {
+      "epoch": 3.6572113463843388,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 9154
+    },
+    {
+      "epoch": 3.6576108669596485,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 9155
+    },
+    {
+      "epoch": 3.658010387534958,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 9156
+    },
+    {
+      "epoch": 3.6584099081102677,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9157
+    },
+    {
+      "epoch": 3.6588094286855775,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0563,
+      "step": 9158
+    },
+    {
+      "epoch": 3.659208949260887,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 9159
+    },
+    {
+      "epoch": 3.6596084698361966,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 9160
+    },
+    {
+      "epoch": 3.660007990411506,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 9161
+    },
+    {
+      "epoch": 3.6604075109868157,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 9162
+    },
+    {
+      "epoch": 3.6608070315621255,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 9163
+    },
+    {
+      "epoch": 3.6612065521374353,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 9164
+    },
+    {
+      "epoch": 3.6616060727127446,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 9165
+    },
+    {
+      "epoch": 3.6620055932880544,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 9166
+    },
+    {
+      "epoch": 3.6624051138633638,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0378,
+      "step": 9167
+    },
+    {
+      "epoch": 3.6628046344386735,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 9168
+    },
+    {
+      "epoch": 3.6632041550139833,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 9169
+    },
+    {
+      "epoch": 3.663603675589293,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 9170
+    },
+    {
+      "epoch": 3.6640031961646025,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 9171
+    },
+    {
+      "epoch": 3.6644027167399122,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 9172
+    },
+    {
+      "epoch": 3.6648022373152216,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 9173
+    },
+    {
+      "epoch": 3.6652017578905314,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9174
+    },
+    {
+      "epoch": 3.665601278465841,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 9175
+    },
+    {
+      "epoch": 3.6660007990411505,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9176
+    },
+    {
+      "epoch": 3.6664003196164603,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 9177
+    },
+    {
+      "epoch": 3.6667998401917696,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0398,
+      "step": 9178
+    },
+    {
+      "epoch": 3.6671993607670794,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9179
+    },
+    {
+      "epoch": 3.667598881342389,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 9180
+    },
+    {
+      "epoch": 3.667998401917699,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9181
+    },
+    {
+      "epoch": 3.6683979224930083,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 9182
+    },
+    {
+      "epoch": 3.668797443068318,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9183
+    },
+    {
+      "epoch": 3.6691969636436275,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.04,
+      "step": 9184
+    },
+    {
+      "epoch": 3.6695964842189372,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 9185
+    },
+    {
+      "epoch": 3.669996004794247,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 9186
+    },
+    {
+      "epoch": 3.6703955253695564,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 9187
+    },
+    {
+      "epoch": 3.670795045944866,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 9188
+    },
+    {
+      "epoch": 3.671194566520176,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 9189
+    },
+    {
+      "epoch": 3.6715940870954853,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0402,
+      "step": 9190
+    },
+    {
+      "epoch": 3.671993607670795,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 9191
+    },
+    {
+      "epoch": 3.672393128246105,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 9192
+    },
+    {
+      "epoch": 3.672792648821414,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 9193
+    },
+    {
+      "epoch": 3.673192169396724,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 9194
+    },
+    {
+      "epoch": 3.6735916899720333,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0315,
+      "step": 9195
+    },
+    {
+      "epoch": 3.673991210547343,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 9196
+    },
+    {
+      "epoch": 3.674390731122653,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9197
+    },
+    {
+      "epoch": 3.6747902516979627,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9198
+    },
+    {
+      "epoch": 3.675189772273272,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 9199
+    },
+    {
+      "epoch": 3.675589292848582,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0356,
+      "step": 9200
+    },
+    {
+      "epoch": 3.675988813423891,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0366,
+      "step": 9201
+    },
+    {
+      "epoch": 3.676388333999201,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0378,
+      "step": 9202
+    },
+    {
+      "epoch": 3.6767878545745107,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0404,
+      "step": 9203
+    },
+    {
+      "epoch": 3.67718737514982,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 9204
+    },
+    {
+      "epoch": 3.67758689572513,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 9205
+    },
+    {
+      "epoch": 3.677986416300439,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 9206
+    },
+    {
+      "epoch": 3.678385936875749,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0404,
+      "step": 9207
+    },
+    {
+      "epoch": 3.678785457451059,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 9208
+    },
+    {
+      "epoch": 3.6791849780263686,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 9209
+    },
+    {
+      "epoch": 3.679584498601678,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 9210
+    },
+    {
+      "epoch": 3.6799840191769877,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 9211
+    },
+    {
+      "epoch": 3.680383539752297,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 9212
+    },
+    {
+      "epoch": 3.680783060327607,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9213
+    },
+    {
+      "epoch": 3.6811825809029166,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 9214
+    },
+    {
+      "epoch": 3.6815821014782264,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9215
+    },
+    {
+      "epoch": 3.6819816220535357,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0378,
+      "step": 9216
+    },
+    {
+      "epoch": 3.6823811426288455,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0649,
+      "step": 9217
+    },
+    {
+      "epoch": 3.682780663204155,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 9218
+    },
+    {
+      "epoch": 3.6831801837794647,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9219
+    },
+    {
+      "epoch": 3.6835797043547744,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 9220
+    },
+    {
+      "epoch": 3.683979224930084,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 9221
+    },
+    {
+      "epoch": 3.6843787455053936,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9222
+    },
+    {
+      "epoch": 3.684778266080703,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9223
+    },
+    {
+      "epoch": 3.6851777866560127,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 9224
+    },
+    {
+      "epoch": 3.6855773072313225,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 9225
+    },
+    {
+      "epoch": 3.6859768278066323,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 9226
+    },
+    {
+      "epoch": 3.6863763483819416,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9227
+    },
+    {
+      "epoch": 3.6867758689572514,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0507,
+      "step": 9228
+    },
+    {
+      "epoch": 3.6871753895325607,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0427,
+      "step": 9229
+    },
+    {
+      "epoch": 3.6875749101078705,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 9230
+    },
+    {
+      "epoch": 3.6879744306831803,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 9231
+    },
+    {
+      "epoch": 3.6883739512584897,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0326,
+      "step": 9232
+    },
+    {
+      "epoch": 3.6887734718337994,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0601,
+      "step": 9233
+    },
+    {
+      "epoch": 3.6891729924091092,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 9234
+    },
+    {
+      "epoch": 3.6895725129844186,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0391,
+      "step": 9235
+    },
+    {
+      "epoch": 3.6899720335597284,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 9236
+    },
+    {
+      "epoch": 3.690371554135038,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 9237
+    },
+    {
+      "epoch": 3.6907710747103475,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 9238
+    },
+    {
+      "epoch": 3.6911705952856573,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 9239
+    },
+    {
+      "epoch": 3.6915701158609666,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 9240
+    },
+    {
+      "epoch": 3.6919696364362764,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 9241
+    },
+    {
+      "epoch": 3.692369157011586,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 9242
+    },
+    {
+      "epoch": 3.692768677586896,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9243
+    },
+    {
+      "epoch": 3.6931681981622053,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 9244
+    },
+    {
+      "epoch": 3.693567718737515,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 9245
+    },
+    {
+      "epoch": 3.6939672393128244,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 9246
+    },
+    {
+      "epoch": 3.6943667598881342,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 9247
+    },
+    {
+      "epoch": 3.694766280463444,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0599,
+      "step": 9248
+    },
+    {
+      "epoch": 3.6951658010387534,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9249
+    },
+    {
+      "epoch": 3.695565321614063,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0656,
+      "step": 9250
+    },
+    {
+      "epoch": 3.6959648421893725,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 9251
+    },
+    {
+      "epoch": 3.6963643627646823,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9252
+    },
+    {
+      "epoch": 3.696763883339992,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9253
+    },
+    {
+      "epoch": 3.697163403915302,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9254
+    },
+    {
+      "epoch": 3.697562924490611,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 9255
+    },
+    {
+      "epoch": 3.697962445065921,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 9256
+    },
+    {
+      "epoch": 3.6983619656412303,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9257
+    },
+    {
+      "epoch": 3.69876148621654,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 9258
+    },
+    {
+      "epoch": 3.69916100679185,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9259
+    },
+    {
+      "epoch": 3.6995605273671597,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 9260
+    },
+    {
+      "epoch": 3.699960047942469,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 9261
+    },
+    {
+      "epoch": 3.700359568517779,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9262
+    },
+    {
+      "epoch": 3.700759089093088,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 9263
+    },
+    {
+      "epoch": 3.701158609668398,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0576,
+      "step": 9264
+    },
+    {
