diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,34193 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 4880,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00020491803278688525,
+      "grad_norm": 70.52354431152344,
+      "learning_rate": 1.360544217687075e-07,
+      "loss": 4.4634,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004098360655737705,
+      "grad_norm": 62.63434600830078,
+      "learning_rate": 2.72108843537415e-07,
+      "loss": 3.3301,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006147540983606558,
+      "grad_norm": 42.43904495239258,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 4.3627,
+      "step": 3
+    },
+    {
+      "epoch": 0.000819672131147541,
+      "grad_norm": 104.70809173583984,
+      "learning_rate": 5.4421768707483e-07,
+      "loss": 4.4047,
+      "step": 4
+    },
+    {
+      "epoch": 0.0010245901639344263,
+      "grad_norm": 77.74440002441406,
+      "learning_rate": 6.802721088435376e-07,
+      "loss": 4.1214,
+      "step": 5
+    },
+    {
+      "epoch": 0.0012295081967213116,
+      "grad_norm": 114.56156921386719,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 3.9439,
+      "step": 6
+    },
+    {
+      "epoch": 0.0014344262295081967,
+      "grad_norm": 56.30268478393555,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 4.3415,
+      "step": 7
+    },
+    {
+      "epoch": 0.001639344262295082,
+      "grad_norm": 95.06924438476562,
+      "learning_rate": 1.08843537414966e-06,
+      "loss": 3.1273,
+      "step": 8
+    },
+    {
+      "epoch": 0.0018442622950819673,
+      "grad_norm": 87.25345611572266,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 4.4863,
+      "step": 9
+    },
+    {
+      "epoch": 0.0020491803278688526,
+      "grad_norm": 93.61027526855469,
+      "learning_rate": 1.3605442176870751e-06,
+      "loss": 3.5536,
+      "step": 10
+    },
+    {
+      "epoch": 0.002254098360655738,
+      "grad_norm": 31.413633346557617,
+      "learning_rate": 1.4965986394557825e-06,
+      "loss": 2.5695,
+      "step": 11
+    },
+    {
+      "epoch": 0.002459016393442623,
+      "grad_norm": 98.80310821533203,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 3.6669,
+      "step": 12
+    },
+    {
+      "epoch": 0.002663934426229508,
+      "grad_norm": 90.9544677734375,
+      "learning_rate": 1.7687074829931975e-06,
+      "loss": 3.6284,
+      "step": 13
+    },
+    {
+      "epoch": 0.0028688524590163933,
+      "grad_norm": 117.36346435546875,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 3.5488,
+      "step": 14
+    },
+    {
+      "epoch": 0.0030737704918032786,
+      "grad_norm": 58.44469451904297,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 2.1571,
+      "step": 15
+    },
+    {
+      "epoch": 0.003278688524590164,
+      "grad_norm": 56.09517288208008,
+      "learning_rate": 2.17687074829932e-06,
+      "loss": 2.3233,
+      "step": 16
+    },
+    {
+      "epoch": 0.0034836065573770492,
+      "grad_norm": 92.69772338867188,
+      "learning_rate": 2.3129251700680273e-06,
+      "loss": 2.2183,
+      "step": 17
+    },
+    {
+      "epoch": 0.0036885245901639345,
+      "grad_norm": 96.25579071044922,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 2.2459,
+      "step": 18
+    },
+    {
+      "epoch": 0.00389344262295082,
+      "grad_norm": 39.21778869628906,
+      "learning_rate": 2.5850340136054425e-06,
+      "loss": 1.7109,
+      "step": 19
+    },
+    {
+      "epoch": 0.004098360655737705,
+      "grad_norm": 22.96857452392578,
+      "learning_rate": 2.7210884353741503e-06,
+      "loss": 1.8037,
+      "step": 20
+    },
+    {
+      "epoch": 0.00430327868852459,
+      "grad_norm": 29.3774471282959,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.79,
+      "step": 21
+    },
+    {
+      "epoch": 0.004508196721311476,
+      "grad_norm": 43.857173919677734,
+      "learning_rate": 2.993197278911565e-06,
+      "loss": 2.3311,
+      "step": 22
+    },
+    {
+      "epoch": 0.004713114754098361,
+      "grad_norm": 68.78257751464844,
+      "learning_rate": 3.1292517006802725e-06,
+      "loss": 1.4279,
+      "step": 23
+    },
+    {
+      "epoch": 0.004918032786885246,
+      "grad_norm": 58.9012565612793,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 1.7674,
+      "step": 24
+    },
+    {
+      "epoch": 0.005122950819672131,
+      "grad_norm": 19.282386779785156,
+      "learning_rate": 3.4013605442176872e-06,
+      "loss": 1.7499,
+      "step": 25
+    },
+    {
+      "epoch": 0.005327868852459016,
+      "grad_norm": 12.08267593383789,
+      "learning_rate": 3.537414965986395e-06,
+      "loss": 1.4584,
+      "step": 26
+    },
+    {
+      "epoch": 0.005532786885245901,
+      "grad_norm": 19.2326602935791,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 1.708,
+      "step": 27
+    },
+    {
+      "epoch": 0.005737704918032787,
+      "grad_norm": 17.703121185302734,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 1.6633,
+      "step": 28
+    },
+    {
+      "epoch": 0.005942622950819672,
+      "grad_norm": 19.123910903930664,
+      "learning_rate": 3.945578231292517e-06,
+      "loss": 1.7394,
+      "step": 29
+    },
+    {
+      "epoch": 0.006147540983606557,
+      "grad_norm": 18.27283477783203,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 1.5864,
+      "step": 30
+    },
+    {
+      "epoch": 0.0063524590163934426,
+      "grad_norm": 16.20029640197754,
+      "learning_rate": 4.217687074829933e-06,
+      "loss": 1.5277,
+      "step": 31
+    },
+    {
+      "epoch": 0.006557377049180328,
+      "grad_norm": 27.13144874572754,
+      "learning_rate": 4.35374149659864e-06,
+      "loss": 1.2805,
+      "step": 32
+    },
+    {
+      "epoch": 0.006762295081967213,
+      "grad_norm": 13.550676345825195,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 1.6316,
+      "step": 33
+    },
+    {
+      "epoch": 0.0069672131147540985,
+      "grad_norm": 12.758017539978027,
+      "learning_rate": 4.6258503401360546e-06,
+      "loss": 1.5128,
+      "step": 34
+    },
+    {
+      "epoch": 0.007172131147540984,
+      "grad_norm": 11.11542797088623,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 1.2861,
+      "step": 35
+    },
+    {
+      "epoch": 0.007377049180327869,
+      "grad_norm": 9.994466781616211,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 1.4709,
+      "step": 36
+    },
+    {
+      "epoch": 0.007581967213114754,
+      "grad_norm": 13.414884567260742,
+      "learning_rate": 5.034013605442177e-06,
+      "loss": 1.3745,
+      "step": 37
+    },
+    {
+      "epoch": 0.00778688524590164,
+      "grad_norm": 17.804180145263672,
+      "learning_rate": 5.170068027210885e-06,
+      "loss": 1.5925,
+      "step": 38
+    },
+    {
+      "epoch": 0.007991803278688524,
+      "grad_norm": 28.957603454589844,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 1.3823,
+      "step": 39
+    },
+    {
+      "epoch": 0.00819672131147541,
+      "grad_norm": 32.831172943115234,
+      "learning_rate": 5.442176870748301e-06,
+      "loss": 1.508,
+      "step": 40
+    },
+    {
+      "epoch": 0.008401639344262295,
+      "grad_norm": 8.307755470275879,
+      "learning_rate": 5.578231292517007e-06,
+      "loss": 1.3784,
+      "step": 41
+    },
+    {
+      "epoch": 0.00860655737704918,
+      "grad_norm": 8.693304061889648,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.2506,
+      "step": 42
+    },
+    {
+      "epoch": 0.008811475409836065,
+      "grad_norm": 17.223217010498047,
+      "learning_rate": 5.850340136054422e-06,
+      "loss": 1.5501,
+      "step": 43
+    },
+    {
+      "epoch": 0.009016393442622951,
+      "grad_norm": 7.655211925506592,
+      "learning_rate": 5.98639455782313e-06,
+      "loss": 1.1933,
+      "step": 44
+    },
+    {
+      "epoch": 0.009221311475409836,
+      "grad_norm": 10.601262092590332,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 1.3662,
+      "step": 45
+    },
+    {
+      "epoch": 0.009426229508196722,
+      "grad_norm": 12.74623966217041,
+      "learning_rate": 6.258503401360545e-06,
+      "loss": 1.2939,
+      "step": 46
+    },
+    {
+      "epoch": 0.009631147540983606,
+      "grad_norm": 25.677146911621094,
+      "learning_rate": 6.394557823129253e-06,
+      "loss": 1.6364,
+      "step": 47
+    },
+    {
+      "epoch": 0.009836065573770493,
+      "grad_norm": 33.271453857421875,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 1.0837,
+      "step": 48
+    },
+    {
+      "epoch": 0.010040983606557377,
+      "grad_norm": 19.500932693481445,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.3672,
+      "step": 49
+    },
+    {
+      "epoch": 0.010245901639344262,
+      "grad_norm": 12.357169151306152,
+      "learning_rate": 6.8027210884353745e-06,
+      "loss": 1.2843,
+      "step": 50
+    },
+    {
+      "epoch": 0.010450819672131148,
+      "grad_norm": 20.660541534423828,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 1.5166,
+      "step": 51
+    },
+    {
+      "epoch": 0.010655737704918032,
+      "grad_norm": 6.690814971923828,
+      "learning_rate": 7.07482993197279e-06,
+      "loss": 1.2924,
+      "step": 52
+    },
+    {
+      "epoch": 0.010860655737704918,
+      "grad_norm": 9.240177154541016,
+      "learning_rate": 7.210884353741497e-06,
+      "loss": 1.3927,
+      "step": 53
+    },
+    {
+      "epoch": 0.011065573770491803,
+      "grad_norm": 7.247419834136963,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 1.3985,
+      "step": 54
+    },
+    {
+      "epoch": 0.011270491803278689,
+      "grad_norm": 6.519885063171387,
+      "learning_rate": 7.482993197278913e-06,
+      "loss": 1.4811,
+      "step": 55
+    },
+    {
+      "epoch": 0.011475409836065573,
+      "grad_norm": 8.092867851257324,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 1.4866,
+      "step": 56
+    },
+    {
+      "epoch": 0.01168032786885246,
+      "grad_norm": 8.256175994873047,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 1.3912,
+      "step": 57
+    },
+    {
+      "epoch": 0.011885245901639344,
+      "grad_norm": 7.044185638427734,
+      "learning_rate": 7.891156462585034e-06,
+      "loss": 0.8956,
+      "step": 58
+    },
+    {
+      "epoch": 0.01209016393442623,
+      "grad_norm": 7.225033760070801,
+      "learning_rate": 8.027210884353741e-06,
+      "loss": 1.2556,
+      "step": 59
+    },
+    {
+      "epoch": 0.012295081967213115,
+      "grad_norm": 6.698459625244141,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 1.3912,
+      "step": 60
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 7.6336164474487305,
+      "learning_rate": 8.299319727891157e-06,
+      "loss": 1.4159,
+      "step": 61
+    },
+    {
+      "epoch": 0.012704918032786885,
+      "grad_norm": 6.3351311683654785,
+      "learning_rate": 8.435374149659866e-06,
+      "loss": 1.208,
+      "step": 62
+    },
+    {
+      "epoch": 0.012909836065573771,
+      "grad_norm": 10.987747192382812,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.293,
+      "step": 63
+    },
+    {
+      "epoch": 0.013114754098360656,
+      "grad_norm": 11.872108459472656,
+      "learning_rate": 8.70748299319728e-06,
+      "loss": 1.2438,
+      "step": 64
+    },
+    {
+      "epoch": 0.01331967213114754,
+      "grad_norm": 17.030614852905273,
+      "learning_rate": 8.843537414965987e-06,
+      "loss": 1.4289,
+      "step": 65
+    },
+    {
+      "epoch": 0.013524590163934426,
+      "grad_norm": 9.360306739807129,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 1.0959,
+      "step": 66
+    },
+    {
+      "epoch": 0.01372950819672131,
+      "grad_norm": 7.583203315734863,
+      "learning_rate": 9.115646258503402e-06,
+      "loss": 0.9907,
+      "step": 67
+    },
+    {
+      "epoch": 0.013934426229508197,
+      "grad_norm": 6.066573619842529,
+      "learning_rate": 9.251700680272109e-06,
+      "loss": 1.1113,
+      "step": 68
+    },
+    {
+      "epoch": 0.014139344262295081,
+      "grad_norm": 10.424674034118652,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 1.3842,
+      "step": 69
+    },
+    {
+      "epoch": 0.014344262295081968,
+      "grad_norm": 10.948161125183105,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 1.102,
+      "step": 70
+    },
+    {
+      "epoch": 0.014549180327868852,
+      "grad_norm": 7.992804527282715,
+      "learning_rate": 9.659863945578232e-06,
+      "loss": 1.3135,
+      "step": 71
+    },
+    {
+      "epoch": 0.014754098360655738,
+      "grad_norm": 7.234769821166992,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 1.2466,
+      "step": 72
+    },
+    {
+      "epoch": 0.014959016393442623,
+      "grad_norm": 7.50241231918335,
+      "learning_rate": 9.931972789115647e-06,
+      "loss": 1.0503,
+      "step": 73
+    },
+    {
+      "epoch": 0.015163934426229509,
+      "grad_norm": 8.018733978271484,
+      "learning_rate": 1.0068027210884354e-05,
+      "loss": 1.2381,
+      "step": 74
+    },
+    {
+      "epoch": 0.015368852459016393,
+      "grad_norm": 7.580471992492676,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 1.2495,
+      "step": 75
+    },
+    {
+      "epoch": 0.01557377049180328,
+      "grad_norm": 12.087300300598145,
+      "learning_rate": 1.034013605442177e-05,
+      "loss": 1.4176,
+      "step": 76
+    },
+    {
+      "epoch": 0.015778688524590165,
+      "grad_norm": 9.444079399108887,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 1.1955,
+      "step": 77
+    },
+    {
+      "epoch": 0.015983606557377048,
+      "grad_norm": 6.125025272369385,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 1.2054,
+      "step": 78
+    },
+    {
+      "epoch": 0.016188524590163934,
+      "grad_norm": 14.449728012084961,
+      "learning_rate": 1.0748299319727893e-05,
+      "loss": 1.2344,
+      "step": 79
+    },
+    {
+      "epoch": 0.01639344262295082,
+      "grad_norm": 7.792777061462402,
+      "learning_rate": 1.0884353741496601e-05,
+      "loss": 0.9641,
+      "step": 80
+    },
+    {
+      "epoch": 0.016598360655737703,
+      "grad_norm": 15.913436889648438,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 1.1531,
+      "step": 81
+    },
+    {
+      "epoch": 0.01680327868852459,
+      "grad_norm": 6.101311683654785,
+      "learning_rate": 1.1156462585034013e-05,
+      "loss": 1.4566,
+      "step": 82
+    },
+    {
+      "epoch": 0.017008196721311476,