+      "epoch": 3.7015581302437077,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 9265
+    },
+    {
+      "epoch": 3.701957650819017,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0366,
+      "step": 9266
+    },
+    {
+      "epoch": 3.702357171394327,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 9267
+    },
+    {
+      "epoch": 3.702756691969636,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9268
+    },
+    {
+      "epoch": 3.703156212544946,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9269
+    },
+    {
+      "epoch": 3.7035557331202558,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 9270
+    },
+    {
+      "epoch": 3.7039552536955656,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 9271
+    },
+    {
+      "epoch": 3.704354774270875,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0383,
+      "step": 9272
+    },
+    {
+      "epoch": 3.7047542948461847,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 9273
+    },
+    {
+      "epoch": 3.705153815421494,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0399,
+      "step": 9274
+    },
+    {
+      "epoch": 3.705553335996804,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 9275
+    },
+    {
+      "epoch": 3.7059528565721136,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.047,
+      "step": 9276
+    },
+    {
+      "epoch": 3.706352377147423,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9277
+    },
+    {
+      "epoch": 3.7067518977227327,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 9278
+    },
+    {
+      "epoch": 3.7071514182980425,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9279
+    },
+    {
+      "epoch": 3.707550938873352,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 9280
+    },
+    {
+      "epoch": 3.7079504594486616,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9281
+    },
+    {
+      "epoch": 3.7083499800239714,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 9282
+    },
+    {
+      "epoch": 3.7087495005992808,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9283
+    },
+    {
+      "epoch": 3.7091490211745906,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0458,
+      "step": 9284
+    },
+    {
+      "epoch": 3.7095485417499,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9285
+    },
+    {
+      "epoch": 3.7099480623252097,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 9286
+    },
+    {
+      "epoch": 3.7103475829005195,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 9287
+    },
+    {
+      "epoch": 3.7107471034758293,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 9288
+    },
+    {
+      "epoch": 3.7111466240511386,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 9289
+    },
+    {
+      "epoch": 3.7115461446264484,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 9290
+    },
+    {
+      "epoch": 3.7119456652017577,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 9291
+    },
+    {
+      "epoch": 3.7123451857770675,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 9292
+    },
+    {
+      "epoch": 3.7127447063523773,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0511,
+      "step": 9293
+    },
+    {
+      "epoch": 3.7131442269276866,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 9294
+    },
+    {
+      "epoch": 3.7135437475029964,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 9295
+    },
+    {
+      "epoch": 3.7139432680783058,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 9296
+    },
+    {
+      "epoch": 3.7143427886536156,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9297
+    },
+    {
+      "epoch": 3.7147423092289253,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 9298
+    },
+    {
+      "epoch": 3.715141829804235,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 9299
+    },
+    {
+      "epoch": 3.7155413503795445,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0526,
+      "step": 9300
+    },
+    {
+      "epoch": 3.7159408709548543,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 9301
+    },
+    {
+      "epoch": 3.7163403915301636,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0603,
+      "step": 9302
+    },
+    {
+      "epoch": 3.7167399121054734,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 9303
+    },
+    {
+      "epoch": 3.717139432680783,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0608,
+      "step": 9304
+    },
+    {
+      "epoch": 3.717538953256093,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 9305
+    },
+    {
+      "epoch": 3.7179384738314023,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0562,
+      "step": 9306
+    },
+    {
+      "epoch": 3.718337994406712,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0661,
+      "step": 9307
+    },
+    {
+      "epoch": 3.7187375149820214,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0405,
+      "step": 9308
+    },
+    {
+      "epoch": 3.719137035557331,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 9309
+    },
+    {
+      "epoch": 3.719536556132641,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 9310
+    },
+    {
+      "epoch": 3.7199360767079503,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9311
+    },
+    {
+      "epoch": 3.72033559728326,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0651,
+      "step": 9312
+    },
+    {
+      "epoch": 3.7207351178585695,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9313
+    },
+    {
+      "epoch": 3.7211346384338793,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 9314
+    },
+    {
+      "epoch": 3.721534159009189,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 9315
+    },
+    {
+      "epoch": 3.721933679584499,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 9316
+    },
+    {
+      "epoch": 3.722333200159808,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 9317
+    },
+    {
+      "epoch": 3.722732720735118,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 9318
+    },
+    {
+      "epoch": 3.7231322413104273,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9319
+    },
+    {
+      "epoch": 3.723531761885737,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 9320
+    },
+    {
+      "epoch": 3.723931282461047,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0347,
+      "step": 9321
+    },
+    {
+      "epoch": 3.724330803036356,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 9322
+    },
+    {
+      "epoch": 3.724730323611666,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 9323
+    },
+    {
+      "epoch": 3.725129844186976,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0445,
+      "step": 9324
+    },
+    {
+      "epoch": 3.725529364762285,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9325
+    },
+    {
+      "epoch": 3.725928885337595,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 9326
+    },
+    {
+      "epoch": 3.7263284059129047,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 9327
+    },
+    {
+      "epoch": 3.726727926488214,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 9328
+    },
+    {
+      "epoch": 3.727127447063524,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 9329
+    },
+    {
+      "epoch": 3.727526967638833,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0357,
+      "step": 9330
+    },
+    {
+      "epoch": 3.727926488214143,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.034,
+      "step": 9331
+    },
+    {
+      "epoch": 3.7283260087894528,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0393,
+      "step": 9332
+    },
+    {
+      "epoch": 3.7287255293647625,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0394,
+      "step": 9333
+    },
+    {
+      "epoch": 3.729125049940072,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9334
+    },
+    {
+      "epoch": 3.7295245705153817,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0381,
+      "step": 9335
+    },
+    {
+      "epoch": 3.729924091090691,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 9336
+    },
+    {
+      "epoch": 3.730323611666001,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9337
+    },
+    {
+      "epoch": 3.7307231322413106,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0626,
+      "step": 9338
+    },
+    {
+      "epoch": 3.73112265281662,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9339
+    },
+    {
+      "epoch": 3.7315221733919297,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0375,
+      "step": 9340
+    },
+    {
+      "epoch": 3.731921693967239,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 9341
+    },
+    {
+      "epoch": 3.732321214542549,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 9342
+    },
+    {
+      "epoch": 3.7327207351178586,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0387,
+      "step": 9343
+    },
+    {
+      "epoch": 3.7331202556931684,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 9344
+    },
+    {
+      "epoch": 3.7335197762684778,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 9345
+    },
+    {
+      "epoch": 3.7339192968437875,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 9346
+    },
+    {
+      "epoch": 3.734318817419097,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 9347
+    },
+    {
+      "epoch": 3.7347183379944067,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 9348
+    },
+    {
+      "epoch": 3.7351178585697165,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 9349
+    },
+    {
+      "epoch": 3.7355173791450262,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0631,
+      "step": 9350
+    },
+    {
+      "epoch": 3.7359168997203356,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0462,
+      "step": 9351
+    },
+    {
+      "epoch": 3.7363164202956454,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 9352
+    },
+    {
+      "epoch": 3.7367159408709547,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 9353
+    },
+    {
+      "epoch": 3.7371154614462645,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.058,
+      "step": 9354
+    },
+    {
+      "epoch": 3.7375149820215743,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0373,
+      "step": 9355
+    },
+    {
+      "epoch": 3.7379145025968836,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 9356
+    },
+    {
+      "epoch": 3.7383140231721934,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9357
+    },
+    {
+      "epoch": 3.7387135437475028,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0588,
+      "step": 9358
+    },
+    {
+      "epoch": 3.7391130643228125,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0411,
+      "step": 9359
+    },
+    {
+      "epoch": 3.7395125848981223,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 9360
+    },
+    {
+      "epoch": 3.739912105473432,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 9361
+    },
+    {
+      "epoch": 3.7403116260487415,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 9362
+    },
+    {
+      "epoch": 3.7407111466240512,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0673,
+      "step": 9363
+    },
+    {
+      "epoch": 3.7411106671993606,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 9364
+    },
+    {
+      "epoch": 3.7415101877746704,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 9365
+    },
+    {
+      "epoch": 3.74190970834998,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9366
+    },
+    {
+      "epoch": 3.7423092289252895,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 9367
+    },
+    {
+      "epoch": 3.7427087495005993,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9368
+    },
+    {
+      "epoch": 3.743108270075909,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0355,
+      "step": 9369
+    },
+    {
+      "epoch": 3.7435077906512184,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0377,
+      "step": 9370
+    },
+    {
+      "epoch": 3.743907311226528,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 9371
+    },
+    {
+      "epoch": 3.744306831801838,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 9372
+    },
+    {
+      "epoch": 3.7447063523771473,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 9373
+    },
+    {
+      "epoch": 3.745105872952457,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 9374
+    },
+    {
+      "epoch": 3.7455053935277665,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0398,
+      "step": 9375
+    },
+    {
+      "epoch": 3.7459049141030762,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 9376
+    },
+    {
+      "epoch": 3.746304434678386,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 9377
+    },
+    {
+      "epoch": 3.746703955253696,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 9378
+    },
+    {
+      "epoch": 3.747103475829005,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 9379
+    },
+    {
+      "epoch": 3.747502996404315,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9380
+    },
+    {
+      "epoch": 3.7479025169796243,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 9381
+    },
+    {
+      "epoch": 3.748302037554934,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9382
+    },
+    {
+      "epoch": 3.748701558130244,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 9383
+    },
+    {
+      "epoch": 3.749101078705553,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0598,
+      "step": 9384
+    },
+    {
+      "epoch": 3.749500599280863,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9385
+    },
+    {
+      "epoch": 3.7499001198561723,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 9386
+    },
+    {
+      "epoch": 3.750299640431482,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 9387
+    },
+    {
+      "epoch": 3.750699161006792,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 9388
+    },
+    {
+      "epoch": 3.7510986815821017,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0404,
+      "step": 9389
+    },
+    {
+      "epoch": 3.751498202157411,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9390
+    },
+    {
+      "epoch": 3.751897722732721,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9391
+    },
+    {
+      "epoch": 3.75229724330803,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9392
+    },
+    {
+      "epoch": 3.75269676388334,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 9393
+    },
+    {
+      "epoch": 3.7530962844586497,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.034,
+      "step": 9394
+    },
+    {
+      "epoch": 3.7534958050339595,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 9395
+    },
+    {
+      "epoch": 3.753895325609269,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 9396
+    },
+    {
+      "epoch": 3.7542948461845786,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 9397
+    },
+    {
+      "epoch": 3.754694366759888,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 9398
+    },
+    {
+      "epoch": 3.7550938873351978,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 9399
+    },
+    {
+      "epoch": 3.7554934079105076,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0438,
+      "step": 9400
+    },
+    {
+      "epoch": 3.755892928485817,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 9401
+    },
+    {
+      "epoch": 3.7562924490611267,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9402
+    },
+    {
+      "epoch": 3.756691969636436,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 9403
+    },
+    {
+      "epoch": 3.757091490211746,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 9404
+    },
+    {
+      "epoch": 3.7574910107870556,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 9405
+    },
+    {
+      "epoch": 3.7578905313623654,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 9406
+    },
+    {
+      "epoch": 3.7582900519376747,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0635,
+      "step": 9407
+    },
+    {
+      "epoch": 3.7586895725129845,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 9408
+    },
+    {
+      "epoch": 3.759089093088294,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9409
+    },
+    {
+      "epoch": 3.7594886136636037,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9410
+    },
+    {
+      "epoch": 3.7598881342389134,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 9411
+    },
+    {
+      "epoch": 3.760287654814223,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 9412
+    },
+    {
+      "epoch": 3.7606871753895326,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 9413
+    },
+    {
+      "epoch": 3.7610866959648424,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0415,
+      "step": 9414
+    },
+    {
+      "epoch": 3.7614862165401517,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 9415
+    },
+    {
+      "epoch": 3.7618857371154615,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 9416
+    },
+    {
+      "epoch": 3.7622852576907713,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 9417
+    },
+    {
+      "epoch": 3.7626847782660806,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 9418
+    },
+    {
+      "epoch": 3.7630842988413904,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 9419
+    },
+    {
+      "epoch": 3.7634838194166997,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 9420
+    },
+    {
+      "epoch": 3.7638833399920095,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9421
+    },
+    {
+      "epoch": 3.7642828605673193,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9422
+    },
+    {
+      "epoch": 3.764682381142629,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 9423
+    },
+    {
+      "epoch": 3.7650819017179384,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 9424
+    },
+    {
+      "epoch": 3.7654814222932482,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 9425
+    },
+    {
+      "epoch": 3.7658809428685576,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 9426
+    },
+    {
+      "epoch": 3.7662804634438674,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 9427
+    },
+    {
+      "epoch": 3.766679984019177,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 9428
+    },
+    {
+      "epoch": 3.7670795045944865,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 9429
+    },
+    {
+      "epoch": 3.7674790251697963,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9430
+    },
+    {
+      "epoch": 3.7678785457451056,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0632,
+      "step": 9431
+    },
+    {
+      "epoch": 3.7682780663204154,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 9432
+    },
+    {
+      "epoch": 3.768677586895725,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0623,
+      "step": 9433
+    },
+    {
+      "epoch": 3.769077107471035,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9434
+    },
+    {
+      "epoch": 3.7694766280463443,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9435
+    },
+    {
+      "epoch": 3.769876148621654,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0541,
+      "step": 9436
+    },
+    {
+      "epoch": 3.7702756691969634,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0481,
+      "step": 9437
+    },
+    {
+      "epoch": 3.7706751897722732,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9438
+    },
+    {
+      "epoch": 3.771074710347583,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 9439
+    },
+    {
+      "epoch": 3.771474230922893,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 9440
+    },
+    {
+      "epoch": 3.771873751498202,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 9441
+    },
+    {
+      "epoch": 3.772273272073512,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 9442
+    },
+    {
+      "epoch": 3.7726727926488213,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 9443
+    },
+    {
+      "epoch": 3.773072313224131,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0633,
+      "step": 9444
+    },
+    {
+      "epoch": 3.773471833799441,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 9445
+    },
+    {
+      "epoch": 3.77387135437475,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 9446
+    },
+    {
+      "epoch": 3.77427087495006,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.062,
+      "step": 9447
+    },
+    {
+      "epoch": 3.7746703955253693,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9448
+    },
+    {
+      "epoch": 3.775069916100679,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0357,
+      "step": 9449
+    },
+    {
+      "epoch": 3.775469436675989,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 9450
+    },
+    {
+      "epoch": 3.7758689572512987,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 9451
+    },
+    {
+      "epoch": 3.776268477826608,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0372,
+      "step": 9452
+    },
+    {
+      "epoch": 3.776667998401918,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0545,
+      "step": 9453
+    },
+    {
+      "epoch": 3.777067518977227,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9454
+    },
+    {
+      "epoch": 3.777467039552537,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 9455
+    },
+    {
+      "epoch": 3.7778665601278467,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 9456
+    },
+    {
+      "epoch": 3.778266080703156,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 9457
+    },
+    {
+      "epoch": 3.778665601278466,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0383,
+      "step": 9458
+    },
+    {
+      "epoch": 3.7790651218537756,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9459
+    },
+    {
+      "epoch": 3.779464642429085,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0359,
+      "step": 9460
+    },
+    {
+      "epoch": 3.7798641630043948,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9461
+    },
+    {
+      "epoch": 3.7802636835797045,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0452,
+      "step": 9462
+    },
+    {
+      "epoch": 3.780663204155014,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9463
+    },
+    {
+      "epoch": 3.7810627247303237,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 9464
+    },
+    {
+      "epoch": 3.781462245305633,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 9465
+    },
+    {
+      "epoch": 3.781861765880943,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 9466
+    },
+    {
+      "epoch": 3.7822612864562526,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0338,
+      "step": 9467
+    },
+    {
+      "epoch": 3.7826608070315624,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 9468
+    },
+    {
+      "epoch": 3.7830603276068717,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 9469
+    },
+    {
+      "epoch": 3.7834598481821815,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 9470
+    },
+    {
+      "epoch": 3.783859368757491,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0515,
+      "step": 9471
+    },
+    {
+      "epoch": 3.7842588893328006,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9472
+    },
+    {
+      "epoch": 3.7846584099081104,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 9473
+    },
+    {
+      "epoch": 3.7850579304834198,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0589,
+      "step": 9474
+    },
+    {
+      "epoch": 3.7854574510587295,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9475
+    },
+    {
+      "epoch": 3.785856971634039,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 9476
+    },
+    {
+      "epoch": 3.7862564922093487,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0554,
+      "step": 9477
+    },
+    {
+      "epoch": 3.7866560127846585,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 9478
+    },
+    {
+      "epoch": 3.7870555333599683,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9479
+    },
+    {
+      "epoch": 3.7874550539352776,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 9480
+    },
+    {
+      "epoch": 3.7878545745105874,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 9481
+    },
+    {
+      "epoch": 3.7882540950858967,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0367,
+      "step": 9482
+    },
+    {
+      "epoch": 3.7886536156612065,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9483
+    },
+    {
+      "epoch": 3.7890531362365163,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 9484