+      "grad_norm": 8.605325698852539,
+      "learning_rate": 1.1292517006802722e-05,
+      "loss": 1.4164,
+      "step": 83
+    },
+    {
+      "epoch": 0.01721311475409836,
+      "grad_norm": 12.320062637329102,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.1472,
+      "step": 84
+    },
+    {
+      "epoch": 0.017418032786885244,
+      "grad_norm": 9.834012985229492,
+      "learning_rate": 1.1564625850340136e-05,
+      "loss": 1.0011,
+      "step": 85
+    },
+    {
+      "epoch": 0.01762295081967213,
+      "grad_norm": 6.471250057220459,
+      "learning_rate": 1.1700680272108845e-05,
+      "loss": 1.3704,
+      "step": 86
+    },
+    {
+      "epoch": 0.017827868852459017,
+      "grad_norm": 11.808831214904785,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 1.4195,
+      "step": 87
+    },
+    {
+      "epoch": 0.018032786885245903,
+      "grad_norm": 8.477483749389648,
+      "learning_rate": 1.197278911564626e-05,
+      "loss": 1.0781,
+      "step": 88
+    },
+    {
+      "epoch": 0.018237704918032786,
+      "grad_norm": 7.347333908081055,
+      "learning_rate": 1.2108843537414967e-05,
+      "loss": 1.4389,
+      "step": 89
+    },
+    {
+      "epoch": 0.018442622950819672,
+      "grad_norm": 6.9269585609436035,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 1.2749,
+      "step": 90
+    },
+    {
+      "epoch": 0.018647540983606558,
+      "grad_norm": 10.419594764709473,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 0.9977,
+      "step": 91
+    },
+    {
+      "epoch": 0.018852459016393444,
+      "grad_norm": 5.431417942047119,
+      "learning_rate": 1.251700680272109e-05,
+      "loss": 1.2084,
+      "step": 92
+    },
+    {
+      "epoch": 0.019057377049180327,
+      "grad_norm": 9.720732688903809,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 1.3194,
+      "step": 93
+    },
+    {
+      "epoch": 0.019262295081967213,
+      "grad_norm": 5.313255310058594,
+      "learning_rate": 1.2789115646258505e-05,
+      "loss": 1.1175,
+      "step": 94
+    },
+    {
+      "epoch": 0.0194672131147541,
+      "grad_norm": 8.45032787322998,
+      "learning_rate": 1.2925170068027212e-05,
+      "loss": 1.0982,
+      "step": 95
+    },
+    {
+      "epoch": 0.019672131147540985,
+      "grad_norm": 4.430140018463135,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 0.9337,
+      "step": 96
+    },
+    {
+      "epoch": 0.019877049180327868,
+      "grad_norm": 12.27338981628418,
+      "learning_rate": 1.3197278911564626e-05,
+      "loss": 1.6666,
+      "step": 97
+    },
+    {
+      "epoch": 0.020081967213114754,
+      "grad_norm": 5.76198673248291,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.249,
+      "step": 98
+    },
+    {
+      "epoch": 0.02028688524590164,
+      "grad_norm": 6.549588203430176,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 1.125,
+      "step": 99
+    },
+    {
+      "epoch": 0.020491803278688523,
+      "grad_norm": 17.627540588378906,
+      "learning_rate": 1.3605442176870749e-05,
+      "loss": 1.094,
+      "step": 100
+    },
+    {
+      "epoch": 0.02069672131147541,
+      "grad_norm": 5.4481425285339355,
+      "learning_rate": 1.3741496598639456e-05,
+      "loss": 1.0699,
+      "step": 101
+    },
+    {
+      "epoch": 0.020901639344262295,
+      "grad_norm": 6.589564800262451,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 1.0399,
+      "step": 102
+    },
+    {
+      "epoch": 0.02110655737704918,
+      "grad_norm": 8.633831977844238,
+      "learning_rate": 1.4013605442176872e-05,
+      "loss": 1.0855,
+      "step": 103
+    },
+    {
+      "epoch": 0.021311475409836064,
+      "grad_norm": 5.375726222991943,
+      "learning_rate": 1.414965986394558e-05,
+      "loss": 1.0079,
+      "step": 104
+    },
+    {
+      "epoch": 0.02151639344262295,
+      "grad_norm": 13.008928298950195,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.7567,
+      "step": 105
+    },
+    {
+      "epoch": 0.021721311475409837,
+      "grad_norm": 9.22192096710205,
+      "learning_rate": 1.4421768707482994e-05,
+      "loss": 1.3072,
+      "step": 106
+    },
+    {
+      "epoch": 0.021926229508196723,
+      "grad_norm": 15.61291790008545,
+      "learning_rate": 1.4557823129251703e-05,
+      "loss": 1.3382,
+      "step": 107
+    },
+    {
+      "epoch": 0.022131147540983605,
+      "grad_norm": 7.4089579582214355,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 1.3322,
+      "step": 108
+    },
+    {
+      "epoch": 0.02233606557377049,
+      "grad_norm": 8.459456443786621,
+      "learning_rate": 1.4829931972789118e-05,
+      "loss": 1.3497,
+      "step": 109
+    },
+    {
+      "epoch": 0.022540983606557378,
+      "grad_norm": 9.232161521911621,
+      "learning_rate": 1.4965986394557825e-05,
+      "loss": 1.0648,
+      "step": 110
+    },
+    {
+      "epoch": 0.022745901639344264,
+      "grad_norm": 7.919378757476807,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 1.2031,
+      "step": 111
+    },
+    {
+      "epoch": 0.022950819672131147,
+      "grad_norm": 5.073938369750977,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 1.1524,
+      "step": 112
+    },
+    {
+      "epoch": 0.023155737704918033,
+      "grad_norm": 8.450615882873535,
+      "learning_rate": 1.5374149659863945e-05,
+      "loss": 1.281,
+      "step": 113
+    },
+    {
+      "epoch": 0.02336065573770492,
+      "grad_norm": 9.293567657470703,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 0.9808,
+      "step": 114
+    },
+    {
+      "epoch": 0.0235655737704918,
+      "grad_norm": 16.753215789794922,
+      "learning_rate": 1.5646258503401362e-05,
+      "loss": 1.2217,
+      "step": 115
+    },
+    {
+      "epoch": 0.023770491803278688,
+      "grad_norm": 4.330772399902344,
+      "learning_rate": 1.578231292517007e-05,
+      "loss": 1.0051,
+      "step": 116
+    },
+    {
+      "epoch": 0.023975409836065574,
+      "grad_norm": 10.495442390441895,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 0.9206,
+      "step": 117
+    },
+    {
+      "epoch": 0.02418032786885246,
+      "grad_norm": 6.8509297370910645,
+      "learning_rate": 1.6054421768707483e-05,
+      "loss": 1.2529,
+      "step": 118
+    },
+    {
+      "epoch": 0.024385245901639343,
+      "grad_norm": 7.595362186431885,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.9256,
+      "step": 119
+    },
+    {
+      "epoch": 0.02459016393442623,
+      "grad_norm": 8.258394241333008,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 1.1673,
+      "step": 120
+    },
+    {
+      "epoch": 0.024795081967213115,
+      "grad_norm": 14.05384635925293,
+      "learning_rate": 1.6462585034013607e-05,
+      "loss": 1.2332,
+      "step": 121
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 8.516852378845215,
+      "learning_rate": 1.6598639455782314e-05,
+      "loss": 1.0019,
+      "step": 122
+    },
+    {
+      "epoch": 0.025204918032786884,
+      "grad_norm": 14.138521194458008,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 1.3111,
+      "step": 123
+    },
+    {
+      "epoch": 0.02540983606557377,
+      "grad_norm": 7.923465728759766,
+      "learning_rate": 1.687074829931973e-05,
+      "loss": 1.2941,
+      "step": 124
+    },
+    {
+      "epoch": 0.025614754098360656,
+      "grad_norm": 6.187522888183594,
+      "learning_rate": 1.7006802721088435e-05,
+      "loss": 1.0597,
+      "step": 125
+    },
+    {
+      "epoch": 0.025819672131147543,
+      "grad_norm": 7.313263893127441,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.9949,
+      "step": 126
+    },
+    {
+      "epoch": 0.026024590163934425,
+      "grad_norm": 5.4770989418029785,
+      "learning_rate": 1.7278911564625852e-05,
+      "loss": 0.8755,
+      "step": 127
+    },
+    {
+      "epoch": 0.02622950819672131,
+      "grad_norm": 4.413930892944336,
+      "learning_rate": 1.741496598639456e-05,
+      "loss": 1.1793,
+      "step": 128
+    },
+    {
+      "epoch": 0.026434426229508198,
+      "grad_norm": 5.8976149559021,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 1.1475,
+      "step": 129
+    },
+    {
+      "epoch": 0.02663934426229508,
+      "grad_norm": 8.168647766113281,
+      "learning_rate": 1.7687074829931973e-05,
+      "loss": 1.0569,
+      "step": 130
+    },
+    {
+      "epoch": 0.026844262295081966,
+      "grad_norm": 4.450958251953125,
+      "learning_rate": 1.782312925170068e-05,
+      "loss": 1.145,
+      "step": 131
+    },
+    {
+      "epoch": 0.027049180327868853,
+      "grad_norm": 5.435624599456787,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 1.1596,
+      "step": 132
+    },
+    {
+      "epoch": 0.02725409836065574,
+      "grad_norm": 7.785508155822754,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 1.0785,
+      "step": 133
+    },
+    {
+      "epoch": 0.02745901639344262,
+      "grad_norm": 3.5530316829681396,
+      "learning_rate": 1.8231292517006804e-05,
+      "loss": 0.9686,
+      "step": 134
+    },
+    {
+      "epoch": 0.027663934426229508,
+      "grad_norm": 4.79872465133667,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 1.0595,
+      "step": 135
+    },
+    {
+      "epoch": 0.027868852459016394,
+      "grad_norm": 9.013531684875488,
+      "learning_rate": 1.8503401360544218e-05,
+      "loss": 1.1157,
+      "step": 136
+    },
+    {
+      "epoch": 0.02807377049180328,
+      "grad_norm": 4.757910251617432,
+      "learning_rate": 1.863945578231293e-05,
+      "loss": 1.1223,
+      "step": 137
+    },
+    {
+      "epoch": 0.028278688524590163,
+      "grad_norm": 9.590606689453125,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 1.379,
+      "step": 138
+    },
+    {
+      "epoch": 0.02848360655737705,
+      "grad_norm": 10.98641586303711,
+      "learning_rate": 1.8911564625850343e-05,
+      "loss": 1.3612,
+      "step": 139
+    },
+    {
+      "epoch": 0.028688524590163935,
+      "grad_norm": 6.515561103820801,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 1.4689,
+      "step": 140
+    },
+    {
+      "epoch": 0.02889344262295082,
+      "grad_norm": 8.915470123291016,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 1.0698,
+      "step": 141
+    },
+    {
+      "epoch": 0.029098360655737704,
+      "grad_norm": 14.602327346801758,
+      "learning_rate": 1.9319727891156463e-05,
+      "loss": 1.1287,
+      "step": 142
+    },
+    {
+      "epoch": 0.02930327868852459,
+      "grad_norm": 8.850377082824707,
+      "learning_rate": 1.945578231292517e-05,
+      "loss": 1.27,
+      "step": 143
+    },
+    {
+      "epoch": 0.029508196721311476,
+      "grad_norm": 14.269759178161621,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 1.0585,
+      "step": 144
+    },
+    {
+      "epoch": 0.02971311475409836,
+      "grad_norm": 7.54500675201416,
+      "learning_rate": 1.9727891156462588e-05,
+      "loss": 1.3498,
+      "step": 145
+    },
+    {
+      "epoch": 0.029918032786885245,
+      "grad_norm": 8.48276138305664,
+      "learning_rate": 1.9863945578231295e-05,
+      "loss": 1.0192,
+      "step": 146
+    },
+    {
+      "epoch": 0.03012295081967213,
+      "grad_norm": 22.97007179260254,
+      "learning_rate": 2e-05,
+      "loss": 1.3309,
+      "step": 147
+    },
+    {
+      "epoch": 0.030327868852459017,
+      "grad_norm": 5.385742664337158,
+      "learning_rate": 1.999999779709015e-05,
+      "loss": 1.1295,
+      "step": 148
+    },
+    {
+      "epoch": 0.0305327868852459,
+      "grad_norm": 6.592761993408203,
+      "learning_rate": 1.999999118836157e-05,
+      "loss": 1.2529,
+      "step": 149
+    },
+    {
+      "epoch": 0.030737704918032786,
+      "grad_norm": 5.611164569854736,
+      "learning_rate": 1.999998017381717e-05,
+      "loss": 1.3107,
+      "step": 150
+    },
+    {
+      "epoch": 0.030942622950819672,
+      "grad_norm": 10.386234283447266,
+      "learning_rate": 1.999996475346181e-05,
+      "loss": 1.3309,
+      "step": 151
+    },
+    {
+      "epoch": 0.03114754098360656,
+      "grad_norm": 13.743348121643066,
+      "learning_rate": 1.9999944927302274e-05,
+      "loss": 1.4131,
+      "step": 152
+    },
+    {
+      "epoch": 0.03135245901639344,
+      "grad_norm": 5.390098571777344,
+      "learning_rate": 1.9999920695347306e-05,
+      "loss": 1.1264,
+      "step": 153
+    },
+    {
+      "epoch": 0.03155737704918033,
+      "grad_norm": 7.308399200439453,
+      "learning_rate": 1.9999892057607574e-05,
+      "loss": 1.2324,
+      "step": 154
+    },
+    {
+      "epoch": 0.031762295081967214,
+      "grad_norm": 11.579313278198242,
+      "learning_rate": 1.9999859014095705e-05,
+      "loss": 1.4816,
+      "step": 155
+    },
+    {
+      "epoch": 0.031967213114754096,
+      "grad_norm": 5.121943473815918,
+      "learning_rate": 1.9999821564826244e-05,
+      "loss": 1.0447,
+      "step": 156
+    },
+    {
+      "epoch": 0.032172131147540986,
+      "grad_norm": 4.620405197143555,
+      "learning_rate": 1.9999779709815704e-05,
+      "loss": 1.0411,
+      "step": 157
+    },
+    {
+      "epoch": 0.03237704918032787,
+      "grad_norm": 6.9878830909729,
+      "learning_rate": 1.9999733449082514e-05,
+      "loss": 1.0816,
+      "step": 158
+    },
+    {
+      "epoch": 0.03258196721311475,
+      "grad_norm": 10.145071029663086,
+      "learning_rate": 1.9999682782647064e-05,
+      "loss": 0.9762,
+      "step": 159
+    },
+    {
+      "epoch": 0.03278688524590164,
+      "grad_norm": 19.290319442749023,
+      "learning_rate": 1.9999627710531675e-05,
+      "loss": 1.1309,
+      "step": 160
+    },
+    {
+      "epoch": 0.032991803278688524,
+      "grad_norm": 4.637401580810547,
+      "learning_rate": 1.999956823276061e-05,
+      "loss": 1.1428,
+      "step": 161
+    },
+    {
+      "epoch": 0.033196721311475406,
+      "grad_norm": 9.85617733001709,
+      "learning_rate": 1.9999504349360072e-05,
+      "loss": 1.1815,
+      "step": 162
+    },
+    {
+      "epoch": 0.033401639344262296,
+      "grad_norm": 5.839721202850342,
+      "learning_rate": 1.9999436060358208e-05,
+      "loss": 1.3807,
+      "step": 163
+    },
+    {
+      "epoch": 0.03360655737704918,
+      "grad_norm": 9.82158374786377,
+      "learning_rate": 1.9999363365785106e-05,
+      "loss": 1.1515,
+      "step": 164
+    },
+    {
+      "epoch": 0.03381147540983607,
+      "grad_norm": 24.74509620666504,
+      "learning_rate": 1.9999286265672794e-05,
+      "loss": 1.176,
+      "step": 165