+    },
+    {
+      "epoch": 3.789452656811826,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9485
+    },
+    {
+      "epoch": 3.7898521773871354,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 9486
+    },
+    {
+      "epoch": 3.790251697962445,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9487
+    },
+    {
+      "epoch": 3.7906512185377546,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 9488
+    },
+    {
+      "epoch": 3.7910507391130643,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9489
+    },
+    {
+      "epoch": 3.791450259688374,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0582,
+      "step": 9490
+    },
+    {
+      "epoch": 3.7918497802636835,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0546,
+      "step": 9491
+    },
+    {
+      "epoch": 3.7922493008389933,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0392,
+      "step": 9492
+    },
+    {
+      "epoch": 3.7926488214143026,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 9493
+    },
+    {
+      "epoch": 3.7930483419896124,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 9494
+    },
+    {
+      "epoch": 3.793447862564922,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 9495
+    },
+    {
+      "epoch": 3.793847383140232,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 9496
+    },
+    {
+      "epoch": 3.7942469037155413,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 9497
+    },
+    {
+      "epoch": 3.794646424290851,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0398,
+      "step": 9498
+    },
+    {
+      "epoch": 3.7950459448661604,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0427,
+      "step": 9499
+    },
+    {
+      "epoch": 3.79544546544147,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9500
+    },
+    {
+      "epoch": 3.79584498601678,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 9501
+    },
+    {
+      "epoch": 3.7962445065920893,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0383,
+      "step": 9502
+    },
+    {
+      "epoch": 3.796644027167399,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0354,
+      "step": 9503
+    },
+    {
+      "epoch": 3.797043547742709,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9504
+    },
+    {
+      "epoch": 3.7974430683180183,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 9505
+    },
+    {
+      "epoch": 3.797842588893328,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9506
+    },
+    {
+      "epoch": 3.798242109468638,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 9507
+    },
+    {
+      "epoch": 3.798641630043947,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9508
+    },
+    {
+      "epoch": 3.799041150619257,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 9509
+    },
+    {
+      "epoch": 3.7994406711945663,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0343,
+      "step": 9510
+    },
+    {
+      "epoch": 3.799840191769876,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 9511
+    },
+    {
+      "epoch": 3.800239712345186,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 9512
+    },
+    {
+      "epoch": 3.8006392329204957,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 9513
+    },
+    {
+      "epoch": 3.801038753495805,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 9514
+    },
+    {
+      "epoch": 3.801438274071115,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 9515
+    },
+    {
+      "epoch": 3.801837794646424,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0423,
+      "step": 9516
+    },
+    {
+      "epoch": 3.802237315221734,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0418,
+      "step": 9517
+    },
+    {
+      "epoch": 3.8026368357970437,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9518
+    },
+    {
+      "epoch": 3.803036356372353,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.066,
+      "step": 9519
+    },
+    {
+      "epoch": 3.803435876947663,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 9520
+    },
+    {
+      "epoch": 3.803835397522972,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 9521
+    },
+    {
+      "epoch": 3.804234918098282,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 9522
+    },
+    {
+      "epoch": 3.8046344386735917,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0369,
+      "step": 9523
+    },
+    {
+      "epoch": 3.8050339592489015,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 9524
+    },
+    {
+      "epoch": 3.805433479824211,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0316,
+      "step": 9525
+    },
+    {
+      "epoch": 3.8058330003995207,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 9526
+    },
+    {
+      "epoch": 3.80623252097483,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 9527
+    },
+    {
+      "epoch": 3.80663204155014,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9528
+    },
+    {
+      "epoch": 3.8070315621254496,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9529
+    },
+    {
+      "epoch": 3.8074310827007594,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 9530
+    },
+    {
+      "epoch": 3.8078306032760687,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0395,
+      "step": 9531
+    },
+    {
+      "epoch": 3.8082301238513785,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0386,
+      "step": 9532
+    },
+    {
+      "epoch": 3.808629644426688,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 9533
+    },
+    {
+      "epoch": 3.8090291650019976,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0637,
+      "step": 9534
+    },
+    {
+      "epoch": 3.8094286855773074,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 9535
+    },
+    {
+      "epoch": 3.8098282061526167,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 9536
+    },
+    {
+      "epoch": 3.8102277267279265,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9537
+    },
+    {
+      "epoch": 3.810627247303236,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9538
+    },
+    {
+      "epoch": 3.8110267678785457,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0489,
+      "step": 9539
+    },
+    {
+      "epoch": 3.8114262884538554,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9540
+    },
+    {
+      "epoch": 3.8118258090291652,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9541
+    },
+    {
+      "epoch": 3.8122253296044746,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9542
+    },
+    {
+      "epoch": 3.8126248501797844,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0386,
+      "step": 9543
+    },
+    {
+      "epoch": 3.8130243707550937,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9544
+    },
+    {
+      "epoch": 3.8134238913304035,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0625,
+      "step": 9545
+    },
+    {
+      "epoch": 3.8138234119057133,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9546
+    },
+    {
+      "epoch": 3.8142229324810226,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9547
+    },
+    {
+      "epoch": 3.8146224530563324,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 9548
+    },
+    {
+      "epoch": 3.815021973631642,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 9549
+    },
+    {
+      "epoch": 3.8154214942069515,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 9550
+    },
+    {
+      "epoch": 3.8158210147822613,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 9551
+    },
+    {
+      "epoch": 3.816220535357571,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 9552
+    },
+    {
+      "epoch": 3.8166200559328805,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 9553
+    },
+    {
+      "epoch": 3.8170195765081902,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 9554
+    },
+    {
+      "epoch": 3.8174190970834996,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9555
+    },
+    {
+      "epoch": 3.8178186176588094,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0419,
+      "step": 9556
+    },
+    {
+      "epoch": 3.818218138234119,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0558,
+      "step": 9557
+    },
+    {
+      "epoch": 3.818617658809429,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 9558
+    },
+    {
+      "epoch": 3.8190171793847383,
+      "grad_norm": 0.04296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0382,
+      "step": 9559
+    },
+    {
+      "epoch": 3.819416699960048,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 9560
+    },
+    {
+      "epoch": 3.8198162205353574,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.065,
+      "step": 9561
+    },
+    {
+      "epoch": 3.820215741110667,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 9562
+    },
+    {
+      "epoch": 3.820615261685977,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9563
+    },
+    {
+      "epoch": 3.8210147822612863,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9564
+    },
+    {
+      "epoch": 3.821414302836596,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9565
+    },
+    {
+      "epoch": 3.8218138234119055,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 9566
+    },
+    {
+      "epoch": 3.8222133439872152,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 9567
+    },
+    {
+      "epoch": 3.822612864562525,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0469,
+      "step": 9568
+    },
+    {
+      "epoch": 3.823012385137835,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0642,
+      "step": 9569
+    },
+    {
+      "epoch": 3.823411905713144,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 9570
+    },
+    {
+      "epoch": 3.823811426288454,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 9571
+    },
+    {
+      "epoch": 3.8242109468637633,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0382,
+      "step": 9572
+    },
+    {
+      "epoch": 3.824610467439073,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0402,
+      "step": 9573
+    },
+    {
+      "epoch": 3.825009988014383,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0461,
+      "step": 9574
+    },
+    {
+      "epoch": 3.8254095085896926,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9575
+    },
+    {
+      "epoch": 3.825809029165002,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9576
+    },
+    {
+      "epoch": 3.8262085497403118,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 9577
+    },
+    {
+      "epoch": 3.826608070315621,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 9578
+    },
+    {
+      "epoch": 3.827007590890931,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0653,
+      "step": 9579
+    },
+    {
+      "epoch": 3.8274071114662407,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 9580
+    },
+    {
+      "epoch": 3.82780663204155,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0577,
+      "step": 9581
+    },
+    {
+      "epoch": 3.82820615261686,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 9582
+    },
+    {
+      "epoch": 3.828605673192169,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9583
+    },
+    {
+      "epoch": 3.829005193767479,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0346,
+      "step": 9584
+    },
+    {
+      "epoch": 3.8294047143427887,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9585
+    },
+    {
+      "epoch": 3.8298042349180985,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 9586
+    },
+    {
+      "epoch": 3.830203755493408,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9587
+    },
+    {
+      "epoch": 3.8306032760687176,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 9588
+    },
+    {
+      "epoch": 3.831002796644027,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9589
+    },
+    {
+      "epoch": 3.8314023172193368,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 9590
+    },
+    {
+      "epoch": 3.8318018377946466,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 9591
+    },
+    {
+      "epoch": 3.832201358369956,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0597,
+      "step": 9592
+    },
+    {
+      "epoch": 3.8326008789452657,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 9593
+    },
+    {
+      "epoch": 3.8330003995205755,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 9594
+    },
+    {
+      "epoch": 3.833399920095885,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0344,
+      "step": 9595
+    },
+    {
+      "epoch": 3.8337994406711946,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 9596
+    },
+    {
+      "epoch": 3.8341989612465044,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9597
+    },
+    {
+      "epoch": 3.8345984818218137,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0285,
+      "step": 9598
+    },
+    {
+      "epoch": 3.8349980023971235,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 9599
+    },
+    {
+      "epoch": 3.835397522972433,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 9600
+    },
+    {
+      "epoch": 3.8357970435477426,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0621,
+      "step": 9601
+    },
+    {
+      "epoch": 3.8361965641230524,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9602
+    },
+    {
+      "epoch": 3.836596084698362,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0624,
+      "step": 9603
+    },
+    {
+      "epoch": 3.8369956052736716,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0406,
+      "step": 9604
+    },
+    {
+      "epoch": 3.8373951258489813,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 9605
+    },
+    {
+      "epoch": 3.8377946464242907,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.054,
+      "step": 9606
+    },
+    {
+      "epoch": 3.8381941669996005,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 9607
+    },
+    {
+      "epoch": 3.8385936875749103,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0385,
+      "step": 9608
+    },
+    {
+      "epoch": 3.8389932081502196,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 9609
+    },
+    {
+      "epoch": 3.8393927287255294,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9610
+    },
+    {
+      "epoch": 3.8397922493008387,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 9611
+    },
+    {
+      "epoch": 3.8401917698761485,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0465,
+      "step": 9612
+    },
+    {
+      "epoch": 3.8405912904514583,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9613
+    },
+    {
+      "epoch": 3.840990811026768,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9614
+    },
+    {
+      "epoch": 3.8413903316020774,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0571,
+      "step": 9615
+    },
+    {
+      "epoch": 3.841789852177387,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9616
+    },
+    {
+      "epoch": 3.8421893727526966,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0619,
+      "step": 9617
+    },
+    {
+      "epoch": 3.8425888933280063,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 9618
+    },
+    {
+      "epoch": 3.842988413903316,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0379,
+      "step": 9619
+    },
+    {
+      "epoch": 3.843387934478626,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9620
+    },
+    {
+      "epoch": 3.8437874550539353,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 9621
+    },
+    {
+      "epoch": 3.844186975629245,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0433,
+      "step": 9622
+    },
+    {
+      "epoch": 3.8445864962045544,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9623
+    },
+    {
+      "epoch": 3.844986016779864,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0569,
+      "step": 9624
+    },
+    {
+      "epoch": 3.845385537355174,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 9625
+    },
+    {
+      "epoch": 3.8457850579304833,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0405,
+      "step": 9626
+    },
+    {
+      "epoch": 3.846184578505793,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 9627
+    },
+    {
+      "epoch": 3.8465840990811024,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 9628
+    },
+    {
+      "epoch": 3.8469836196564122,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 9629
+    },
+    {
+      "epoch": 3.847383140231722,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0446,
+      "step": 9630
+    },
+    {
+      "epoch": 3.847782660807032,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 9631
+    },
+    {
+      "epoch": 3.848182181382341,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9632
+    },
+    {
+      "epoch": 3.848581701957651,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0557,
+      "step": 9633
+    },
+    {
+      "epoch": 3.8489812225329603,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9634
+    },
+    {
+      "epoch": 3.84938074310827,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 9635
+    },
+    {
+      "epoch": 3.84978026368358,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 9636
+    },
+    {
+      "epoch": 3.850179784258889,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0397,
+      "step": 9637
+    },
+    {
+      "epoch": 3.850579304834199,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0615,
+      "step": 9638
+    },
+    {
+      "epoch": 3.8509788254095088,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9639
+    },
+    {
+      "epoch": 3.851378345984818,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9640
+    },
+    {
+      "epoch": 3.851777866560128,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0533,
+      "step": 9641
+    },
+    {
+      "epoch": 3.8521773871354377,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0365,
+      "step": 9642
+    },
+    {
+      "epoch": 3.852576907710747,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 9643
+    },
+    {
+      "epoch": 3.852976428286057,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 9644
+    },
+    {
+      "epoch": 3.853375948861366,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0586,
+      "step": 9645
+    },
+    {
+      "epoch": 3.853775469436676,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 9646
+    },
+    {
+      "epoch": 3.8541749900119857,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0566,
+      "step": 9647
+    },
+    {
+      "epoch": 3.8545745105872955,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 9648
+    },
+    {
+      "epoch": 3.854974031162605,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.034,
+      "step": 9649
+    },
+    {
+      "epoch": 3.8553735517379146,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9650
+    },
+    {
+      "epoch": 3.855773072313224,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0484,
+      "step": 9651
+    },
+    {
+      "epoch": 3.8561725928885338,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 9652
+    },
+    {
+      "epoch": 3.8565721134638435,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0488,
+      "step": 9653
+    },
+    {
+      "epoch": 3.856971634039153,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.035,
+      "step": 9654
+    },
+    {
+      "epoch": 3.8573711546144627,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9655
+    },
+    {
+      "epoch": 3.857770675189772,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 9656
+    },
+    {
+      "epoch": 3.858170195765082,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9657
+    },
+    {
+      "epoch": 3.8585697163403916,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 9658
+    },
+    {
+      "epoch": 3.8589692369157014,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.056,
+      "step": 9659
+    },
+    {
+      "epoch": 3.8593687574910107,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 9660
+    },
+    {
+      "epoch": 3.8597682780663205,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9661
+    },
+    {
+      "epoch": 3.86016779864163,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0371,
+      "step": 9662
+    },
+    {
+      "epoch": 3.8605673192169396,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9663
+    },
+    {
+      "epoch": 3.8609668397922494,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 9664
+    },
+    {
+      "epoch": 3.861366360367559,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0388,
+      "step": 9665
+    },
+    {
+      "epoch": 3.8617658809428685,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 9666
+    },
+    {
+      "epoch": 3.8621654015181783,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 9667
+    },
+    {
+      "epoch": 3.8625649220934877,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 9668
+    },
+    {
+      "epoch": 3.8629644426687975,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 9669
+    },
+    {
+      "epoch": 3.8633639632441072,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9670
+    },
+    {
+      "epoch": 3.8637634838194166,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9671
+    },
+    {
+      "epoch": 3.8641630043947264,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9672
+    },
+    {
+      "epoch": 3.8645625249700357,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0434,
+      "step": 9673
+    },
+    {
+      "epoch": 3.8649620455453455,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9674
+    },
+    {
+      "epoch": 3.8653615661206553,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 9675
+    },
+    {
+      "epoch": 3.865761086695965,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0439,
+      "step": 9676
+    },
+    {
+      "epoch": 3.8661606072712744,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0358,
+      "step": 9677
+    },
+    {
+      "epoch": 3.866560127846584,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0487,
+      "step": 9678
+    },
+    {
+      "epoch": 3.8669596484218935,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 9679
+    },
+    {
+      "epoch": 3.8673591689972033,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.053,
+      "step": 9680
+    },
+    {
+      "epoch": 3.867758689572513,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0585,
+      "step": 9681
+    },
+    {
+      "epoch": 3.8681582101478225,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0535,
+      "step": 9682
+    },
+    {
+      "epoch": 3.8685577307231322,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9683
+    },
+    {
+      "epoch": 3.868957251298442,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 9684
+    },
+    {
+      "epoch": 3.8693567718737514,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 9685
+    },
+    {
+      "epoch": 3.869756292449061,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9686
+    },
+    {
+      "epoch": 3.870155813024371,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0592,
+      "step": 9687
+    },
+    {
+      "epoch": 3.8705553335996803,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0581,
+      "step": 9688
+    },
+    {
+      "epoch": 3.87095485417499,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.045,
+      "step": 9689
+    },
+    {
+      "epoch": 3.8713543747502994,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 9690
+    },
+    {
+      "epoch": 3.871753895325609,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 9691
+    },
+    {
+      "epoch": 3.872153415900919,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 9692
+    },
+    {
+      "epoch": 3.872552936476229,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 9693
+    },
+    {
+      "epoch": 3.872952457051538,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0382,
+      "step": 9694
+    },
+    {
+      "epoch": 3.873351977626848,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0544,
+      "step": 9695
+    },
+    {
+      "epoch": 3.8737514982021573,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9696
+    },
+    {
+      "epoch": 3.874151018777467,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 9697
+    },
+    {
+      "epoch": 3.874550539352777,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0342,
+      "step": 9698
+    },
+    {
+      "epoch": 3.874950059928086,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0641,
+      "step": 9699
+    },
+    {