+    },
+    {
+      "epoch": 0.03401639344262295,
+      "grad_norm": 34.61911392211914,
+      "learning_rate": 1.999920476005524e-05,
+      "loss": 1.859,
+      "step": 166
+    },
+    {
+      "epoch": 0.034221311475409834,
+      "grad_norm": 7.277236461639404,
+      "learning_rate": 1.9999118848968355e-05,
+      "loss": 1.2156,
+      "step": 167
+    },
+    {
+      "epoch": 0.03442622950819672,
+      "grad_norm": 5.595851898193359,
+      "learning_rate": 1.999902853244999e-05,
+      "loss": 1.0688,
+      "step": 168
+    },
+    {
+      "epoch": 0.034631147540983606,
+      "grad_norm": 11.632538795471191,
+      "learning_rate": 1.999893381053993e-05,
+      "loss": 0.9313,
+      "step": 169
+    },
+    {
+      "epoch": 0.03483606557377049,
+      "grad_norm": 7.2690534591674805,
+      "learning_rate": 1.9998834683279922e-05,
+      "loss": 1.1968,
+      "step": 170
+    },
+    {
+      "epoch": 0.03504098360655738,
+      "grad_norm": 4.0165300369262695,
+      "learning_rate": 1.9998731150713623e-05,
+      "loss": 0.9762,
+      "step": 171
+    },
+    {
+      "epoch": 0.03524590163934426,
+      "grad_norm": 8.033493041992188,
+      "learning_rate": 1.9998623212886662e-05,
+      "loss": 1.4223,
+      "step": 172
+    },
+    {
+      "epoch": 0.03545081967213115,
+      "grad_norm": 6.4791388511657715,
+      "learning_rate": 1.9998510869846586e-05,
+      "loss": 1.2952,
+      "step": 173
+    },
+    {
+      "epoch": 0.035655737704918034,
+      "grad_norm": 7.656297206878662,
+      "learning_rate": 1.999839412164289e-05,
+      "loss": 1.4808,
+      "step": 174
+    },
+    {
+      "epoch": 0.035860655737704916,
+      "grad_norm": 5.531428337097168,
+      "learning_rate": 1.9998272968327023e-05,
+      "loss": 1.0726,
+      "step": 175
+    },
+    {
+      "epoch": 0.036065573770491806,
+      "grad_norm": 7.457252025604248,
+      "learning_rate": 1.9998147409952348e-05,
+      "loss": 1.3183,
+      "step": 176
+    },
+    {
+      "epoch": 0.03627049180327869,
+      "grad_norm": 4.94792366027832,
+      "learning_rate": 1.9998017446574195e-05,
+      "loss": 1.2482,
+      "step": 177
+    },
+    {
+      "epoch": 0.03647540983606557,
+      "grad_norm": 5.549227714538574,
+      "learning_rate": 1.9997883078249816e-05,
+      "loss": 1.2386,
+      "step": 178
+    },
+    {
+      "epoch": 0.03668032786885246,
+      "grad_norm": 8.272525787353516,
+      "learning_rate": 1.9997744305038415e-05,
+      "loss": 1.3698,
+      "step": 179
+    },
+    {
+      "epoch": 0.036885245901639344,
+      "grad_norm": 7.554256439208984,
+      "learning_rate": 1.999760112700113e-05,
+      "loss": 1.434,
+      "step": 180
+    },
+    {
+      "epoch": 0.037090163934426226,
+      "grad_norm": 4.72944450378418,
+      "learning_rate": 1.9997453544201048e-05,
+      "loss": 1.0464,
+      "step": 181
+    },
+    {
+      "epoch": 0.037295081967213116,
+      "grad_norm": 6.418345928192139,
+      "learning_rate": 1.999730155670319e-05,
+      "loss": 1.2032,
+      "step": 182
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 5.083874702453613,
+      "learning_rate": 1.9997145164574516e-05,
+      "loss": 1.3472,
+      "step": 183
+    },
+    {
+      "epoch": 0.03770491803278689,
+      "grad_norm": 5.423723220825195,
+      "learning_rate": 1.9996984367883927e-05,
+      "loss": 1.0647,
+      "step": 184
+    },
+    {
+      "epoch": 0.03790983606557377,
+      "grad_norm": 6.5748467445373535,
+      "learning_rate": 1.9996819166702274e-05,
+      "loss": 0.9607,
+      "step": 185
+    },
+    {
+      "epoch": 0.038114754098360654,
+      "grad_norm": 6.006103038787842,
+      "learning_rate": 1.999664956110234e-05,
+      "loss": 1.163,
+      "step": 186
+    },
+    {
+      "epoch": 0.03831967213114754,
+      "grad_norm": 9.075021743774414,
+      "learning_rate": 1.9996475551158846e-05,
+      "loss": 1.7353,
+      "step": 187
+    },
+    {
+      "epoch": 0.038524590163934426,
+      "grad_norm": 6.526434421539307,
+      "learning_rate": 1.9996297136948462e-05,
+      "loss": 1.1265,
+      "step": 188
+    },
+    {
+      "epoch": 0.03872950819672131,
+      "grad_norm": 8.984017372131348,
+      "learning_rate": 1.9996114318549787e-05,
+      "loss": 1.2191,
+      "step": 189
+    },
+    {
+      "epoch": 0.0389344262295082,
+      "grad_norm": 18.00693702697754,
+      "learning_rate": 1.999592709604338e-05,
+      "loss": 1.1598,
+      "step": 190
+    },
+    {
+      "epoch": 0.03913934426229508,
+      "grad_norm": 6.52983283996582,
+      "learning_rate": 1.999573546951172e-05,
+      "loss": 1.1296,
+      "step": 191
+    },
+    {
+      "epoch": 0.03934426229508197,
+      "grad_norm": 15.859800338745117,
+      "learning_rate": 1.9995539439039232e-05,
+      "loss": 1.2346,
+      "step": 192
+    },
+    {
+      "epoch": 0.03954918032786885,
+      "grad_norm": 5.02130126953125,
+      "learning_rate": 1.999533900471229e-05,
+      "loss": 1.3252,
+      "step": 193
+    },
+    {
+      "epoch": 0.039754098360655736,
+      "grad_norm": 5.483813285827637,
+      "learning_rate": 1.999513416661919e-05,
+      "loss": 1.2937,
+      "step": 194
+    },
+    {
+      "epoch": 0.039959016393442626,
+      "grad_norm": 11.05758285522461,
+      "learning_rate": 1.9994924924850196e-05,
+      "loss": 1.2846,
+      "step": 195
+    },
+    {
+      "epoch": 0.04016393442622951,
+      "grad_norm": 5.94422721862793,
+      "learning_rate": 1.9994711279497488e-05,
+      "loss": 1.0385,
+      "step": 196
+    },
+    {
+      "epoch": 0.04036885245901639,
+      "grad_norm": 3.9777755737304688,
+      "learning_rate": 1.9994493230655193e-05,
+      "loss": 1.1731,
+      "step": 197
+    },
+    {
+      "epoch": 0.04057377049180328,
+      "grad_norm": 7.8311767578125,
+      "learning_rate": 1.9994270778419383e-05,
+      "loss": 1.3207,
+      "step": 198
+    },
+    {
+      "epoch": 0.04077868852459016,
+      "grad_norm": 9.597280502319336,
+      "learning_rate": 1.999404392288806e-05,
+      "loss": 0.9855,
+      "step": 199
+    },
+    {
+      "epoch": 0.040983606557377046,
+      "grad_norm": 8.792752265930176,
+      "learning_rate": 1.999381266416118e-05,
+      "loss": 1.3882,
+      "step": 200
+    },
+    {
+      "epoch": 0.041188524590163936,
+      "grad_norm": 4.946412086486816,
+      "learning_rate": 1.9993577002340628e-05,
+      "loss": 1.1343,
+      "step": 201
+    },
+    {
+      "epoch": 0.04139344262295082,
+      "grad_norm": 6.381492614746094,
+      "learning_rate": 1.999333693753023e-05,
+      "loss": 1.3434,
+      "step": 202
+    },
+    {
+      "epoch": 0.04159836065573771,
+      "grad_norm": 8.755156517028809,
+      "learning_rate": 1.999309246983576e-05,
+      "loss": 1.3132,
+      "step": 203
+    },
+    {
+      "epoch": 0.04180327868852459,
+      "grad_norm": 7.955817222595215,
+      "learning_rate": 1.9992843599364925e-05,
+      "loss": 1.3179,
+      "step": 204
+    },
+    {
+      "epoch": 0.042008196721311473,
+      "grad_norm": 6.036494731903076,
+      "learning_rate": 1.999259032622737e-05,
+      "loss": 1.0803,
+      "step": 205
+    },
+    {
+      "epoch": 0.04221311475409836,
+      "grad_norm": 4.87514591217041,
+      "learning_rate": 1.999233265053468e-05,
+      "loss": 1.1388,
+      "step": 206
+    },
+    {
+      "epoch": 0.042418032786885246,
+      "grad_norm": 7.108249664306641,
+      "learning_rate": 1.999207057240039e-05,
+      "loss": 1.071,
+      "step": 207
+    },
+    {
+      "epoch": 0.04262295081967213,
+      "grad_norm": 4.926052093505859,
+      "learning_rate": 1.999180409193996e-05,
+      "loss": 1.28,
+      "step": 208
+    },
+    {
+      "epoch": 0.04282786885245902,
+      "grad_norm": 6.915220737457275,
+      "learning_rate": 1.99915332092708e-05,
+      "loss": 1.0994,
+      "step": 209
+    },
+    {
+      "epoch": 0.0430327868852459,
+      "grad_norm": 6.172661781311035,
+      "learning_rate": 1.9991257924512252e-05,
+      "loss": 1.1291,
+      "step": 210
+    },
+    {
+      "epoch": 0.043237704918032784,
+      "grad_norm": 5.505499839782715,
+      "learning_rate": 1.9990978237785606e-05,
+      "loss": 1.1309,
+      "step": 211
+    },
+    {
+      "epoch": 0.04344262295081967,
+      "grad_norm": 5.355732440948486,
+      "learning_rate": 1.9990694149214088e-05,
+      "loss": 1.1633,
+      "step": 212
+    },
+    {
+      "epoch": 0.043647540983606556,
+      "grad_norm": 5.880125522613525,
+      "learning_rate": 1.999040565892286e-05,
+      "loss": 1.0888,
+      "step": 213
+    },
+    {
+      "epoch": 0.043852459016393446,
+      "grad_norm": 6.315762519836426,
+      "learning_rate": 1.999011276703902e-05,
+      "loss": 1.2143,
+      "step": 214
+    },
+    {
+      "epoch": 0.04405737704918033,
+      "grad_norm": 6.737286567687988,
+      "learning_rate": 1.9989815473691623e-05,
+      "loss": 1.0742,
+      "step": 215
+    },
+    {
+      "epoch": 0.04426229508196721,
+      "grad_norm": 5.956904411315918,
+      "learning_rate": 1.998951377901164e-05,
+      "loss": 1.0593,
+      "step": 216
+    },
+    {
+      "epoch": 0.0444672131147541,
+      "grad_norm": 12.217978477478027,
+      "learning_rate": 1.9989207683131998e-05,
+      "loss": 1.156,
+      "step": 217
+    },
+    {
+      "epoch": 0.04467213114754098,
+      "grad_norm": 6.168445587158203,
+      "learning_rate": 1.9988897186187555e-05,
+      "loss": 1.1134,
+      "step": 218
+    },
+    {
+      "epoch": 0.044877049180327866,
+      "grad_norm": 3.915191888809204,
+      "learning_rate": 1.9988582288315115e-05,
+      "loss": 1.0915,
+      "step": 219
+    },
+    {
+      "epoch": 0.045081967213114756,
+      "grad_norm": 6.489718437194824,
+      "learning_rate": 1.998826298965341e-05,
+      "loss": 1.2011,
+      "step": 220
+    },
+    {
+      "epoch": 0.04528688524590164,
+      "grad_norm": 12.514413833618164,
+      "learning_rate": 1.998793929034312e-05,
+      "loss": 1.2259,
+      "step": 221
+    },
+    {
+      "epoch": 0.04549180327868853,
+      "grad_norm": 4.124410152435303,
+      "learning_rate": 1.998761119052686e-05,
+      "loss": 1.0253,
+      "step": 222
+    },
+    {
+      "epoch": 0.04569672131147541,
+      "grad_norm": 6.930433750152588,
+      "learning_rate": 1.9987278690349186e-05,
+      "loss": 0.887,
+      "step": 223
+    },
+    {
+      "epoch": 0.04590163934426229,
+      "grad_norm": 4.603182315826416,
+      "learning_rate": 1.9986941789956593e-05,
+      "loss": 1.2128,
+      "step": 224
+    },
+    {
+      "epoch": 0.04610655737704918,
+      "grad_norm": 4.752573490142822,
+      "learning_rate": 1.998660048949751e-05,
+      "loss": 1.1586,
+      "step": 225
+    },
+    {
+      "epoch": 0.046311475409836066,
+      "grad_norm": 6.711463928222656,
+      "learning_rate": 1.998625478912231e-05,
+      "loss": 0.946,
+      "step": 226
+    },
+    {
+      "epoch": 0.04651639344262295,
+      "grad_norm": 5.96216344833374,
+      "learning_rate": 1.9985904688983303e-05,
+      "loss": 1.1318,
+      "step": 227
+    },
+    {
+      "epoch": 0.04672131147540984,
+      "grad_norm": 9.748722076416016,
+      "learning_rate": 1.9985550189234733e-05,
+      "loss": 1.3274,
+      "step": 228
+    },
+    {
+      "epoch": 0.04692622950819672,
+      "grad_norm": 8.127034187316895,
+      "learning_rate": 1.9985191290032794e-05,
+      "loss": 1.3141,
+      "step": 229
+    },
+    {
+      "epoch": 0.0471311475409836,
+      "grad_norm": 6.898861408233643,
+      "learning_rate": 1.9984827991535602e-05,
+      "loss": 1.2519,
+      "step": 230
+    },
+    {
+      "epoch": 0.04733606557377049,
+      "grad_norm": 5.1891374588012695,
+      "learning_rate": 1.998446029390322e-05,
+      "loss": 1.2144,
+      "step": 231
+    },
+    {
+      "epoch": 0.047540983606557376,
+      "grad_norm": 4.222479820251465,
+      "learning_rate": 1.9984088197297654e-05,
+      "loss": 1.3694,
+      "step": 232
+    },
+    {
+      "epoch": 0.047745901639344265,
+      "grad_norm": 4.857460975646973,
+      "learning_rate": 1.9983711701882844e-05,
+      "loss": 1.0404,
+      "step": 233
+    },
+    {
+      "epoch": 0.04795081967213115,
+      "grad_norm": 8.265710830688477,
+      "learning_rate": 1.9983330807824664e-05,
+      "loss": 1.1649,
+      "step": 234
+    },
+    {
+      "epoch": 0.04815573770491803,
+      "grad_norm": 11.158879280090332,
+      "learning_rate": 1.9982945515290926e-05,
+      "loss": 1.4552,
+      "step": 235
+    },
+    {
+      "epoch": 0.04836065573770492,
+      "grad_norm": 2.9954006671905518,
+      "learning_rate": 1.998255582445139e-05,
+      "loss": 1.0138,
+      "step": 236
+    },
+    {
+      "epoch": 0.0485655737704918,
+      "grad_norm": 8.821298599243164,
+      "learning_rate": 1.998216173547774e-05,
+      "loss": 1.1865,
+      "step": 237
+    },
+    {
+      "epoch": 0.048770491803278686,
+      "grad_norm": 5.960107326507568,
+      "learning_rate": 1.998176324854361e-05,
+      "loss": 1.0127,
+      "step": 238
+    },
+    {
+      "epoch": 0.048975409836065575,
+      "grad_norm": 6.328037261962891,
+      "learning_rate": 1.998136036382456e-05,
+      "loss": 1.1963,
+      "step": 239
+    },
+    {
+      "epoch": 0.04918032786885246,
+      "grad_norm": 4.278129577636719,
+      "learning_rate": 1.99809530814981e-05,
+      "loss": 1.1689,
+      "step": 240
+    },
+    {
+      "epoch": 0.04938524590163934,
+      "grad_norm": 5.412634372711182,
+      "learning_rate": 1.9980541401743666e-05,
+      "loss": 1.1342,
+      "step": 241
+    },
+    {
+      "epoch": 0.04959016393442623,
+      "grad_norm": 6.392723560333252,
+      "learning_rate": 1.998012532474264e-05,
+      "loss": 1.1325,
+      "step": 242
+    },
+    {
+      "epoch": 0.04979508196721311,
+      "grad_norm": 4.8896164894104,
+      "learning_rate": 1.9979704850678342e-05,
+      "loss": 1.1705,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.368144989013672,
+      "learning_rate": 1.9979279979736016e-05,
+      "loss": 0.9812,
+      "step": 244
+    },
+    {
+      "epoch": 0.050204918032786885,
+      "grad_norm": 7.6163177490234375,
+      "learning_rate": 1.997885071210286e-05,
+      "loss": 1.1348,
+      "step": 245
+    },
+    {
+      "epoch": 0.05040983606557377,
+      "grad_norm": 6.862471103668213,
+      "learning_rate": 1.9978417047967996e-05,
+      "loss": 1.1732,
+      "step": 246
+    },
+    {
+      "epoch": 0.05061475409836066,
+      "grad_norm": 5.668833255767822,
+      "learning_rate": 1.997797898752249e-05,
+      "loss": 1.0762,
+      "step": 247
+    },
+    {
+      "epoch": 0.05081967213114754,
+      "grad_norm": 5.717274188995361,
+      "learning_rate": 1.997753653095935e-05,
+      "loss": 1.1673,
+      "step": 248
+    },
+    {
+      "epoch": 0.05102459016393442,
+      "grad_norm": 5.3873610496521,
+      "learning_rate": 1.9977089678473505e-05,
+      "loss": 1.1433,
+      "step": 249
+    },
+    {
+      "epoch": 0.05122950819672131,
+      "grad_norm": 10.489523887634277,
+      "learning_rate": 1.9976638430261836e-05,
+      "loss": 1.4107,
+      "step": 250
+    },
+    {
+      "epoch": 0.051434426229508196,
+      "grad_norm": 4.740592002868652,
+      "learning_rate": 1.9976182786523154e-05,
+      "loss": 1.0148,
+      "step": 251
+    },
+    {
+      "epoch": 0.051639344262295085,
+      "grad_norm": 4.611098766326904,
+      "learning_rate": 1.9975722747458203e-05,
+      "loss": 1.4058,
+      "step": 252
+    },
+    {
+      "epoch": 0.05184426229508197,
+      "grad_norm": 4.639456272125244,
+      "learning_rate": 1.9975258313269674e-05,
+      "loss": 1.1427,
+      "step": 253
+    },
+    {
+      "epoch": 0.05204918032786885,
+      "grad_norm": 4.333242893218994,
+      "learning_rate": 1.9974789484162187e-05,
+      "loss": 1.2699,
+      "step": 254
+    },
+    {
+      "epoch": 0.05225409836065574,
+      "grad_norm": 4.731597423553467,
+      "learning_rate": 1.99743162603423e-05,
+      "loss": 1.0687,
+      "step": 255
+    },
+    {
+      "epoch": 0.05245901639344262,
+      "grad_norm": 6.71632194519043,
+      "learning_rate": 1.9973838642018498e-05,
+      "loss": 1.057,
+      "step": 256
+    },
+    {
+      "epoch": 0.052663934426229506,
+      "grad_norm": 10.081603050231934,
+      "learning_rate": 1.9973356629401225e-05,
+      "loss": 1.0457,
+      "step": 257
+    },
+    {
+      "epoch": 0.052868852459016395,
+      "grad_norm": 4.13563871383667,
+      "learning_rate": 1.997287022270284e-05,
+      "loss": 1.0271,
+      "step": 258
+    },
+    {
+      "epoch": 0.05307377049180328,
+      "grad_norm": 13.729360580444336,
+      "learning_rate": 1.997237942213764e-05,
+      "loss": 0.9735,
+      "step": 259
+    },
+    {
+      "epoch": 0.05327868852459016,
+      "grad_norm": 8.7944974899292,
+      "learning_rate": 1.9971884227921872e-05,
+      "loss": 1.175,
+      "step": 260
+    },
+    {
+      "epoch": 0.05348360655737705,
+      "grad_norm": 3.3799877166748047,
+      "learning_rate": 1.997138464027371e-05,
+      "loss": 1.21,
+      "step": 261
+    },
+    {
+      "epoch": 0.05368852459016393,
+      "grad_norm": 7.876162052154541,
+      "learning_rate": 1.9970880659413253e-05,
+      "loss": 1.4368,
+      "step": 262
+    },
+    {
+      "epoch": 0.05389344262295082,
+      "grad_norm": 10.842589378356934,
+      "learning_rate": 1.9970372285562556e-05,
+      "loss": 1.0508,
+      "step": 263
+    },
+    {
+      "epoch": 0.054098360655737705,
+      "grad_norm": 3.8059465885162354,
+      "learning_rate": 1.9969859518945592e-05,
+      "loss": 1.0718,
+      "step": 264
+    },
+    {
+      "epoch": 0.05430327868852459,
+      "grad_norm": 7.259851932525635,
+      "learning_rate": 1.9969342359788283e-05,
+      "loss": 1.2139,
+      "step": 265
+    },
+    {
+      "epoch": 0.05450819672131148,
+      "grad_norm": 7.442305564880371,
+      "learning_rate": 1.996882080831848e-05,
+      "loss": 1.3024,
+      "step": 266
+    },
+    {
+      "epoch": 0.05471311475409836,
+      "grad_norm": 10.548867225646973,
+      "learning_rate": 1.996829486476596e-05,
+      "loss": 1.1474,
+      "step": 267
+    },
+    {
+      "epoch": 0.05491803278688524,
+      "grad_norm": 7.676911354064941,
+      "learning_rate": 1.9967764529362456e-05,
+      "loss": 1.0151,
+      "step": 268
+    },
+    {
+      "epoch": 0.05512295081967213,
+      "grad_norm": 4.766316890716553,
+      "learning_rate": 1.9967229802341616e-05,
+      "loss": 1.0427,
+      "step": 269
+    },
+    {
+      "epoch": 0.055327868852459015,
+      "grad_norm": 9.710952758789062,
+      "learning_rate": 1.9966690683939033e-05,
+      "loss": 1.1689,
+      "step": 270
+    },
+    {
+      "epoch": 0.055532786885245905,
+      "grad_norm": 5.249614238739014,
+      "learning_rate": 1.996614717439224e-05,
+      "loss": 0.9576,
+      "step": 271
+    },
+    {
+      "epoch": 0.05573770491803279,
+      "grad_norm": 4.558048725128174,
+      "learning_rate": 1.9965599273940682e-05,
+      "loss": 1.1562,
+      "step": 272
+    },
+    {
+      "epoch": 0.05594262295081967,
+      "grad_norm": 2.996655225753784,
+      "learning_rate": 1.996504698282577e-05,
+      "loss": 1.1962,
+      "step": 273
+    },
+    {
+      "epoch": 0.05614754098360656,
+      "grad_norm": 8.341348648071289,
+      "learning_rate": 1.9964490301290827e-05,
+      "loss": 1.247,
+      "step": 274
+    },
+    {
+      "epoch": 0.05635245901639344,
+      "grad_norm": 6.110327243804932,
+      "learning_rate": 1.9963929229581118e-05,
+      "loss": 1.306,
+      "step": 275
+    },
+    {
+      "epoch": 0.056557377049180325,
+      "grad_norm": 9.551907539367676,
+      "learning_rate": 1.9963363767943832e-05,
+      "loss": 1.2698,
+      "step": 276
+    },
+    {
+      "epoch": 0.056762295081967215,
+      "grad_norm": 4.445561408996582,
+      "learning_rate": 1.9962793916628117e-05,
+      "loss": 1.2703,
+      "step": 277
+    },
+    {
+      "epoch": 0.0569672131147541,
+      "grad_norm": 4.591422080993652,
+      "learning_rate": 1.9962219675885026e-05,
+      "loss": 1.148,
+      "step": 278
+    },
+    {
+      "epoch": 0.05717213114754098,
+      "grad_norm": 7.946553707122803,
+      "learning_rate": 1.9961641045967564e-05,
+      "loss": 0.9461,
+      "step": 279
+    },
+    {
+      "epoch": 0.05737704918032787,
+      "grad_norm": 7.723875999450684,
+      "learning_rate": 1.996105802713067e-05,
+      "loss": 1.2234,
+      "step": 280
+    },
+    {
+      "epoch": 0.05758196721311475,
+      "grad_norm": 5.6502509117126465,
+      "learning_rate": 1.9960470619631206e-05,
+      "loss": 1.2315,
+      "step": 281
+    },
+    {
+      "epoch": 0.05778688524590164,
+      "grad_norm": 4.188591480255127,
+      "learning_rate": 1.9959878823727972e-05,
+      "loss": 1.0057,
+      "step": 282
+    },
+    {
+      "epoch": 0.057991803278688525,
+      "grad_norm": 6.996484279632568,
+      "learning_rate": 1.995928263968171e-05,
+      "loss": 1.3121,
+      "step": 283
+    },
+    {
+      "epoch": 0.05819672131147541,
+      "grad_norm": 6.087568759918213,
+      "learning_rate": 1.9958682067755073e-05,
+      "loss": 1.4539,
+      "step": 284
+    },
+    {
+      "epoch": 0.0584016393442623,
+      "grad_norm": 4.1526408195495605,
+      "learning_rate": 1.995807710821268e-05,
+      "loss": 1.0459,
+      "step": 285
+    },
+    {
+      "epoch": 0.05860655737704918,
+      "grad_norm": 4.8373026847839355,
+      "learning_rate": 1.9957467761321053e-05,
+      "loss": 1.3683,
+      "step": 286
+    },
+    {
+      "epoch": 0.05881147540983606,
+      "grad_norm": 5.23197603225708,
+      "learning_rate": 1.9956854027348664e-05,
+      "loss": 0.9701,
+      "step": 287
+    },
+    {
+      "epoch": 0.05901639344262295,
+      "grad_norm": 4.430776596069336,
+      "learning_rate": 1.9956235906565912e-05,
+      "loss": 1.2406,
+      "step": 288
+    },
+    {
+      "epoch": 0.059221311475409835,
+      "grad_norm": 4.473460674285889,
+      "learning_rate": 1.9955613399245132e-05,
+      "loss": 1.1204,
+      "step": 289
+    },
+    {
+      "epoch": 0.05942622950819672,
+      "grad_norm": 4.795286655426025,
+      "learning_rate": 1.9954986505660587e-05,
+      "loss": 1.2583,
+      "step": 290
+    },
+    {
+      "epoch": 0.05963114754098361,
+      "grad_norm": 5.8868608474731445,
+      "learning_rate": 1.9954355226088472e-05,
+      "loss": 1.137,
+      "step": 291
+    },
+    {
+      "epoch": 0.05983606557377049,
+      "grad_norm": 5.758181095123291,
+      "learning_rate": 1.9953719560806925e-05,
+      "loss": 1.0261,
+      "step": 292
+    },
+    {
+      "epoch": 0.06004098360655738,
+      "grad_norm": 7.626696586608887,
+      "learning_rate": 1.9953079510096004e-05,
+      "loss": 1.011,
+      "step": 293
+    },
+    {
+      "epoch": 0.06024590163934426,
+      "grad_norm": 6.3893914222717285,
+      "learning_rate": 1.9952435074237703e-05,
+      "loss": 1.0937,
+      "step": 294
+    },
+    {
+      "epoch": 0.060450819672131145,
+      "grad_norm": 6.119849681854248,
+      "learning_rate": 1.995178625351595e-05,
+      "loss": 1.0437,
+      "step": 295
+    },
+    {
+      "epoch": 0.060655737704918035,
+      "grad_norm": 7.111720561981201,
+      "learning_rate": 1.9951133048216608e-05,
+      "loss": 1.163,
+      "step": 296
+    },
+    {
+      "epoch": 0.06086065573770492,
+      "grad_norm": 6.381929397583008,
+      "learning_rate": 1.9950475458627463e-05,
+      "loss": 0.9624,
+      "step": 297
+    },
+    {
+      "epoch": 0.0610655737704918,
+      "grad_norm": 6.1409101486206055,
+      "learning_rate": 1.9949813485038237e-05,
+      "loss": 1.0425,
+      "step": 298
+    },
+    {
+      "epoch": 0.06127049180327869,
+      "grad_norm": 5.093874454498291,
+      "learning_rate": 1.9949147127740585e-05,
+      "loss": 1.1769,
+      "step": 299
+    },
+    {
+      "epoch": 0.06147540983606557,
+      "grad_norm": 5.422652721405029,
+      "learning_rate": 1.994847638702809e-05,
+      "loss": 1.3558,
+      "step": 300
+    },
+    {
+      "epoch": 0.06168032786885246,
+      "grad_norm": 7.927038669586182,
+      "learning_rate": 1.994780126319627e-05,
+      "loss": 1.2094,
+      "step": 301
+    },
+    {
+      "epoch": 0.061885245901639345,
+      "grad_norm": 4.5818047523498535,
+      "learning_rate": 1.9947121756542573e-05,
+      "loss": 1.0354,
+      "step": 302
+    },
+    {
+      "epoch": 0.06209016393442623,
+      "grad_norm": 7.267391681671143,
+      "learning_rate": 1.9946437867366373e-05,
+      "loss": 1.0968,
+      "step": 303
+    },
+    {
+      "epoch": 0.06229508196721312,
+      "grad_norm": 3.929089307785034,
+      "learning_rate": 1.9945749595968985e-05,
+      "loss": 1.1785,
+      "step": 304
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 5.291886806488037,
+      "learning_rate": 1.994505694265365e-05,
+      "loss": 1.3982,
+      "step": 305
+    },
+    {
+      "epoch": 0.06270491803278688,
+      "grad_norm": 3.852639675140381,
+      "learning_rate": 1.994435990772553e-05,
+      "loss": 1.1604,
+      "step": 306
+    },
+    {
+      "epoch": 0.06290983606557377,
+      "grad_norm": 11.10228157043457,
+      "learning_rate": 1.9943658491491734e-05,
+      "loss": 1.3263,
+      "step": 307
+    },
+    {
+      "epoch": 0.06311475409836066,
+      "grad_norm": 4.713088512420654,
+      "learning_rate": 1.994295269426129e-05,
+      "loss": 1.147,
+      "step": 308
+    },
+    {
+      "epoch": 0.06331967213114754,
+      "grad_norm": 6.204249858856201,
+      "learning_rate": 1.994224251634516e-05,
+      "loss": 1.3193,
+      "step": 309
+    },
+    {
+      "epoch": 0.06352459016393443,
+      "grad_norm": 10.645562171936035,
+      "learning_rate": 1.9941527958056235e-05,
+      "loss": 1.178,
+      "step": 310
+    },
+    {
+      "epoch": 0.06372950819672131,
+      "grad_norm": 4.138010501861572,
+      "learning_rate": 1.994080901970934e-05,
+      "loss": 1.0434,
+      "step": 311
+    },
+    {
+      "epoch": 0.06393442622950819,
+      "grad_norm": 8.684898376464844,
+      "learning_rate": 1.994008570162122e-05,
+      "loss": 1.1404,
+      "step": 312
+    },
+    {
+      "epoch": 0.06413934426229508,
+      "grad_norm": 4.833040237426758,
+      "learning_rate": 1.9939358004110562e-05,
+      "loss": 1.0985,
+      "step": 313
+    },
+    {
+      "epoch": 0.06434426229508197,
+      "grad_norm": 9.461345672607422,
+      "learning_rate": 1.993862592749797e-05,
+      "loss": 1.316,
+      "step": 314
+    },
+    {
+      "epoch": 0.06454918032786885,
+      "grad_norm": 5.740072727203369,
+      "learning_rate": 1.9937889472105988e-05,
+      "loss": 1.1458,
+      "step": 315
+    },
+    {
+      "epoch": 0.06475409836065574,
+      "grad_norm": 10.704619407653809,
+      "learning_rate": 1.993714863825909e-05,
+      "loss": 1.5645,
+      "step": 316
+    },
+    {
+      "epoch": 0.06495901639344262,
+      "grad_norm": 6.295975685119629,
+      "learning_rate": 1.9936403426283665e-05,
+      "loss": 1.3762,
+      "step": 317
+    },
+    {
+      "epoch": 0.0651639344262295,
+      "grad_norm": 3.8343141078948975,
+      "learning_rate": 1.9935653836508042e-05,
+      "loss": 1.0756,
+      "step": 318
+    },
+    {
+      "epoch": 0.0653688524590164,
+      "grad_norm": 3.62160325050354,
+      "learning_rate": 1.993489986926248e-05,
+      "loss": 0.8286,
+      "step": 319
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 4.521775245666504,
+      "learning_rate": 1.9934141524879163e-05,
+      "loss": 1.2028,
+      "step": 320
+    },
+    {
+      "epoch": 0.06577868852459016,
+      "grad_norm": 6.055569171905518,
+      "learning_rate": 1.9933378803692197e-05,
+      "loss": 1.1916,
+      "step": 321
+    },
+    {
+      "epoch": 0.06598360655737705,
+      "grad_norm": 4.80081844329834,
+      "learning_rate": 1.9932611706037635e-05,
+      "loss": 1.1373,
+      "step": 322
+    },
+    {
+      "epoch": 0.06618852459016393,
+      "grad_norm": 6.640129089355469,
+      "learning_rate": 1.9931840232253436e-05,
+      "loss": 1.0955,
+      "step": 323
+    },
+    {
+      "epoch": 0.06639344262295081,
+      "grad_norm": 5.279435634613037,
+      "learning_rate": 1.9931064382679502e-05,
+      "loss": 1.2492,
+      "step": 324
+    },
+    {
+      "epoch": 0.06659836065573771,
+      "grad_norm": 4.472193717956543,
+      "learning_rate": 1.993028415765766e-05,
+      "loss": 1.2006,
+      "step": 325
+    },
+    {
+      "epoch": 0.06680327868852459,
+      "grad_norm": 7.69242525100708,
+      "learning_rate": 1.992949955753166e-05,
+      "loss": 1.0976,
+      "step": 326
+    },
+    {
+      "epoch": 0.06700819672131147,
+      "grad_norm": 7.352753162384033,
+      "learning_rate": 1.9928710582647183e-05,
+      "loss": 1.191,
+      "step": 327
+    },
+    {
+      "epoch": 0.06721311475409836,
+      "grad_norm": 3.708878517150879,