+      "epoch": 3.875349580503396,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 9700
+    },
+    {
+      "epoch": 3.8757491010787053,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0402,
+      "step": 9701
+    },
+    {
+      "epoch": 3.876148621654015,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0634,
+      "step": 9702
+    },
+    {
+      "epoch": 3.876548142229325,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9703
+    },
+    {
+      "epoch": 3.8769476628046347,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0388,
+      "step": 9704
+    },
+    {
+      "epoch": 3.877347183379944,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 9705
+    },
+    {
+      "epoch": 3.877746703955254,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0595,
+      "step": 9706
+    },
+    {
+      "epoch": 3.878146224530563,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9707
+    },
+    {
+      "epoch": 3.878545745105873,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9708
+    },
+    {
+      "epoch": 3.8789452656811827,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0413,
+      "step": 9709
+    },
+    {
+      "epoch": 3.8793447862564925,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0579,
+      "step": 9710
+    },
+    {
+      "epoch": 3.879744306831802,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 9711
+    },
+    {
+      "epoch": 3.8801438274071116,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0522,
+      "step": 9712
+    },
+    {
+      "epoch": 3.880543347982421,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9713
+    },
+    {
+      "epoch": 3.8809428685577307,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9714
+    },
+    {
+      "epoch": 3.8813423891330405,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 9715
+    },
+    {
+      "epoch": 3.88174190970835,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0521,
+      "step": 9716
+    },
+    {
+      "epoch": 3.8821414302836597,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 9717
+    },
+    {
+      "epoch": 3.882540950858969,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 9718
+    },
+    {
+      "epoch": 3.882940471434279,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 9719
+    },
+    {
+      "epoch": 3.8833399920095886,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0613,
+      "step": 9720
+    },
+    {
+      "epoch": 3.8837395125848984,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0519,
+      "step": 9721
+    },
+    {
+      "epoch": 3.8841390331602077,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9722
+    },
+    {
+      "epoch": 3.8845385537355175,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 9723
+    },
+    {
+      "epoch": 3.884938074310827,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9724
+    },
+    {
+      "epoch": 3.8853375948861366,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 9725
+    },
+    {
+      "epoch": 3.8857371154614464,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9726
+    },
+    {
+      "epoch": 3.8861366360367557,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 9727
+    },
+    {
+      "epoch": 3.8865361566120655,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9728
+    },
+    {
+      "epoch": 3.8869356771873753,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.051,
+      "step": 9729
+    },
+    {
+      "epoch": 3.8873351977626847,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0425,
+      "step": 9730
+    },
+    {
+      "epoch": 3.8877347183379944,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0559,
+      "step": 9731
+    },
+    {
+      "epoch": 3.8881342389133042,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9732
+    },
+    {
+      "epoch": 3.8885337594886136,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0375,
+      "step": 9733
+    },
+    {
+      "epoch": 3.8889332800639234,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0573,
+      "step": 9734
+    },
+    {
+      "epoch": 3.8893328006392327,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0338,
+      "step": 9735
+    },
+    {
+      "epoch": 3.8897323212145425,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0594,
+      "step": 9736
+    },
+    {
+      "epoch": 3.8901318417898523,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9737
+    },
+    {
+      "epoch": 3.890531362365162,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 9738
+    },
+    {
+      "epoch": 3.8909308829404714,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0432,
+      "step": 9739
+    },
+    {
+      "epoch": 3.891330403515781,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9740
+    },
+    {
+      "epoch": 3.8917299240910905,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9741
+    },
+    {
+      "epoch": 3.8921294446664003,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 9742
+    },
+    {
+      "epoch": 3.89252896524171,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0574,
+      "step": 9743
+    },
+    {
+      "epoch": 3.8929284858170194,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0403,
+      "step": 9744
+    },
+    {
+      "epoch": 3.8933280063923292,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0478,
+      "step": 9745
+    },
+    {
+      "epoch": 3.8937275269676386,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9746
+    },
+    {
+      "epoch": 3.8941270475429484,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 9747
+    },
+    {
+      "epoch": 3.894526568118258,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9748
+    },
+    {
+      "epoch": 3.894926088693568,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0543,
+      "step": 9749
+    },
+    {
+      "epoch": 3.8953256092688773,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0424,
+      "step": 9750
+    },
+    {
+      "epoch": 3.895725129844187,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9751
+    },
+    {
+      "epoch": 3.8961246504194964,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 9752
+    },
+    {
+      "epoch": 3.896524170994806,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9753
+    },
+    {
+      "epoch": 3.896923691570116,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0372,
+      "step": 9754
+    },
+    {
+      "epoch": 3.8973232121454258,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9755
+    },
+    {
+      "epoch": 3.897722732720735,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0417,
+      "step": 9756
+    },
+    {
+      "epoch": 3.898122253296045,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 9757
+    },
+    {
+      "epoch": 3.8985217738713542,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 9758
+    },
+    {
+      "epoch": 3.898921294446664,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9759
+    },
+    {
+      "epoch": 3.899320815021974,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0561,
+      "step": 9760
+    },
+    {
+      "epoch": 3.899720335597283,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0627,
+      "step": 9761
+    },
+    {
+      "epoch": 3.900119856172593,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0407,
+      "step": 9762
+    },
+    {
+      "epoch": 3.9005193767479023,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 9763
+    },
+    {
+      "epoch": 3.900918897323212,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0591,
+      "step": 9764
+    },
+    {
+      "epoch": 3.901318417898522,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9765
+    },
+    {
+      "epoch": 3.9017179384738316,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 9766
+    },
+    {
+      "epoch": 3.902117459049141,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 9767
+    },
+    {
+      "epoch": 3.9025169796244508,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9768
+    },
+    {
+      "epoch": 3.90291650019976,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0375,
+      "step": 9769
+    },
+    {
+      "epoch": 3.90331602077507,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9770
+    },
+    {
+      "epoch": 3.9037155413503797,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 9771
+    },
+    {
+      "epoch": 3.904115061925689,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0496,
+      "step": 9772
+    },
+    {
+      "epoch": 3.904514582500999,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0449,
+      "step": 9773
+    },
+    {
+      "epoch": 3.9049141030763086,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0497,
+      "step": 9774
+    },
+    {
+      "epoch": 3.905313623651618,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 9775
+    },
+    {
+      "epoch": 3.9057131442269277,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 9776
+    },
+    {
+      "epoch": 3.9061126648022375,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0448,
+      "step": 9777
+    },
+    {
+      "epoch": 3.906512185377547,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0454,
+      "step": 9778
+    },
+    {
+      "epoch": 3.9069117059528566,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0553,
+      "step": 9779
+    },
+    {
+      "epoch": 3.907311226528166,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 9780
+    },
+    {
+      "epoch": 3.9077107471034758,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 9781
+    },
+    {
+      "epoch": 3.9081102676787856,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0395,
+      "step": 9782
+    },
+    {
+      "epoch": 3.9085097882540953,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 9783
+    },
+    {
+      "epoch": 3.9089093088294047,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005,
+      "loss": 1.042,
+      "step": 9784
+    },
+    {
+      "epoch": 3.9093088294047145,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0433,
+      "step": 9785
+    },
+    {
+      "epoch": 3.909708349980024,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0467,
+      "step": 9786
+    },
+    {
+      "epoch": 3.9101078705553336,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 9787
+    },
+    {
+      "epoch": 3.9105073911306434,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9788
+    },
+    {
+      "epoch": 3.9109069117059527,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0443,
+      "step": 9789
+    },
+    {
+      "epoch": 3.9113064322812625,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0311,
+      "step": 9790
+    },
+    {
+      "epoch": 3.911705952856572,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0552,
+      "step": 9791
+    },
+    {
+      "epoch": 3.9121054734318816,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9792
+    },
+    {
+      "epoch": 3.9125049940071914,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0527,
+      "step": 9793
+    },
+    {
+      "epoch": 3.912904514582501,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0536,
+      "step": 9794
+    },
+    {
+      "epoch": 3.9133040351578106,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0364,
+      "step": 9795
+    },
+    {
+      "epoch": 3.9137035557331203,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 9796
+    },
+    {
+      "epoch": 3.9141030763084297,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0377,
+      "step": 9797
+    },
+    {
+      "epoch": 3.9145025968837395,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005,
+      "loss": 1.057,
+      "step": 9798
+    },
+    {
+      "epoch": 3.9149021174590493,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 9799
+    },
+    {
+      "epoch": 3.915301638034359,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9800
+    },
+    {
+      "epoch": 3.9157011586096684,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0433,
+      "step": 9801
+    },
+    {
+      "epoch": 3.916100679184978,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0529,
+      "step": 9802
+    },
+    {
+      "epoch": 3.9165001997602875,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9803
+    },
+    {
+      "epoch": 3.9168997203355973,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0471,
+      "step": 9804
+    },
+    {