+      "learning_rate": 1.992791723335184e-05,
+      "loss": 0.8764,
+      "step": 328
+    },
+    {
+      "epoch": 0.06741803278688524,
+      "grad_norm": 9.256171226501465,
+      "learning_rate": 1.992711950999516e-05,
+      "loss": 1.0268,
+      "step": 329
+    },
+    {
+      "epoch": 0.06762295081967214,
+      "grad_norm": 8.49113655090332,
+      "learning_rate": 1.9926317412928612e-05,
+      "loss": 1.0476,
+      "step": 330
+    },
+    {
+      "epoch": 0.06782786885245902,
+      "grad_norm": 5.7638349533081055,
+      "learning_rate": 1.9925510942505583e-05,
+      "loss": 0.9546,
+      "step": 331
+    },
+    {
+      "epoch": 0.0680327868852459,
+      "grad_norm": 6.705662727355957,
+      "learning_rate": 1.992470009908139e-05,
+      "loss": 1.2159,
+      "step": 332
+    },
+    {
+      "epoch": 0.06823770491803278,
+      "grad_norm": 6.94514799118042,
+      "learning_rate": 1.992388488301327e-05,
+      "loss": 1.2782,
+      "step": 333
+    },
+    {
+      "epoch": 0.06844262295081967,
+      "grad_norm": 6.104513168334961,
+      "learning_rate": 1.9923065294660407e-05,
+      "loss": 0.9341,
+      "step": 334
+    },
+    {
+      "epoch": 0.06864754098360656,
+      "grad_norm": 9.277979850769043,
+      "learning_rate": 1.9922241334383884e-05,
+      "loss": 1.6393,
+      "step": 335
+    },
+    {
+      "epoch": 0.06885245901639345,
+      "grad_norm": 3.9294443130493164,
+      "learning_rate": 1.9921413002546724e-05,
+      "loss": 0.9451,
+      "step": 336
+    },
+    {
+      "epoch": 0.06905737704918033,
+      "grad_norm": 5.987906455993652,
+      "learning_rate": 1.9920580299513878e-05,
+      "loss": 1.1584,
+      "step": 337
+    },
+    {
+      "epoch": 0.06926229508196721,
+      "grad_norm": 5.073190212249756,
+      "learning_rate": 1.9919743225652223e-05,
+      "loss": 1.3076,
+      "step": 338
+    },
+    {
+      "epoch": 0.0694672131147541,
+      "grad_norm": 4.323671817779541,
+      "learning_rate": 1.9918901781330556e-05,
+      "loss": 1.1024,
+      "step": 339
+    },
+    {
+      "epoch": 0.06967213114754098,
+      "grad_norm": 4.659067630767822,
+      "learning_rate": 1.9918055966919597e-05,
+      "loss": 1.317,
+      "step": 340
+    },
+    {
+      "epoch": 0.06987704918032787,
+      "grad_norm": 18.332191467285156,
+      "learning_rate": 1.9917205782792007e-05,
+      "loss": 1.0331,
+      "step": 341
+    },
+    {
+      "epoch": 0.07008196721311476,
+      "grad_norm": 4.506664752960205,
+      "learning_rate": 1.9916351229322353e-05,
+      "loss": 1.251,
+      "step": 342
+    },
+    {
+      "epoch": 0.07028688524590164,
+      "grad_norm": 4.723790645599365,
+      "learning_rate": 1.991549230688714e-05,
+      "loss": 1.1034,
+      "step": 343
+    },
+    {
+      "epoch": 0.07049180327868852,
+      "grad_norm": 4.463237285614014,
+      "learning_rate": 1.991462901586479e-05,
+      "loss": 1.1163,
+      "step": 344
+    },
+    {
+      "epoch": 0.0706967213114754,
+      "grad_norm": 4.143716335296631,
+      "learning_rate": 1.991376135663566e-05,
+      "loss": 1.4264,
+      "step": 345
+    },
+    {
+      "epoch": 0.0709016393442623,
+      "grad_norm": 7.441681861877441,
+      "learning_rate": 1.9912889329582018e-05,
+      "loss": 1.3288,
+      "step": 346
+    },
+    {
+      "epoch": 0.07110655737704918,
+      "grad_norm": 5.005547523498535,
+      "learning_rate": 1.991201293508807e-05,
+      "loss": 1.2485,
+      "step": 347
+    },
+    {
+      "epoch": 0.07131147540983607,
+      "grad_norm": 13.519057273864746,
+      "learning_rate": 1.991113217353993e-05,
+      "loss": 1.4056,
+      "step": 348
+    },
+    {
+      "epoch": 0.07151639344262295,
+      "grad_norm": 7.111389636993408,
+      "learning_rate": 1.9910247045325658e-05,
+      "loss": 1.198,
+      "step": 349
+    },
+    {
+      "epoch": 0.07172131147540983,
+      "grad_norm": 10.395705223083496,
+      "learning_rate": 1.9909357550835214e-05,
+      "loss": 1.501,
+      "step": 350
+    },
+    {
+      "epoch": 0.07192622950819672,
+      "grad_norm": 4.163288593292236,
+      "learning_rate": 1.9908463690460506e-05,
+      "loss": 1.1692,
+      "step": 351
+    },
+    {
+      "epoch": 0.07213114754098361,
+      "grad_norm": 4.662473678588867,
+      "learning_rate": 1.990756546459534e-05,
+      "loss": 1.4471,
+      "step": 352
+    },
+    {
+      "epoch": 0.0723360655737705,
+      "grad_norm": 8.750268936157227,
+      "learning_rate": 1.9906662873635463e-05,
+      "loss": 1.0304,
+      "step": 353
+    },
+    {
+      "epoch": 0.07254098360655738,
+      "grad_norm": 8.17618465423584,
+      "learning_rate": 1.9905755917978542e-05,
+      "loss": 1.2393,
+      "step": 354
+    },
+    {
+      "epoch": 0.07274590163934426,
+      "grad_norm": 8.16373348236084,
+      "learning_rate": 1.9904844598024166e-05,
+      "loss": 0.9603,
+      "step": 355
+    },
+    {
+      "epoch": 0.07295081967213114,
+      "grad_norm": 6.389175891876221,
+      "learning_rate": 1.990392891417384e-05,
+      "loss": 1.0792,
+      "step": 356
+    },
+    {
+      "epoch": 0.07315573770491804,
+      "grad_norm": 12.539687156677246,
+      "learning_rate": 1.9903008866831005e-05,
+      "loss": 1.1687,
+      "step": 357
+    },
+    {
+      "epoch": 0.07336065573770492,
+      "grad_norm": 6.5345139503479,
+      "learning_rate": 1.9902084456401013e-05,
+      "loss": 0.8142,
+      "step": 358
+    },
+    {
+      "epoch": 0.0735655737704918,
+      "grad_norm": 5.326274871826172,
+      "learning_rate": 1.9901155683291147e-05,
+      "loss": 1.2746,
+      "step": 359
+    },
+    {
+      "epoch": 0.07377049180327869,
+      "grad_norm": 4.962457656860352,
+      "learning_rate": 1.99002225479106e-05,
+      "loss": 1.0346,
+      "step": 360
+    },
+    {
+      "epoch": 0.07397540983606557,
+      "grad_norm": 4.834855079650879,
+      "learning_rate": 1.9899285050670504e-05,
+      "loss": 1.394,
+      "step": 361
+    },
+    {
+      "epoch": 0.07418032786885245,
+      "grad_norm": 3.858365297317505,
+      "learning_rate": 1.9898343191983893e-05,
+      "loss": 0.8557,
+      "step": 362
+    },
+    {
+      "epoch": 0.07438524590163935,
+      "grad_norm": 4.282461166381836,
+      "learning_rate": 1.9897396972265745e-05,
+      "loss": 0.9287,
+      "step": 363
+    },
+    {
+      "epoch": 0.07459016393442623,
+      "grad_norm": 5.022656440734863,
+      "learning_rate": 1.9896446391932937e-05,
+      "loss": 1.0797,
+      "step": 364
+    },
+    {
+      "epoch": 0.07479508196721311,
+      "grad_norm": 3.3839757442474365,
+      "learning_rate": 1.989549145140428e-05,
+      "loss": 1.2377,
+      "step": 365
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 6.768479347229004,
+      "learning_rate": 1.9894532151100505e-05,
+      "loss": 1.2873,
+      "step": 366
+    },
+    {
+      "epoch": 0.07520491803278688,
+      "grad_norm": 3.928938627243042,
+      "learning_rate": 1.9893568491444265e-05,
+      "loss": 1.247,
+      "step": 367
+    },
+    {
+      "epoch": 0.07540983606557378,
+      "grad_norm": 4.926082611083984,
+      "learning_rate": 1.9892600472860127e-05,
+      "loss": 1.1658,
+      "step": 368
+    },
+    {
+      "epoch": 0.07561475409836066,
+      "grad_norm": 4.916016101837158,
+      "learning_rate": 1.9891628095774586e-05,
+      "loss": 1.241,
+      "step": 369
+    },
+    {
+      "epoch": 0.07581967213114754,
+      "grad_norm": 4.226110935211182,
+      "learning_rate": 1.989065136061605e-05,
+      "loss": 0.9728,
+      "step": 370
+    },
+    {
+      "epoch": 0.07602459016393442,
+      "grad_norm": 4.8184051513671875,
+      "learning_rate": 1.9889670267814852e-05,
+      "loss": 1.1645,
+      "step": 371
+    },
+    {
+      "epoch": 0.07622950819672131,
+      "grad_norm": 4.868440628051758,
+      "learning_rate": 1.9888684817803244e-05,
+      "loss": 1.194,
+      "step": 372
+    },
+    {
+      "epoch": 0.07643442622950819,
+      "grad_norm": 5.104640483856201,
+      "learning_rate": 1.98876950110154e-05,
+      "loss": 1.2531,
+      "step": 373
+    },
+    {
+      "epoch": 0.07663934426229509,
+      "grad_norm": 3.473788022994995,
+      "learning_rate": 1.988670084788741e-05,
+      "loss": 1.1471,
+      "step": 374
+    },
+    {
+      "epoch": 0.07684426229508197,
+      "grad_norm": 5.2743754386901855,
+      "learning_rate": 1.988570232885728e-05,
+      "loss": 1.2023,
+      "step": 375
+    },
+    {
+      "epoch": 0.07704918032786885,
+      "grad_norm": 12.526911735534668,
+      "learning_rate": 1.9884699454364943e-05,
+      "loss": 0.9044,
+      "step": 376
+    },
+    {
+      "epoch": 0.07725409836065573,
+      "grad_norm": 16.59382438659668,
+      "learning_rate": 1.988369222485225e-05,
+      "loss": 0.8985,
+      "step": 377
+    },
+    {
+      "epoch": 0.07745901639344262,
+      "grad_norm": 3.4979331493377686,
+      "learning_rate": 1.9882680640762966e-05,
+      "loss": 0.9137,
+      "step": 378
+    },
+    {
+      "epoch": 0.07766393442622951,
+      "grad_norm": 6.846384525299072,
+      "learning_rate": 1.9881664702542775e-05,
+      "loss": 1.1426,
+      "step": 379
+    },
+    {
+      "epoch": 0.0778688524590164,
+      "grad_norm": 6.987870216369629,
+      "learning_rate": 1.988064441063928e-05,
+      "loss": 1.2147,
+      "step": 380
+    },
+    {
+      "epoch": 0.07807377049180328,
+      "grad_norm": 5.567429542541504,
+      "learning_rate": 1.9879619765502008e-05,
+      "loss": 1.2676,
+      "step": 381
+    },
+    {
+      "epoch": 0.07827868852459016,
+      "grad_norm": 4.196963310241699,
+      "learning_rate": 1.9878590767582398e-05,
+      "loss": 1.0465,
+      "step": 382
+    },
+    {
+      "epoch": 0.07848360655737704,
+      "grad_norm": 3.4539272785186768,
+      "learning_rate": 1.98775574173338e-05,
+      "loss": 1.0422,
+      "step": 383
+    },
+    {
+      "epoch": 0.07868852459016394,
+      "grad_norm": 8.80219554901123,
+      "learning_rate": 1.9876519715211504e-05,
+      "loss": 1.4982,
+      "step": 384
+    },
+    {
+      "epoch": 0.07889344262295082,
+      "grad_norm": 4.855871677398682,
+      "learning_rate": 1.9875477661672692e-05,
+      "loss": 1.2142,
+      "step": 385
+    },
+    {
+      "epoch": 0.0790983606557377,
+      "grad_norm": 4.491757392883301,
+      "learning_rate": 1.9874431257176474e-05,
+      "loss": 1.2971,
+      "step": 386
+    },
+    {
+      "epoch": 0.07930327868852459,
+      "grad_norm": 7.661441802978516,
+      "learning_rate": 1.987338050218388e-05,
+      "loss": 1.0419,
+      "step": 387
+    },
+    {
+      "epoch": 0.07950819672131147,
+      "grad_norm": 4.889249801635742,
+      "learning_rate": 1.9872325397157856e-05,
+      "loss": 1.3603,
+      "step": 388
+    },
+    {
+      "epoch": 0.07971311475409835,
+      "grad_norm": 4.540982246398926,
+      "learning_rate": 1.987126594256326e-05,
+      "loss": 1.1722,
+      "step": 389
+    },
+    {
+      "epoch": 0.07991803278688525,
+      "grad_norm": 4.631284236907959,
+      "learning_rate": 1.987020213886687e-05,
+      "loss": 1.104,
+      "step": 390
+    },
+    {
+      "epoch": 0.08012295081967213,
+      "grad_norm": 7.405433654785156,
+      "learning_rate": 1.9869133986537372e-05,
+      "loss": 1.0899,
+      "step": 391
+    },
+    {
+      "epoch": 0.08032786885245902,
+      "grad_norm": 8.410371780395508,
+      "learning_rate": 1.986806148604538e-05,
+      "loss": 1.3015,
+      "step": 392
+    },
+    {
+      "epoch": 0.0805327868852459,
+      "grad_norm": 4.5368733406066895,
+      "learning_rate": 1.9866984637863424e-05,
+      "loss": 0.9434,
+      "step": 393
+    },
+    {
+      "epoch": 0.08073770491803278,
+      "grad_norm": 9.395214080810547,
+      "learning_rate": 1.9865903442465936e-05,
+      "loss": 1.1777,
+      "step": 394
+    },
+    {
+      "epoch": 0.08094262295081968,
+      "grad_norm": 7.398886203765869,
+      "learning_rate": 1.9864817900329273e-05,
+      "loss": 1.2935,
+      "step": 395
+    },
+    {
+      "epoch": 0.08114754098360656,
+      "grad_norm": 4.210666656494141,
+      "learning_rate": 1.9863728011931703e-05,
+      "loss": 1.1023,
+      "step": 396
+    },
+    {
+      "epoch": 0.08135245901639344,
+      "grad_norm": 3.5518016815185547,
+      "learning_rate": 1.9862633777753417e-05,
+      "loss": 1.041,
+      "step": 397
+    },
+    {
+      "epoch": 0.08155737704918033,
+      "grad_norm": 5.6375651359558105,
+      "learning_rate": 1.9861535198276507e-05,
+      "loss": 1.302,
+      "step": 398
+    },
+    {
+      "epoch": 0.08176229508196721,
+      "grad_norm": 3.7568047046661377,
+      "learning_rate": 1.9860432273985e-05,
+      "loss": 1.0107,
+      "step": 399
+    },
+    {
+      "epoch": 0.08196721311475409,
+      "grad_norm": 6.214047431945801,
+      "learning_rate": 1.985932500536481e-05,
+      "loss": 1.1708,
+      "step": 400
+    },
+    {
+      "epoch": 0.08217213114754099,
+      "grad_norm": 7.050199508666992,
+      "learning_rate": 1.9858213392903786e-05,
+      "loss": 1.2478,
+      "step": 401
+    },
+    {
+      "epoch": 0.08237704918032787,
+      "grad_norm": 5.843206882476807,
+      "learning_rate": 1.9857097437091683e-05,
+      "loss": 1.038,
+      "step": 402
+    },
+    {
+      "epoch": 0.08258196721311475,
+      "grad_norm": 18.08635711669922,
+      "learning_rate": 1.9855977138420174e-05,
+      "loss": 0.9655,
+      "step": 403
+    },
+    {
+      "epoch": 0.08278688524590164,
+      "grad_norm": 14.273040771484375,
+      "learning_rate": 1.9854852497382843e-05,
+      "loss": 1.1011,
+      "step": 404
+    },
+    {
+      "epoch": 0.08299180327868852,
+      "grad_norm": 3.518399477005005,
+      "learning_rate": 1.9853723514475183e-05,
+      "loss": 1.1698,
+      "step": 405
+    },
+    {
+      "epoch": 0.08319672131147542,
+      "grad_norm": 3.1028213500976562,
+      "learning_rate": 1.9852590190194604e-05,
+      "loss": 1.2404,
+      "step": 406
+    },
+    {
+      "epoch": 0.0834016393442623,
+      "grad_norm": 5.72116231918335,
+      "learning_rate": 1.9851452525040433e-05,
+      "loss": 1.0001,
+      "step": 407
+    },
+    {
+      "epoch": 0.08360655737704918,
+      "grad_norm": 2.904630661010742,
+      "learning_rate": 1.98503105195139e-05,
+      "loss": 0.7577,
+      "step": 408
+    },
+    {
+      "epoch": 0.08381147540983606,
+      "grad_norm": 9.549497604370117,
+      "learning_rate": 1.9849164174118147e-05,
+      "loss": 1.2596,
+      "step": 409
+    },
+    {
+      "epoch": 0.08401639344262295,
+      "grad_norm": 5.525508880615234,
+      "learning_rate": 1.9848013489358245e-05,
+      "loss": 0.948,
+      "step": 410
+    },
+    {
+      "epoch": 0.08422131147540983,
+      "grad_norm": 8.438507080078125,
+      "learning_rate": 1.9846858465741157e-05,
+      "loss": 0.9516,
+      "step": 411
+    },
+    {
+      "epoch": 0.08442622950819673,
+      "grad_norm": 3.7555463314056396,
+      "learning_rate": 1.9845699103775766e-05,
+      "loss": 1.1547,
+      "step": 412
+    },
+    {
+      "epoch": 0.08463114754098361,
+      "grad_norm": 3.939037561416626,
+      "learning_rate": 1.984453540397287e-05,
+      "loss": 1.0833,
+      "step": 413
+    },
+    {
+      "epoch": 0.08483606557377049,
+      "grad_norm": 3.2579450607299805,
+      "learning_rate": 1.9843367366845172e-05,
+      "loss": 1.2377,
+      "step": 414
+    },
+    {
+      "epoch": 0.08504098360655737,
+      "grad_norm": 16.745431900024414,
+      "learning_rate": 1.984219499290729e-05,
+      "loss": 1.124,
+      "step": 415
+    },
+    {
+      "epoch": 0.08524590163934426,
+      "grad_norm": 7.746021747589111,
+      "learning_rate": 1.9841018282675743e-05,
+      "loss": 1.2139,
+      "step": 416
+    },
+    {
+      "epoch": 0.08545081967213115,
+      "grad_norm": 3.2424557209014893,
+      "learning_rate": 1.9839837236668975e-05,
+      "loss": 1.0922,
+      "step": 417
+    },
+    {
+      "epoch": 0.08565573770491804,
+      "grad_norm": 3.7237932682037354,
+      "learning_rate": 1.9838651855407335e-05,
+      "loss": 1.0508,
+      "step": 418
+    },
+    {
+      "epoch": 0.08586065573770492,
+      "grad_norm": 5.400592803955078,
+      "learning_rate": 1.983746213941308e-05,
+      "loss": 1.1481,
+      "step": 419
+    },
+    {
+      "epoch": 0.0860655737704918,
+      "grad_norm": 12.82390308380127,
+      "learning_rate": 1.983626808921037e-05,
+      "loss": 1.2062,
+      "step": 420
+    },
+    {
+      "epoch": 0.08627049180327868,
+      "grad_norm": 5.999417781829834,
+      "learning_rate": 1.9835069705325292e-05,
+      "loss": 1.2194,
+      "step": 421
+    },
+    {
+      "epoch": 0.08647540983606557,
+      "grad_norm": 3.446197032928467,
+      "learning_rate": 1.9833866988285825e-05,
+      "loss": 1.0371,
+      "step": 422
+    },
+    {
+      "epoch": 0.08668032786885246,
+      "grad_norm": 11.168344497680664,
+      "learning_rate": 1.9832659938621866e-05,
+      "loss": 1.1539,
+      "step": 423
+    },
+    {
+      "epoch": 0.08688524590163935,
+      "grad_norm": 3.7594494819641113,
+      "learning_rate": 1.9831448556865225e-05,
+      "loss": 1.0466,
+      "step": 424
+    },
+    {
+      "epoch": 0.08709016393442623,
+      "grad_norm": 8.873649597167969,
+      "learning_rate": 1.9830232843549612e-05,
+      "loss": 1.3276,
+      "step": 425
+    },
+    {
+      "epoch": 0.08729508196721311,
+      "grad_norm": 5.091553211212158,
+      "learning_rate": 1.982901279921064e-05,
+      "loss": 1.0249,
+      "step": 426
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 4.82570743560791,
+      "learning_rate": 1.9827788424385852e-05,
+      "loss": 1.0564,
+      "step": 427
+    },
+    {
+      "epoch": 0.08770491803278689,
+      "grad_norm": 5.105833053588867,
+      "learning_rate": 1.9826559719614678e-05,
+      "loss": 1.0904,
+      "step": 428
+    },
+    {
+      "epoch": 0.08790983606557377,
+      "grad_norm": 10.815126419067383,
+      "learning_rate": 1.982532668543846e-05,
+      "loss": 1.0885,
+      "step": 429
+    },
+    {
+      "epoch": 0.08811475409836066,
+      "grad_norm": 10.724872589111328,
+      "learning_rate": 1.9824089322400457e-05,
+      "loss": 1.4276,
+      "step": 430
+    },
+    {
+      "epoch": 0.08831967213114754,
+      "grad_norm": 6.311023235321045,
+      "learning_rate": 1.982284763104583e-05,
+      "loss": 1.1141,
+      "step": 431
+    },
+    {
+      "epoch": 0.08852459016393442,
+      "grad_norm": 3.94588303565979,
+      "learning_rate": 1.9821601611921635e-05,
+      "loss": 1.1054,
+      "step": 432
+    },
+    {
+      "epoch": 0.08872950819672132,
+      "grad_norm": 12.953496932983398,
+      "learning_rate": 1.982035126557686e-05,
+      "loss": 1.143,
+      "step": 433
+    },
+    {
+      "epoch": 0.0889344262295082,
+      "grad_norm": 3.5794312953948975,
+      "learning_rate": 1.9819096592562377e-05,
+      "loss": 1.0385,
+      "step": 434
+    },
+    {
+      "epoch": 0.08913934426229508,
+      "grad_norm": 5.2894606590271,
+      "learning_rate": 1.981783759343097e-05,
+      "loss": 1.3748,
+      "step": 435
+    },
+    {
+      "epoch": 0.08934426229508197,
+      "grad_norm": 4.756110668182373,
+      "learning_rate": 1.9816574268737335e-05,
+      "loss": 1.0401,
+      "step": 436
+    },
+    {
+      "epoch": 0.08954918032786885,
+      "grad_norm": 5.366695880889893,
+      "learning_rate": 1.9815306619038074e-05,
+      "loss": 0.9262,
+      "step": 437
+    },
+    {
+      "epoch": 0.08975409836065573,
+      "grad_norm": 7.6710357666015625,
+      "learning_rate": 1.9814034644891685e-05,
+      "loss": 1.0466,
+      "step": 438
+    },
+    {
+      "epoch": 0.08995901639344263,
+      "grad_norm": 11.06809139251709,
+      "learning_rate": 1.9812758346858575e-05,
+      "loss": 0.9149,
+      "step": 439
+    },
+    {
+      "epoch": 0.09016393442622951,
+      "grad_norm": 6.2112860679626465,
+      "learning_rate": 1.9811477725501062e-05,
+      "loss": 1.3283,
+      "step": 440
+    },
+    {
+      "epoch": 0.0903688524590164,
+      "grad_norm": 4.560141086578369,
+      "learning_rate": 1.9810192781383368e-05,
+      "loss": 1.0309,
+      "step": 441
+    },
+    {
+      "epoch": 0.09057377049180328,
+      "grad_norm": 6.564510822296143,
+      "learning_rate": 1.980890351507161e-05,
+      "loss": 1.0267,
+      "step": 442
+    },
+    {
+      "epoch": 0.09077868852459016,
+      "grad_norm": 5.00251579284668,
+      "learning_rate": 1.9807609927133815e-05,
+      "loss": 1.1372,
+      "step": 443
+    },
+    {
+      "epoch": 0.09098360655737706,
+      "grad_norm": 5.195645809173584,
+      "learning_rate": 1.980631201813992e-05,
+      "loss": 1.1602,
+      "step": 444
+    },
+    {
+      "epoch": 0.09118852459016394,
+      "grad_norm": 7.138591766357422,
+      "learning_rate": 1.9805009788661754e-05,
+      "loss": 1.0639,
+      "step": 445
+    },
+    {
+      "epoch": 0.09139344262295082,
+      "grad_norm": 2.8527333736419678,
+      "learning_rate": 1.980370323927306e-05,
+      "loss": 0.9744,
+      "step": 446
+    },
+    {
+      "epoch": 0.0915983606557377,
+      "grad_norm": 3.473731756210327,
+      "learning_rate": 1.980239237054948e-05,
+      "loss": 1.0944,
+      "step": 447
+    },
+    {
+      "epoch": 0.09180327868852459,
+      "grad_norm": 4.834840774536133,
+      "learning_rate": 1.9801077183068563e-05,
+      "loss": 1.2849,
+      "step": 448
+    },
+    {
+      "epoch": 0.09200819672131147,
+      "grad_norm": 3.979938507080078,
+      "learning_rate": 1.9799757677409747e-05,
+      "loss": 1.1443,
+      "step": 449
+    },
+    {
+      "epoch": 0.09221311475409837,
+      "grad_norm": 4.512206554412842,
+      "learning_rate": 1.9798433854154387e-05,
+      "loss": 1.1485,
+      "step": 450
+    },
+    {
+      "epoch": 0.09241803278688525,
+      "grad_norm": 3.106778621673584,
+      "learning_rate": 1.979710571388574e-05,
+      "loss": 0.9873,
+      "step": 451
+    },
+    {
+      "epoch": 0.09262295081967213,
+      "grad_norm": 3.794093608856201,
+      "learning_rate": 1.9795773257188955e-05,
+      "loss": 0.7376,
+      "step": 452
+    },
+    {
+      "epoch": 0.09282786885245901,
+      "grad_norm": 3.374683141708374,
+      "learning_rate": 1.979443648465109e-05,
+      "loss": 1.1689,
+      "step": 453
+    },
+    {
+      "epoch": 0.0930327868852459,
+      "grad_norm": 7.916249752044678,
+      "learning_rate": 1.9793095396861104e-05,
+      "loss": 1.0732,
+      "step": 454
+    },
+    {
+      "epoch": 0.0932377049180328,
+      "grad_norm": 8.963521957397461,
+      "learning_rate": 1.9791749994409856e-05,
+      "loss": 1.2149,
+      "step": 455
+    },
+    {
+      "epoch": 0.09344262295081968,
+      "grad_norm": 4.018504619598389,
+      "learning_rate": 1.9790400277890107e-05,
+      "loss": 1.0471,
+      "step": 456
+    },
+    {
+      "epoch": 0.09364754098360656,
+      "grad_norm": 14.759957313537598,
+      "learning_rate": 1.9789046247896514e-05,
+      "loss": 1.4734,
+      "step": 457
+    },
+    {
+      "epoch": 0.09385245901639344,
+      "grad_norm": 9.469250679016113,
+      "learning_rate": 1.9787687905025642e-05,
+      "loss": 1.1113,
+      "step": 458
+    },
+    {
+      "epoch": 0.09405737704918032,
+      "grad_norm": 7.0343427658081055,
+      "learning_rate": 1.9786325249875945e-05,
+      "loss": 1.1035,
+      "step": 459
+    },
+    {
+      "epoch": 0.0942622950819672,
+      "grad_norm": 3.955578088760376,
+      "learning_rate": 1.97849582830478e-05,
+      "loss": 1.088,
+      "step": 460
+    },
+    {
+      "epoch": 0.0944672131147541,
+      "grad_norm": 3.693833351135254,
+      "learning_rate": 1.978358700514345e-05,
+      "loss": 1.1247,
+      "step": 461
+    },
+    {
+      "epoch": 0.09467213114754099,
+      "grad_norm": 11.533448219299316,
+      "learning_rate": 1.9782211416767066e-05,
+      "loss": 1.5089,
+      "step": 462
+    },
+    {
+      "epoch": 0.09487704918032787,
+      "grad_norm": 7.726495742797852,
+      "learning_rate": 1.9780831518524704e-05,
+      "loss": 1.0064,
+      "step": 463
+    },
+    {
+      "epoch": 0.09508196721311475,
+      "grad_norm": 5.789141654968262,
+      "learning_rate": 1.9779447311024326e-05,
+      "loss": 1.4647,
+      "step": 464
+    },
+    {
+      "epoch": 0.09528688524590163,
+      "grad_norm": 7.863314151763916,
+      "learning_rate": 1.977805879487578e-05,
+      "loss": 1.225,
+      "step": 465
+    },
+    {
+      "epoch": 0.09549180327868853,
+      "grad_norm": 3.453599691390991,
+      "learning_rate": 1.9776665970690834e-05,
+      "loss": 1.0537,
+      "step": 466
+    },
+    {
+      "epoch": 0.09569672131147541,
+      "grad_norm": 6.030725479125977,
+      "learning_rate": 1.977526883908313e-05,
+      "loss": 1.2394,
+      "step": 467
+    },
+    {
+      "epoch": 0.0959016393442623,
+      "grad_norm": 14.207708358764648,
+      "learning_rate": 1.9773867400668224e-05,
+      "loss": 1.0313,
+      "step": 468
+    },
+    {
+      "epoch": 0.09610655737704918,
+      "grad_norm": 5.524641036987305,
+      "learning_rate": 1.9772461656063563e-05,
+      "loss": 1.1591,
+      "step": 469
+    },
+    {
+      "epoch": 0.09631147540983606,
+      "grad_norm": 5.1551103591918945,
+      "learning_rate": 1.97710516058885e-05,
+      "loss": 0.756,
+      "step": 470
+    },
+    {
+      "epoch": 0.09651639344262294,
+      "grad_norm": 3.6393790245056152,
+      "learning_rate": 1.9769637250764262e-05,
+      "loss": 1.0599,
+      "step": 471
+    },
+    {
+      "epoch": 0.09672131147540984,
+      "grad_norm": 4.5832672119140625,
+      "learning_rate": 1.9768218591314003e-05,
+      "loss": 1.2504,
+      "step": 472
+    },
+    {
+      "epoch": 0.09692622950819672,
+      "grad_norm": 3.217679262161255,
+      "learning_rate": 1.9766795628162752e-05,
+      "loss": 1.1328,
+      "step": 473
+    },
+    {
+      "epoch": 0.0971311475409836,
+      "grad_norm": 7.688928604125977,
+      "learning_rate": 1.9765368361937444e-05,
+      "loss": 1.3168,
+      "step": 474
+    },
+    {
+      "epoch": 0.09733606557377049,
+      "grad_norm": 6.236804008483887,
+      "learning_rate": 1.9763936793266905e-05,
+      "loss": 1.1685,
+      "step": 475
+    },
+    {
+      "epoch": 0.09754098360655737,
+      "grad_norm": 3.570645809173584,
+      "learning_rate": 1.9762500922781856e-05,
+      "loss": 1.0228,
+      "step": 476
+    },
+    {
+      "epoch": 0.09774590163934427,
+      "grad_norm": 5.148707389831543,
+      "learning_rate": 1.976106075111492e-05,
+      "loss": 1.0513,
+      "step": 477
+    },
+    {
+      "epoch": 0.09795081967213115,
+      "grad_norm": 5.54697847366333,
+      "learning_rate": 1.9759616278900607e-05,
+      "loss": 1.0121,
+      "step": 478
+    },
+    {
+      "epoch": 0.09815573770491803,
+      "grad_norm": 4.513408184051514,
+      "learning_rate": 1.9758167506775326e-05,
+      "loss": 1.077,
+      "step": 479
+    },
+    {
+      "epoch": 0.09836065573770492,
+      "grad_norm": 6.343353271484375,
+      "learning_rate": 1.9756714435377383e-05,
+      "loss": 0.9622,
+      "step": 480
+    },
+    {
+      "epoch": 0.0985655737704918,
+      "grad_norm": 6.969505786895752,
+      "learning_rate": 1.9755257065346975e-05,
+      "loss": 1.0765,
+      "step": 481
+    },
+    {
+      "epoch": 0.09877049180327868,
+      "grad_norm": 4.847391605377197,
+      "learning_rate": 1.975379539732619e-05,
+      "loss": 1.0316,
+      "step": 482
+    },
+    {
+      "epoch": 0.09897540983606558,
+      "grad_norm": 5.724539279937744,
+      "learning_rate": 1.9752329431959014e-05,
+      "loss": 1.3514,
+      "step": 483
+    },
+    {
+      "epoch": 0.09918032786885246,
+      "grad_norm": 5.527928352355957,
+      "learning_rate": 1.9750859169891325e-05,
+      "loss": 1.1363,
+      "step": 484
+    },
+    {
+      "epoch": 0.09938524590163934,
+      "grad_norm": 7.495690822601318,
+      "learning_rate": 1.974938461177089e-05,
+      "loss": 1.0607,
+      "step": 485
+    },
+    {
+      "epoch": 0.09959016393442623,
+      "grad_norm": 4.207889080047607,
+      "learning_rate": 1.9747905758247375e-05,
+      "loss": 1.2226,
+      "step": 486
+    },
+    {
+      "epoch": 0.09979508196721311,
+      "grad_norm": 10.529947280883789,
+      "learning_rate": 1.9746422609972344e-05,
+      "loss": 1.1581,
+      "step": 487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7691893577575684,
+      "learning_rate": 1.974493516759923e-05,
+      "loss": 1.038,
+      "step": 488
+    },
+    {
+      "epoch": 0.10020491803278689,
+      "grad_norm": 5.862758636474609,
+      "learning_rate": 1.9743443431783392e-05,
+      "loss": 1.3264,
+      "step": 489
+    },
+    {
+      "epoch": 0.10040983606557377,
+      "grad_norm": 6.412524223327637,
+      "learning_rate": 1.974194740318205e-05,
+      "loss": 0.9888,
+      "step": 490
+    },
+    {
+      "epoch": 0.10061475409836065,