+      "epoch": 3.917299240910907,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 9805
+    },
+    {
+      "epoch": 3.9176987614862164,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0414,
+      "step": 9806
+    },
+    {
+      "epoch": 3.918098282061526,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005,
+      "loss": 1.048,
+      "step": 9807
+    },
+    {
+      "epoch": 3.9184978026368356,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0692,
+      "step": 9808
+    },
+    {
+      "epoch": 3.9188973232121453,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0664,
+      "step": 9809
+    },
+    {
+      "epoch": 3.919296843787455,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0504,
+      "step": 9810
+    },
+    {
+      "epoch": 3.919696364362765,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 9811
+    },
+    {
+      "epoch": 3.9200958849380743,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0539,
+      "step": 9812
+    },
+    {
+      "epoch": 3.920495405513384,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 9813
+    },
+    {
+      "epoch": 3.9208949260886934,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 9814
+    },
+    {
+      "epoch": 3.921294446664003,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9815
+    },
+    {
+      "epoch": 3.921693967239313,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0502,
+      "step": 9816
+    },
+    {
+      "epoch": 3.9220934878146223,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 9817
+    },
+    {
+      "epoch": 3.922493008389932,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9818
+    },
+    {
+      "epoch": 3.922892528965242,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0654,
+      "step": 9819
+    },
+    {
+      "epoch": 3.923292049540551,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9820
+    },
+    {
+      "epoch": 3.923691570115861,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 9821
+    },
+    {
+      "epoch": 3.924091090691171,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 9822
+    },
+    {
+      "epoch": 3.92449061126648,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0372,
+      "step": 9823
+    },
+    {
+      "epoch": 3.92489013184179,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 9824
+    },
+    {
+      "epoch": 3.9252896524170993,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0447,
+      "step": 9825
+    },
+    {
+      "epoch": 3.925689172992409,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 9826
+    },
+    {
+      "epoch": 3.926088693567719,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9827
+    },
+    {
+      "epoch": 3.9264882141430286,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 9828
+    },
+    {
+      "epoch": 3.926887734718338,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0513,
+      "step": 9829
+    },
+    {
+      "epoch": 3.9272872552936477,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9830
+    },
+    {
+      "epoch": 3.927686775868957,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 9831
+    },
+    {
+      "epoch": 3.928086296444267,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0516,
+      "step": 9832
+    },
+    {
+      "epoch": 3.9284858170195767,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9833
+    },
+    {
+      "epoch": 3.928885337594886,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0435,
+      "step": 9834
+    },
+    {
+      "epoch": 3.929284858170196,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0568,
+      "step": 9835
+    },
+    {
+      "epoch": 3.929684378745505,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0411,
+      "step": 9836
+    },
+    {
+      "epoch": 3.930083899320815,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 9837
+    },
+    {
+      "epoch": 3.9304834198961247,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.044,
+      "step": 9838
+    },
+    {
+      "epoch": 3.9308829404714345,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 9839
+    },
+    {
+      "epoch": 3.931282461046744,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9840
+    },
+    {
+      "epoch": 3.9316819816220536,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 9841
+    },
+    {
+      "epoch": 3.932081502197363,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 9842
+    },
+    {
+      "epoch": 3.9324810227726728,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0444,
+      "step": 9843
+    },
+    {
+      "epoch": 3.9328805433479825,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0418,
+      "step": 9844
+    },
+    {
+      "epoch": 3.9332800639232923,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0564,
+      "step": 9845
+    },
+    {
+      "epoch": 3.9336795844986017,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0475,
+      "step": 9846
+    },
+    {
+      "epoch": 3.9340791050739115,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0506,
+      "step": 9847
+    },
+    {
+      "epoch": 3.934478625649221,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0466,
+      "step": 9848
+    },
+    {
+      "epoch": 3.9348781462245306,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 9849
+    },
+    {
+      "epoch": 3.9352776667998404,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0445,
+      "step": 9850
+    },
+    {
+      "epoch": 3.9356771873751497,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0386,
+      "step": 9851
+    },
+    {
+      "epoch": 3.9360767079504595,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 9852
+    },
+    {
+      "epoch": 3.936476228525769,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9853
+    },
+    {
+      "epoch": 3.9368757491010786,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0389,
+      "step": 9854
+    },
+    {
+      "epoch": 3.9372752696763884,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0524,
+      "step": 9855
+    },
+    {
+      "epoch": 3.937674790251698,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0477,
+      "step": 9856
+    },
+    {
+      "epoch": 3.9380743108270075,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0311,
+      "step": 9857
+    },
+    {
+      "epoch": 3.9384738314023173,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9858
+    },
+    {
+      "epoch": 3.9388733519776267,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0514,
+      "step": 9859
+    },
+    {
+      "epoch": 3.9392728725529365,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0609,
+      "step": 9860
+    },
+    {
+      "epoch": 3.9396723931282462,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9861
+    },
+    {
+      "epoch": 3.9400719137035556,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.046,
+      "step": 9862
+    },
+    {
+      "epoch": 3.9404714342788654,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0567,
+      "step": 9863
+    },
+    {
+      "epoch": 3.9408709548541747,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9864
+    },
+    {
+      "epoch": 3.9412704754294845,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 9865
+    },
+    {
+      "epoch": 3.9416699960047943,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 9866
+    },
+    {
+      "epoch": 3.942069516580104,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0532,
+      "step": 9867
+    },
+    {
+      "epoch": 3.9424690371554134,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9868
+    },
+    {
+      "epoch": 3.942868557730723,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0495,
+      "step": 9869
+    },
+    {
+      "epoch": 3.9432680783060325,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0583,
+      "step": 9870
+    },
+    {
+      "epoch": 3.9436675988813423,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0442,
+      "step": 9871
+    },
+    {
+      "epoch": 3.944067119456652,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0425,
+      "step": 9872
+    },
+    {
+      "epoch": 3.944466640031962,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0387,
+      "step": 9873
+    },
+    {
+      "epoch": 3.9448661606072712,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9874
+    },
+    {
+      "epoch": 3.945265681182581,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 9875
+    },
+    {
+      "epoch": 3.9456652017578904,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9876
+    },
+    {
+      "epoch": 3.9460647223332,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 9877
+    },
+    {
+      "epoch": 3.94646424290851,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0421,
+      "step": 9878
+    },
+    {
+      "epoch": 3.9468637634838193,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 9879
+    },
+    {
+      "epoch": 3.947263284059129,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0584,
+      "step": 9880
+    },
+    {
+      "epoch": 3.9476628046344384,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9881
+    },
+    {
+      "epoch": 3.948062325209748,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9882
+    },
+    {
+      "epoch": 3.948461845785058,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0492,
+      "step": 9883
+    },
+    {
+      "epoch": 3.9488613663603678,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 9884
+    },
+    {
+      "epoch": 3.949260886935677,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 9885
+    },
+    {
+      "epoch": 3.949660407510987,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0428,
+      "step": 9886
+    },
+    {
+      "epoch": 3.9500599280862962,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0396,
+      "step": 9887
+    },
+    {
+      "epoch": 3.950459448661606,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 9888
+    },
+    {
+      "epoch": 3.950858969236916,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005,
+      "loss": 1.0494,
+      "step": 9889
+    },
+    {
+      "epoch": 3.9512584898122256,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 9890
+    },
+    {
+      "epoch": 3.951658010387535,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0349,
+      "step": 9891
+    },
+    {
+      "epoch": 3.9520575309628447,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9892
+    },
+    {
+      "epoch": 3.952457051538154,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0397,
+      "step": 9893
+    },
+    {
+      "epoch": 3.952856572113464,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.043,
+      "step": 9894
+    },
+    {
+      "epoch": 3.9532560926887736,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0426,
+      "step": 9895
+    },
+    {
+      "epoch": 3.953655613264083,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0441,
+      "step": 9896
+    },
+    {
+      "epoch": 3.9540551338393928,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0392,
+      "step": 9897
+    },
+    {
+      "epoch": 3.954454654414702,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 9898
+    },
+    {
+      "epoch": 3.954854174990012,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0427,
+      "step": 9899
+    },
+    {
+      "epoch": 3.9552536955653217,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0618,
+      "step": 9900
+    },
+    {
+      "epoch": 3.9556532161406315,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 9901
+    },
+    {
+      "epoch": 3.956052736715941,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9902
+    },
+    {
+      "epoch": 3.9564522572912506,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 9903
+    },
+    {
+      "epoch": 3.95685177786656,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0334,
+      "step": 9904
+    },
+    {
+      "epoch": 3.9572512984418697,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0431,
+      "step": 9905
+    },
+    {
+      "epoch": 3.9576508190171795,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0748,
+      "step": 9906
+    },
+    {
+      "epoch": 3.958050339592489,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0409,
+      "step": 9907
+    },
+    {
+      "epoch": 3.9584498601677987,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9908
+    },
+    {
+      "epoch": 3.958849380743108,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0404,
+      "step": 9909
+    },
+    {
+      "epoch": 3.959248901318418,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9910
+    },
+    {
+      "epoch": 3.9596484218937276,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9911
+    },
+    {
+      "epoch": 3.9600479424690374,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 9912
+    },
+    {
+      "epoch": 3.9604474630443467,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0491,
+      "step": 9913
+    },
+    {
+      "epoch": 3.9608469836196565,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 9914
+    },
+    {
+      "epoch": 3.961246504194966,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 0.0005,
+      "loss": 1.0403,
+      "step": 9915
+    },
+    {
+      "epoch": 3.9616460247702756,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0534,
+      "step": 9916
+    },
+    {
+      "epoch": 3.9620455453455854,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9917
+    },
+    {
+      "epoch": 3.962445065920895,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9918
+    },
+    {
+      "epoch": 3.9628445864962045,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0457,
+      "step": 9919
+    },
+    {
+      "epoch": 3.9632441070715143,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0005,
+      "loss": 1.0499,
+      "step": 9920
+    },
+    {
+      "epoch": 3.9636436276468237,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0395,
+      "step": 9921
+    },
+    {
+      "epoch": 3.9640431482221334,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 9922
+    },
+    {
+      "epoch": 3.9644426687974432,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0455,
+      "step": 9923
+    },
+    {
+      "epoch": 3.9648421893727526,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0459,
+      "step": 9924
+    },
+    {
+      "epoch": 3.9652417099480624,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9925
+    },
+    {
+      "epoch": 3.9656412305233717,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 9926
+    },
+    {
+      "epoch": 3.9660407510986815,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.055,
+      "step": 9927
+    },
+    {
+      "epoch": 3.9664402716739913,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0538,
+      "step": 9928
+    },
+    {
+      "epoch": 3.966839792249301,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0421,
+      "step": 9929
+    },
+    {
+      "epoch": 3.9672393128246104,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0565,
+      "step": 9930
+    },
+    {
+      "epoch": 3.96763883339992,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9931
+    },
+    {
+      "epoch": 3.9680383539752295,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0498,
+      "step": 9932
+    },
+    {
+      "epoch": 3.9684378745505393,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0005,
+      "loss": 1.0556,
+      "step": 9933
+    },
+    {
+      "epoch": 3.968837395125849,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0005,
+      "loss": 1.0587,
+      "step": 9934
+    },
+    {
+      "epoch": 3.969236915701159,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005,
+      "loss": 1.0629,
+      "step": 9935
+    },
+    {
+      "epoch": 3.9696364362764682,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.041,
+      "step": 9936
+    },
+    {
+      "epoch": 3.970035956851778,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0412,
+      "step": 9937
+    },
+    {
+      "epoch": 3.9704354774270874,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0381,
+      "step": 9938
+    },
+    {
+      "epoch": 3.970834998002397,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0622,
+      "step": 9939
+    },
+    {
+      "epoch": 3.971234518577707,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005,
+      "loss": 1.0636,
+      "step": 9940
+    },
+    {
+      "epoch": 3.9716340391530163,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0005,
+      "loss": 1.05,
+      "step": 9941
+    },
+    {
+      "epoch": 3.972033559728326,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 0.0005,
+      "loss": 1.0464,
+      "step": 9942
+    },
+    {
+      "epoch": 3.9724330803036354,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 0.0005,
+      "loss": 1.0429,
+      "step": 9943
+    },
+    {
+      "epoch": 3.972832600878945,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0483,
+      "step": 9944
+    },
+    {
+      "epoch": 3.973232121454255,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0508,
+      "step": 9945
+    },
+    {
+      "epoch": 3.9736316420295648,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0476,
+      "step": 9946
+    },
+    {
+      "epoch": 3.974031162604874,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0436,
+      "step": 9947
+    },
+    {
+      "epoch": 3.974430683180184,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0523,
+      "step": 9948
+    },
+    {
+      "epoch": 3.9748302037554932,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0606,
+      "step": 9949
+    },
+    {
+      "epoch": 3.975229724330803,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0005,
+      "loss": 1.0616,
+      "step": 9950
+    },
+    {
+      "epoch": 3.975629244906113,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0479,
+      "step": 9951
+    },
+    {
+      "epoch": 3.976028765481422,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.043,
+      "step": 9952
+    },
+    {
+      "epoch": 3.976428286056732,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0422,
+      "step": 9953
+    },
+    {
+      "epoch": 3.9768278066320413,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0472,
+      "step": 9954
+    },
+    {
+      "epoch": 3.977227327207351,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 9955
+    },
+    {
+      "epoch": 3.977626847782661,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0647,
+      "step": 9956
+    },
+    {
+      "epoch": 3.9780263683579706,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0437,
+      "step": 9957
+    },
+    {
+      "epoch": 3.97842588893328,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005,
+      "loss": 1.0604,
+      "step": 9958
+    },
+    {
+      "epoch": 3.9788254095085898,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0528,
+      "step": 9959
+    },
+    {
+      "epoch": 3.979224930083899,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0425,
+      "step": 9960
+    },
+    {
+      "epoch": 3.979624450659209,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0512,
+      "step": 9961
+    },
+    {
+      "epoch": 3.9800239712345187,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0525,
+      "step": 9962
+    },
+    {
+      "epoch": 3.9804234918098285,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 0.0005,
+      "loss": 1.0474,
+      "step": 9963
+    },
+    {
+      "epoch": 3.980823012385138,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0339,
+      "step": 9964
+    },
+    {
+      "epoch": 3.9812225329604476,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9965
+    },
+    {
+      "epoch": 3.981622053535757,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 9966
+    },
+    {
+      "epoch": 3.9820215741110667,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0503,
+      "step": 9967
+    },
+    {
+      "epoch": 3.9824210946863765,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0518,
+      "step": 9968
+    },
+    {
+      "epoch": 3.982820615261686,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0572,
+      "step": 9969
+    },
+    {
+      "epoch": 3.9832201358369956,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005,
+      "loss": 1.0548,
+      "step": 9970
+    },
+    {
+      "epoch": 3.983619656412305,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005,
+      "loss": 1.0486,
+      "step": 9971
+    },
+    {
+      "epoch": 3.9840191769876148,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005,
+      "loss": 1.0549,
+      "step": 9972
+    },
+    {
+      "epoch": 3.9844186975629245,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 9973
+    },
+    {
+      "epoch": 3.9848182181382343,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0578,
+      "step": 9974
+    },
+    {
+      "epoch": 3.9852177387135437,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0413,
+      "step": 9975
+    },
+    {
+      "epoch": 3.9856172592888535,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0005,
+      "loss": 1.0401,
+      "step": 9976
+    },
+    {
+      "epoch": 3.986016779864163,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0509,
+      "step": 9977
+    },
+    {
+      "epoch": 3.9864163004394726,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0005,
+      "loss": 1.0575,
+      "step": 9978
+    },
+    {
+      "epoch": 3.9868158210147824,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 9979
+    },
+    {
+      "epoch": 3.987215341590092,
+      "grad_norm": 0.048828125,
+      "learning_rate": 0.0005,
+      "loss": 1.0537,
+      "step": 9980
+    },
+    {
+      "epoch": 3.9876148621654015,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0468,
+      "step": 9981
+    },
+    {
+      "epoch": 3.9880143827407113,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0505,
+      "step": 9982
+    },
+    {
+      "epoch": 3.9884139033160206,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005,
+      "loss": 1.049,
+      "step": 9983
+    },
+    {
+      "epoch": 3.9888134238913304,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0463,
+      "step": 9984
+    },
+    {
+      "epoch": 3.98921294446664,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 0.0005,
+      "loss": 1.0493,
+      "step": 9985
+    },
+    {
+      "epoch": 3.9896124650419496,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0005,
+      "loss": 1.0517,
+      "step": 9986
+    },
+    {
+      "epoch": 3.9900119856172593,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 0.0005,
+      "loss": 1.0451,
+      "step": 9987
+    },
+    {
+      "epoch": 3.9904115061925687,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0005,
+      "loss": 1.0485,
+      "step": 9988
+    },
+    {
+      "epoch": 3.9908110267678785,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0542,
+      "step": 9989
+    },
+    {
+      "epoch": 3.9912105473431883,
+      "grad_norm": 0.044921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0695,
+      "step": 9990
+    },
+    {
+      "epoch": 3.991610067918498,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0005,
+      "loss": 1.0547,
+      "step": 9991
+    },
+    {
+      "epoch": 3.9920095884938074,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0005,
+      "loss": 1.037,
+      "step": 9992
+    },
+    {
+      "epoch": 3.992409109069117,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0416,
+      "step": 9993
+    },
+    {
+      "epoch": 3.9928086296444265,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0005,
+      "loss": 1.0453,
+      "step": 9994
+    },
+    {
+      "epoch": 3.9932081502197363,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0005,
+      "loss": 1.0456,
+      "step": 9995
+    },
+    {
+      "epoch": 3.993607670795046,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 0.0005,
+      "loss": 1.0531,
+      "step": 9996
+    },
+    {
+      "epoch": 3.9940071913703554,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005,
+      "loss": 1.0555,
+      "step": 9997
+    },
+    {
+      "epoch": 3.994406711945665,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0005,
+      "loss": 1.0384,
+      "step": 9998
+    },
+    {
+      "epoch": 3.9948062325209746,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0005,
+      "loss": 1.0501,
+      "step": 9999
+    },
+    {
+      "epoch": 3.9952057530962843,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0482,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2503000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1000,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.031984033056686e+18,
+  "train_batch_size": 128,
+  "trial_name": null,
+  "trial_params": null
+}