+      "grad_norm": 11.411065101623535,
+      "learning_rate": 1.9740447082454324e-05,
+      "loss": 1.3166,
+      "step": 491
+    },
+    {
+      "epoch": 0.10081967213114754,
+      "grad_norm": 4.286177158355713,
+      "learning_rate": 1.9738942470261237e-05,
+      "loss": 1.2019,
+      "step": 492
+    },
+    {
+      "epoch": 0.10102459016393443,
+      "grad_norm": 5.590954303741455,
+      "learning_rate": 1.973743356726569e-05,
+      "loss": 1.3095,
+      "step": 493
+    },
+    {
+      "epoch": 0.10122950819672132,
+      "grad_norm": 9.5623779296875,
+      "learning_rate": 1.9735920374132477e-05,
+      "loss": 1.5728,
+      "step": 494
+    },
+    {
+      "epoch": 0.1014344262295082,
+      "grad_norm": 4.693945407867432,
+      "learning_rate": 1.9734402891528287e-05,
+      "loss": 1.1038,
+      "step": 495
+    },
+    {
+      "epoch": 0.10163934426229508,
+      "grad_norm": 4.619269371032715,
+      "learning_rate": 1.9732881120121693e-05,
+      "loss": 0.955,
+      "step": 496
+    },
+    {
+      "epoch": 0.10184426229508196,
+      "grad_norm": 4.818992614746094,
+      "learning_rate": 1.9731355060583163e-05,
+      "loss": 0.8299,
+      "step": 497
+    },
+    {
+      "epoch": 0.10204918032786885,
+      "grad_norm": 3.981853485107422,
+      "learning_rate": 1.972982471358505e-05,
+      "loss": 0.9435,
+      "step": 498
+    },
+    {
+      "epoch": 0.10225409836065574,
+      "grad_norm": 4.845431804656982,
+      "learning_rate": 1.9728290079801594e-05,
+      "loss": 1.4056,
+      "step": 499
+    },
+    {
+      "epoch": 0.10245901639344263,
+      "grad_norm": 6.609842300415039,
+      "learning_rate": 1.9726751159908933e-05,
+      "loss": 0.897,
+      "step": 500
+    },
+    {
+      "epoch": 0.10266393442622951,
+      "grad_norm": 6.319755554199219,
+      "learning_rate": 1.972520795458508e-05,
+      "loss": 1.139,
+      "step": 501
+    },
+    {
+      "epoch": 0.10286885245901639,
+      "grad_norm": 5.723700523376465,
+      "learning_rate": 1.9723660464509947e-05,
+      "loss": 1.0787,
+      "step": 502
+    },
+    {
+      "epoch": 0.10307377049180327,
+      "grad_norm": 4.0129618644714355,
+      "learning_rate": 1.9722108690365334e-05,
+      "loss": 1.2039,
+      "step": 503
+    },
+    {
+      "epoch": 0.10327868852459017,
+      "grad_norm": 5.242337226867676,
+      "learning_rate": 1.972055263283492e-05,
+      "loss": 1.2221,
+      "step": 504
+    },
+    {
+      "epoch": 0.10348360655737705,
+      "grad_norm": 8.100120544433594,
+      "learning_rate": 1.971899229260428e-05,
+      "loss": 1.1796,
+      "step": 505
+    },
+    {
+      "epoch": 0.10368852459016394,
+      "grad_norm": 3.9425466060638428,
+      "learning_rate": 1.9717427670360867e-05,
+      "loss": 1.1289,
+      "step": 506
+    },
+    {
+      "epoch": 0.10389344262295082,
+      "grad_norm": 5.129619121551514,
+      "learning_rate": 1.9715858766794028e-05,
+      "loss": 1.2209,
+      "step": 507
+    },
+    {
+      "epoch": 0.1040983606557377,
+      "grad_norm": 7.54534387588501,
+      "learning_rate": 1.971428558259499e-05,
+      "loss": 1.5715,
+      "step": 508
+    },
+    {
+      "epoch": 0.10430327868852458,
+      "grad_norm": 5.647359848022461,
+      "learning_rate": 1.9712708118456877e-05,
+      "loss": 1.1539,
+      "step": 509
+    },
+    {
+      "epoch": 0.10450819672131148,
+      "grad_norm": 4.6107378005981445,
+      "learning_rate": 1.9711126375074686e-05,
+      "loss": 1.0716,
+      "step": 510
+    },
+    {
+      "epoch": 0.10471311475409836,
+      "grad_norm": 4.168269157409668,
+      "learning_rate": 1.9709540353145306e-05,
+      "loss": 1.0314,
+      "step": 511
+    },
+    {
+      "epoch": 0.10491803278688525,
+      "grad_norm": 3.4925146102905273,
+      "learning_rate": 1.9707950053367506e-05,
+      "loss": 0.9379,
+      "step": 512
+    },
+    {
+      "epoch": 0.10512295081967213,
+      "grad_norm": 5.265373706817627,
+      "learning_rate": 1.9706355476441947e-05,
+      "loss": 1.087,
+      "step": 513
+    },
+    {
+      "epoch": 0.10532786885245901,
+      "grad_norm": 5.916897773742676,
+      "learning_rate": 1.9704756623071172e-05,
+      "loss": 1.2256,
+      "step": 514
+    },
+    {
+      "epoch": 0.10553278688524591,
+      "grad_norm": 5.596438407897949,
+      "learning_rate": 1.9703153493959606e-05,
+      "loss": 1.185,
+      "step": 515
+    },
+    {
+      "epoch": 0.10573770491803279,
+      "grad_norm": 6.046541213989258,
+      "learning_rate": 1.9701546089813556e-05,
+      "loss": 1.229,
+      "step": 516
+    },
+    {
+      "epoch": 0.10594262295081967,
+      "grad_norm": 6.758883953094482,
+      "learning_rate": 1.969993441134122e-05,
+      "loss": 0.9658,
+      "step": 517
+    },
+    {
+      "epoch": 0.10614754098360656,
+      "grad_norm": 7.80905294418335,
+      "learning_rate": 1.969831845925267e-05,
+      "loss": 1.3384,
+      "step": 518
+    },
+    {
+      "epoch": 0.10635245901639344,
+      "grad_norm": 6.035611152648926,
+      "learning_rate": 1.969669823425987e-05,
+      "loss": 1.2211,
+      "step": 519
+    },
+    {
+      "epoch": 0.10655737704918032,
+      "grad_norm": 11.743453979492188,
+      "learning_rate": 1.9695073737076653e-05,
+      "loss": 1.0609,
+      "step": 520
+    },
+    {
+      "epoch": 0.10676229508196722,
+      "grad_norm": 9.643823623657227,
+      "learning_rate": 1.9693444968418753e-05,
+      "loss": 1.1329,
+      "step": 521
+    },
+    {
+      "epoch": 0.1069672131147541,
+      "grad_norm": 6.302698135375977,
+      "learning_rate": 1.9691811929003775e-05,
+      "loss": 0.8874,
+      "step": 522
+    },
+    {
+      "epoch": 0.10717213114754098,
+      "grad_norm": 7.032561302185059,
+      "learning_rate": 1.9690174619551203e-05,
+      "loss": 1.0764,
+      "step": 523
+    },
+    {
+      "epoch": 0.10737704918032787,
+      "grad_norm": 5.467191219329834,
+      "learning_rate": 1.9688533040782404e-05,
+      "loss": 1.1309,
+      "step": 524
+    },
+    {
+      "epoch": 0.10758196721311475,
+      "grad_norm": 3.4174082279205322,
+      "learning_rate": 1.9686887193420635e-05,
+      "loss": 1.1209,
+      "step": 525
+    },
+    {
+      "epoch": 0.10778688524590165,
+      "grad_norm": 9.874616622924805,
+      "learning_rate": 1.968523707819102e-05,
+      "loss": 1.218,
+      "step": 526
+    },
+    {
+      "epoch": 0.10799180327868853,
+      "grad_norm": 5.652812480926514,
+      "learning_rate": 1.9683582695820575e-05,
+      "loss": 1.1327,
+      "step": 527
+    },
+    {
+      "epoch": 0.10819672131147541,
+      "grad_norm": 6.237393379211426,
+      "learning_rate": 1.9681924047038187e-05,
+      "loss": 0.8905,
+      "step": 528
+    },
+    {
+      "epoch": 0.1084016393442623,
+      "grad_norm": 6.403660774230957,
+      "learning_rate": 1.968026113257463e-05,
+      "loss": 1.0695,
+      "step": 529
+    },
+    {
+      "epoch": 0.10860655737704918,
+      "grad_norm": 7.693816661834717,
+      "learning_rate": 1.9678593953162556e-05,
+      "loss": 1.2781,
+      "step": 530
+    },
+    {
+      "epoch": 0.10881147540983606,
+      "grad_norm": 7.445765018463135,
+      "learning_rate": 1.9676922509536486e-05,
+      "loss": 0.9926,
+      "step": 531
+    },
+    {
+      "epoch": 0.10901639344262296,
+      "grad_norm": 6.620217323303223,
+      "learning_rate": 1.9675246802432834e-05,
+      "loss": 1.109,
+      "step": 532
+    },
+    {
+      "epoch": 0.10922131147540984,
+      "grad_norm": 3.8569414615631104,
+      "learning_rate": 1.9673566832589888e-05,
+      "loss": 1.0807,
+      "step": 533
+    },
+    {
+      "epoch": 0.10942622950819672,
+      "grad_norm": 3.502535343170166,
+      "learning_rate": 1.9671882600747806e-05,
+      "loss": 1.0909,
+      "step": 534
+    },
+    {
+      "epoch": 0.1096311475409836,
+      "grad_norm": 6.048972129821777,
+      "learning_rate": 1.967019410764864e-05,
+      "loss": 1.0757,
+      "step": 535
+    },
+    {
+      "epoch": 0.10983606557377049,
+      "grad_norm": 7.330895900726318,
+      "learning_rate": 1.9668501354036298e-05,
+      "loss": 1.0929,
+      "step": 536
+    },
+    {
+      "epoch": 0.11004098360655738,
+      "grad_norm": 3.961709499359131,
+      "learning_rate": 1.9666804340656584e-05,
+      "loss": 1.0843,
+      "step": 537
+    },
+    {
+      "epoch": 0.11024590163934427,
+      "grad_norm": 6.7322845458984375,
+      "learning_rate": 1.9665103068257168e-05,
+      "loss": 0.962,
+      "step": 538
+    },
+    {
+      "epoch": 0.11045081967213115,
+      "grad_norm": 5.309615612030029,
+      "learning_rate": 1.9663397537587604e-05,
+      "loss": 1.4539,
+      "step": 539
+    },
+    {
+      "epoch": 0.11065573770491803,
+      "grad_norm": 4.480984210968018,
+      "learning_rate": 1.9661687749399315e-05,
+      "loss": 1.2021,
+      "step": 540
+    },
+    {
+      "epoch": 0.11086065573770491,
+      "grad_norm": 3.9622044563293457,
+      "learning_rate": 1.9659973704445604e-05,
+      "loss": 1.2742,
+      "step": 541
+    },
+    {
+      "epoch": 0.11106557377049181,
+      "grad_norm": 3.124962568283081,
+      "learning_rate": 1.9658255403481645e-05,
+      "loss": 0.9213,
+      "step": 542
+    },
+    {
+      "epoch": 0.11127049180327869,
+      "grad_norm": 6.439422130584717,
+      "learning_rate": 1.9656532847264497e-05,
+      "loss": 1.1729,
+      "step": 543
+    },
+    {
+      "epoch": 0.11147540983606558,
+      "grad_norm": 3.874537944793701,
+      "learning_rate": 1.965480603655308e-05,
+      "loss": 1.1221,
+      "step": 544
+    },
+    {
+      "epoch": 0.11168032786885246,
+      "grad_norm": 6.43537712097168,
+      "learning_rate": 1.9653074972108203e-05,
+      "loss": 1.2158,
+      "step": 545
+    },
+    {
+      "epoch": 0.11188524590163934,
+      "grad_norm": 5.951541423797607,
+      "learning_rate": 1.9651339654692537e-05,
+      "loss": 1.3796,
+      "step": 546
+    },
+    {
+      "epoch": 0.11209016393442622,
+      "grad_norm": 5.784824371337891,
+      "learning_rate": 1.964960008507063e-05,
+      "loss": 1.1755,
+      "step": 547
+    },
+    {
+      "epoch": 0.11229508196721312,
+      "grad_norm": 5.9174041748046875,
+      "learning_rate": 1.9647856264008913e-05,
+      "loss": 0.9644,
+      "step": 548
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 7.800127029418945,
+      "learning_rate": 1.9646108192275673e-05,
+      "loss": 1.2938,
+      "step": 549
+    },
+    {
+      "epoch": 0.11270491803278689,
+      "grad_norm": 5.84089994430542,
+      "learning_rate": 1.964435587064108e-05,
+      "loss": 1.2522,
+      "step": 550
+    },
+    {
+      "epoch": 0.11290983606557377,
+      "grad_norm": 4.274975776672363,
+      "learning_rate": 1.9642599299877183e-05,
+      "loss": 1.0392,
+      "step": 551
+    },
+    {
+      "epoch": 0.11311475409836065,
+      "grad_norm": 4.058582305908203,
+      "learning_rate": 1.9640838480757887e-05,
+      "loss": 1.0835,
+      "step": 552
+    },
+    {
+      "epoch": 0.11331967213114755,
+      "grad_norm": 5.6682939529418945,
+      "learning_rate": 1.963907341405898e-05,
+      "loss": 0.9911,
+      "step": 553
+    },
+    {
+      "epoch": 0.11352459016393443,
+      "grad_norm": 4.252900123596191,
+      "learning_rate": 1.9637304100558122e-05,
+      "loss": 1.0559,
+      "step": 554
+    },
+    {
+      "epoch": 0.11372950819672131,
+      "grad_norm": 8.527545928955078,
+      "learning_rate": 1.9635530541034835e-05,
+      "loss": 1.0736,
+      "step": 555
+    },
+    {
+      "epoch": 0.1139344262295082,
+      "grad_norm": 3.619999647140503,
+      "learning_rate": 1.963375273627052e-05,
+      "loss": 0.9028,
+      "step": 556
+    },
+    {
+      "epoch": 0.11413934426229508,
+      "grad_norm": 5.17339563369751,
+      "learning_rate": 1.9631970687048446e-05,
+      "loss": 1.1709,
+      "step": 557
+    },
+    {
+      "epoch": 0.11434426229508196,
+      "grad_norm": 6.147169589996338,
+      "learning_rate": 1.9630184394153755e-05,
+      "loss": 1.0099,
+      "step": 558
+    },
+    {
+      "epoch": 0.11454918032786886,
+      "grad_norm": 8.03926944732666,
+      "learning_rate": 1.962839385837345e-05,
+      "loss": 1.1498,
+      "step": 559
+    },
+    {
+      "epoch": 0.11475409836065574,
+      "grad_norm": 5.418529033660889,
+      "learning_rate": 1.9626599080496407e-05,
+      "loss": 1.1062,
+      "step": 560
+    },
+    {
+      "epoch": 0.11495901639344262,
+      "grad_norm": 6.699753761291504,
+      "learning_rate": 1.962480006131338e-05,
+      "loss": 0.9849,
+      "step": 561
+    },
+    {
+      "epoch": 0.1151639344262295,
+      "grad_norm": 3.6027157306671143,
+      "learning_rate": 1.962299680161698e-05,
+      "loss": 1.1867,
+      "step": 562
+    },
+    {
+      "epoch": 0.11536885245901639,
+      "grad_norm": 4.8496928215026855,
+      "learning_rate": 1.9621189302201692e-05,
+      "loss": 1.0063,
+      "step": 563
+    },
+    {
+      "epoch": 0.11557377049180328,
+      "grad_norm": 6.512297630310059,
+      "learning_rate": 1.9619377563863867e-05,
+      "loss": 1.1573,
+      "step": 564
+    },
+    {
+      "epoch": 0.11577868852459017,
+      "grad_norm": 5.721385478973389,
+      "learning_rate": 1.9617561587401726e-05,
+      "loss": 0.6965,
+      "step": 565
+    },
+    {
+      "epoch": 0.11598360655737705,
+      "grad_norm": 6.886279582977295,
+      "learning_rate": 1.9615741373615352e-05,
+      "loss": 1.0906,
+      "step": 566
+    },
+    {
+      "epoch": 0.11618852459016393,
+      "grad_norm": 12.600558280944824,
+      "learning_rate": 1.96139169233067e-05,
+      "loss": 1.0417,
+      "step": 567
+    },
+    {
+      "epoch": 0.11639344262295082,
+      "grad_norm": 4.6830153465271,
+      "learning_rate": 1.9612088237279596e-05,
+      "loss": 1.274,
+      "step": 568
+    },
+    {
+      "epoch": 0.1165983606557377,
+      "grad_norm": 5.533491134643555,
+      "learning_rate": 1.9610255316339712e-05,
+      "loss": 1.2505,
+      "step": 569
+    },
+    {
+      "epoch": 0.1168032786885246,
+      "grad_norm": 4.506268501281738,
+      "learning_rate": 1.9608418161294612e-05,
+      "loss": 1.1488,
+      "step": 570
+    },
+    {
+      "epoch": 0.11700819672131148,
+      "grad_norm": 6.280215263366699,
+      "learning_rate": 1.960657677295371e-05,
+      "loss": 1.0911,
+      "step": 571
+    },
+    {
+      "epoch": 0.11721311475409836,
+      "grad_norm": 6.091683387756348,
+      "learning_rate": 1.9604731152128287e-05,
+      "loss": 0.88,
+      "step": 572
+    },
+    {
+      "epoch": 0.11741803278688524,
+      "grad_norm": 7.105499267578125,
+      "learning_rate": 1.960288129963149e-05,
+      "loss": 1.2173,
+      "step": 573
+    },
+    {
+      "epoch": 0.11762295081967213,
+      "grad_norm": 7.366506099700928,
+      "learning_rate": 1.9601027216278338e-05,
+      "loss": 1.3305,
+      "step": 574
+    },
+    {
+      "epoch": 0.11782786885245902,
+      "grad_norm": 4.746676445007324,
+      "learning_rate": 1.9599168902885696e-05,
+      "loss": 1.2502,
+      "step": 575
+    },
+    {
+      "epoch": 0.1180327868852459,
+      "grad_norm": 3.9490833282470703,
+      "learning_rate": 1.959730636027231e-05,
+      "loss": 1.0232,
+      "step": 576
+    },
+    {
+      "epoch": 0.11823770491803279,
+      "grad_norm": 3.159025192260742,
+      "learning_rate": 1.959543958925878e-05,
+      "loss": 1.1407,
+      "step": 577
+    },
+    {
+      "epoch": 0.11844262295081967,
+      "grad_norm": 4.321066379547119,
+      "learning_rate": 1.959356859066757e-05,
+      "loss": 0.9004,
+      "step": 578
+    },
+    {
+      "epoch": 0.11864754098360655,
+      "grad_norm": 11.040045738220215,
+      "learning_rate": 1.9591693365323015e-05,
+      "loss": 1.4272,
+      "step": 579
+    },
+    {
+      "epoch": 0.11885245901639344,
+      "grad_norm": 6.63756799697876,
+      "learning_rate": 1.9589813914051297e-05,
+      "loss": 1.3064,
+      "step": 580
+    },
+    {
+      "epoch": 0.11905737704918033,
+      "grad_norm": 4.820218563079834,
+      "learning_rate": 1.9587930237680475e-05,
+      "loss": 0.8389,
+      "step": 581
+    },
+    {
+      "epoch": 0.11926229508196722,
+      "grad_norm": 3.8197498321533203,
+      "learning_rate": 1.9586042337040464e-05,
+      "loss": 0.8818,
+      "step": 582
+    },
+    {
+      "epoch": 0.1194672131147541,
+      "grad_norm": 8.314474105834961,
+      "learning_rate": 1.958415021296303e-05,
+      "loss": 1.0655,
+      "step": 583
+    },
+    {
+      "epoch": 0.11967213114754098,
+      "grad_norm": 9.790090560913086,
+      "learning_rate": 1.9582253866281817e-05,
+      "loss": 1.2461,
+      "step": 584
+    },
+    {
+      "epoch": 0.11987704918032786,
+      "grad_norm": 4.669817924499512,
+      "learning_rate": 1.9580353297832317e-05,
+      "loss": 1.1212,
+      "step": 585
+    },
+    {
+      "epoch": 0.12008196721311476,
+      "grad_norm": 7.805581569671631,
+      "learning_rate": 1.957844850845189e-05,
+      "loss": 1.1728,
+      "step": 586
+    },
+    {
+      "epoch": 0.12028688524590164,
+      "grad_norm": 5.179210186004639,
+      "learning_rate": 1.9576539498979745e-05,
+      "loss": 1.0111,
+      "step": 587
+    },
+    {
+      "epoch": 0.12049180327868853,
+      "grad_norm": 3.0293049812316895,
+      "learning_rate": 1.9574626270256963e-05,
+      "loss": 0.9105,
+      "step": 588
+    },
+    {
+      "epoch": 0.12069672131147541,
+      "grad_norm": 7.239634037017822,
+      "learning_rate": 1.9572708823126477e-05,
+      "loss": 1.0846,
+      "step": 589
+    },
+    {
+      "epoch": 0.12090163934426229,
+      "grad_norm": 4.604331970214844,
+      "learning_rate": 1.957078715843308e-05,
+      "loss": 1.1328,
+      "step": 590
+    },
+    {
+      "epoch": 0.12110655737704919,
+      "grad_norm": 4.732018947601318,
+      "learning_rate": 1.956886127702342e-05,
+      "loss": 1.1961,
+      "step": 591
+    },
+    {
+      "epoch": 0.12131147540983607,
+      "grad_norm": 5.725955963134766,
+      "learning_rate": 1.956693117974601e-05,
+      "loss": 0.8697,
+      "step": 592
+    },
+    {
+      "epoch": 0.12151639344262295,
+      "grad_norm": 10.362656593322754,
+      "learning_rate": 1.9564996867451214e-05,
+      "loss": 1.0119,
+      "step": 593
+    },
+    {
+      "epoch": 0.12172131147540984,
+      "grad_norm": 5.288120269775391,
+      "learning_rate": 1.9563058340991254e-05,
+      "loss": 1.0287,
+      "step": 594
+    },
+    {
+      "epoch": 0.12192622950819672,
+      "grad_norm": 4.008474826812744,
+      "learning_rate": 1.956111560122021e-05,
+      "loss": 1.1528,
+      "step": 595
+    },
+    {
+      "epoch": 0.1221311475409836,
+      "grad_norm": 5.368836402893066,
+      "learning_rate": 1.955916864899402e-05,
+      "loss": 0.9499,
+      "step": 596
+    },
+    {
+      "epoch": 0.1223360655737705,
+      "grad_norm": 4.160040855407715,
+      "learning_rate": 1.9557217485170473e-05,
+      "loss": 1.2802,
+      "step": 597
+    },
+    {
+      "epoch": 0.12254098360655738,
+      "grad_norm": 5.125066757202148,
+      "learning_rate": 1.955526211060922e-05,
+      "loss": 0.9663,
+      "step": 598
+    },
+    {
+      "epoch": 0.12274590163934426,
+      "grad_norm": 2.93587327003479,
+      "learning_rate": 1.955330252617176e-05,
+      "loss": 0.9865,
+      "step": 599
+    },
+    {
+      "epoch": 0.12295081967213115,
+      "grad_norm": 4.461883068084717,
+      "learning_rate": 1.9551338732721458e-05,
+      "loss": 0.9501,
+      "step": 600
+    },
+    {
+      "epoch": 0.12315573770491803,
+      "grad_norm": 6.433072566986084,
+      "learning_rate": 1.9549370731123516e-05,
+      "loss": 0.9412,
+      "step": 601
+    },
+    {
+      "epoch": 0.12336065573770492,
+      "grad_norm": 14.074542045593262,
+      "learning_rate": 1.9547398522245007e-05,
+      "loss": 1.1904,
+      "step": 602
+    },
+    {
+      "epoch": 0.12356557377049181,
+      "grad_norm": 3.641433000564575,
+      "learning_rate": 1.9545422106954848e-05,
+      "loss": 1.04,
+      "step": 603
+    },
+    {
+      "epoch": 0.12377049180327869,
+      "grad_norm": 4.920017242431641,
+      "learning_rate": 1.954344148612381e-05,
+      "loss": 1.0674,
+      "step": 604
+    },
+    {
+      "epoch": 0.12397540983606557,
+      "grad_norm": 3.3146302700042725,
+      "learning_rate": 1.9541456660624525e-05,
+      "loss": 1.0376,
+      "step": 605
+    },
+    {
+      "epoch": 0.12418032786885246,
+      "grad_norm": 4.544986724853516,
+      "learning_rate": 1.9539467631331465e-05,
+      "loss": 1.4075,
+      "step": 606
+    },
+    {
+      "epoch": 0.12438524590163934,
+      "grad_norm": 4.605386257171631,
+      "learning_rate": 1.9537474399120965e-05,
+      "loss": 1.2512,
+      "step": 607
+    },
+    {
+      "epoch": 0.12459016393442623,
+      "grad_norm": 4.086893558502197,
+      "learning_rate": 1.95354769648712e-05,
+      "loss": 1.1966,
+      "step": 608
+    },
+    {
+      "epoch": 0.12479508196721312,
+      "grad_norm": 2.9726295471191406,
+      "learning_rate": 1.9533475329462213e-05,
+      "loss": 0.9302,
+      "step": 609
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 4.090008735656738,
+      "learning_rate": 1.9531469493775884e-05,
+      "loss": 0.8683,
+      "step": 610
+    },
+    {
+      "epoch": 0.12520491803278688,
+      "grad_norm": 6.6718926429748535,
+      "learning_rate": 1.952945945869595e-05,
+      "loss": 1.047,
+      "step": 611
+    },
+    {
+      "epoch": 0.12540983606557377,
+      "grad_norm": 5.645918369293213,
+      "learning_rate": 1.9527445225107993e-05,
+      "loss": 1.0374,
+      "step": 612
+    },
+    {
+      "epoch": 0.12561475409836065,
+      "grad_norm": 3.9847586154937744,
+      "learning_rate": 1.9525426793899447e-05,
+      "loss": 1.0976,
+      "step": 613
+    },
+    {
+      "epoch": 0.12581967213114753,
+      "grad_norm": 7.126397609710693,
+      "learning_rate": 1.95234041659596e-05,
+      "loss": 1.2713,
+      "step": 614
+    },
+    {
+      "epoch": 0.1260245901639344,
+      "grad_norm": 6.036407470703125,
+      "learning_rate": 1.9521377342179586e-05,
+      "loss": 1.2156,
+      "step": 615
+    },
+    {
+      "epoch": 0.12622950819672132,
+      "grad_norm": 5.816885471343994,
+      "learning_rate": 1.9519346323452383e-05,
+      "loss": 1.1698,
+      "step": 616
+    },
+    {
+      "epoch": 0.1264344262295082,
+      "grad_norm": 3.994100332260132,
+      "learning_rate": 1.9517311110672824e-05,
+      "loss": 1.2983,
+      "step": 617
+    },
+    {
+      "epoch": 0.1266393442622951,
+      "grad_norm": 4.188429355621338,
+      "learning_rate": 1.951527170473759e-05,
+      "loss": 0.9621,
+      "step": 618
+    },
+    {
+      "epoch": 0.12684426229508197,
+      "grad_norm": 5.648606300354004,
+      "learning_rate": 1.95132281065452e-05,
+      "loss": 1.584,
+      "step": 619
+    },
+    {
+      "epoch": 0.12704918032786885,
+      "grad_norm": 4.516109466552734,
+      "learning_rate": 1.951118031699603e-05,
+      "loss": 1.194,
+      "step": 620
+    },
+    {
+      "epoch": 0.12725409836065574,
+      "grad_norm": 3.120361804962158,
+      "learning_rate": 1.9509128336992294e-05,
+      "loss": 1.2997,
+      "step": 621
+    },
+    {
+      "epoch": 0.12745901639344262,
+      "grad_norm": 3.050447940826416,
+      "learning_rate": 1.950707216743807e-05,
+      "loss": 0.9475,
+      "step": 622
+    },
+    {
+      "epoch": 0.1276639344262295,
+      "grad_norm": 7.873384475708008,
+      "learning_rate": 1.9505011809239252e-05,
+      "loss": 1.115,
+      "step": 623
+    },
+    {
+      "epoch": 0.12786885245901639,
+      "grad_norm": 15.878609657287598,
+      "learning_rate": 1.9502947263303615e-05,
+      "loss": 0.9305,
+      "step": 624
+    },
+    {
+      "epoch": 0.12807377049180327,
+      "grad_norm": 12.19124698638916,
+      "learning_rate": 1.9500878530540748e-05,
+      "loss": 1.3524,
+      "step": 625
+    },
+    {
+      "epoch": 0.12827868852459015,
+      "grad_norm": 3.5837903022766113,
+      "learning_rate": 1.94988056118621e-05,
+      "loss": 0.8883,
+      "step": 626
+    },
+    {
+      "epoch": 0.12848360655737706,
+      "grad_norm": 3.578158140182495,
+      "learning_rate": 1.9496728508180963e-05,
+      "loss": 1.0984,
+      "step": 627
+    },
+    {
+      "epoch": 0.12868852459016394,
+      "grad_norm": 6.101957321166992,
+      "learning_rate": 1.9494647220412474e-05,
+      "loss": 1.0817,
+      "step": 628
+    },
+    {
+      "epoch": 0.12889344262295083,
+      "grad_norm": 4.479754447937012,
+      "learning_rate": 1.9492561749473604e-05,
+      "loss": 0.867,
+      "step": 629
+    },
+    {
+      "epoch": 0.1290983606557377,
+      "grad_norm": 7.559701919555664,
+      "learning_rate": 1.9490472096283183e-05,
+      "loss": 1.1914,
+      "step": 630
+    },
+    {
+      "epoch": 0.1293032786885246,
+      "grad_norm": 7.564691543579102,
+      "learning_rate": 1.9488378261761866e-05,
+      "loss": 1.3573,
+      "step": 631
+    },
+    {
+      "epoch": 0.12950819672131147,
+      "grad_norm": 9.614606857299805,
+      "learning_rate": 1.948628024683216e-05,
+      "loss": 0.9161,
+      "step": 632
+    },
+    {
+      "epoch": 0.12971311475409836,
+      "grad_norm": 5.205758571624756,
+      "learning_rate": 1.948417805241842e-05,
+      "loss": 1.3436,
+      "step": 633
+    },
+    {
+      "epoch": 0.12991803278688524,
+      "grad_norm": 14.779903411865234,
+      "learning_rate": 1.9482071679446823e-05,
+      "loss": 1.0218,
+      "step": 634
+    },
+    {
+      "epoch": 0.13012295081967212,
+      "grad_norm": 6.070825576782227,
+      "learning_rate": 1.9479961128845407e-05,
+      "loss": 0.8954,
+      "step": 635
+    },
+    {
+      "epoch": 0.130327868852459,
+      "grad_norm": 5.7479753494262695,
+      "learning_rate": 1.9477846401544044e-05,
+      "loss": 1.5296,
+      "step": 636
+    },
+    {
+      "epoch": 0.1305327868852459,
+      "grad_norm": 5.432805061340332,
+      "learning_rate": 1.9475727498474442e-05,
+      "loss": 1.0263,
+      "step": 637
+    },
+    {
+      "epoch": 0.1307377049180328,
+      "grad_norm": 5.643026828765869,
+      "learning_rate": 1.9473604420570148e-05,
+      "loss": 0.9657,
+      "step": 638
+    },
+    {
+      "epoch": 0.13094262295081968,
+      "grad_norm": 7.0423150062561035,
+      "learning_rate": 1.947147716876656e-05,
+      "loss": 1.3281,
+      "step": 639
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 5.388788223266602,
+      "learning_rate": 1.94693457440009e-05,
+      "loss": 1.5126,
+      "step": 640
+    },
+    {
+      "epoch": 0.13135245901639345,
+      "grad_norm": 3.805838108062744,
+      "learning_rate": 1.9467210147212235e-05,
+      "loss": 0.8937,
+      "step": 641
+    },
+    {
+      "epoch": 0.13155737704918033,
+      "grad_norm": 5.584044933319092,
+      "learning_rate": 1.9465070379341475e-05,
+      "loss": 1.1011,
+      "step": 642
+    },
+    {
+      "epoch": 0.1317622950819672,
+      "grad_norm": 3.7881829738616943,
+      "learning_rate": 1.946292644133136e-05,
+      "loss": 1.286,
+      "step": 643
+    },
+    {
+      "epoch": 0.1319672131147541,
+      "grad_norm": 4.664256572723389,
+      "learning_rate": 1.946077833412647e-05,
+      "loss": 1.1114,
+      "step": 644
+    },
+    {
+      "epoch": 0.13217213114754098,
+      "grad_norm": 4.145486831665039,
+      "learning_rate": 1.945862605867323e-05,
+      "loss": 1.0098,
+      "step": 645
+    },
+    {
+      "epoch": 0.13237704918032786,
+      "grad_norm": 5.787230014801025,
+      "learning_rate": 1.9456469615919876e-05,
+      "loss": 1.4001,
+      "step": 646
+    },
+    {
+      "epoch": 0.13258196721311474,
+      "grad_norm": 4.139947891235352,
+      "learning_rate": 1.9454309006816517e-05,
+      "loss": 1.2157,
+      "step": 647
+    },
+    {
+      "epoch": 0.13278688524590163,
+      "grad_norm": 11.448017120361328,
+      "learning_rate": 1.945214423231507e-05,
+      "loss": 1.1634,
+      "step": 648
+    },
+    {
+      "epoch": 0.13299180327868854,
+      "grad_norm": 4.783749580383301,
+      "learning_rate": 1.9449975293369296e-05,
+      "loss": 1.0837,
+      "step": 649
+    },
+    {
+      "epoch": 0.13319672131147542,
+      "grad_norm": 18.227947235107422,
+      "learning_rate": 1.9447802190934788e-05,
+      "loss": 1.2333,
+      "step": 650
+    },
+    {
+      "epoch": 0.1334016393442623,
+      "grad_norm": 3.372831106185913,
+      "learning_rate": 1.944562492596898e-05,
+      "loss": 0.9972,
+      "step": 651
+    },
+    {
+      "epoch": 0.13360655737704918,
+      "grad_norm": 5.098379135131836,
+      "learning_rate": 1.9443443499431133e-05,
+      "loss": 1.2551,
+      "step": 652
+    },
+    {
+      "epoch": 0.13381147540983607,
+      "grad_norm": 5.866696834564209,
+      "learning_rate": 1.9441257912282347e-05,
+      "loss": 1.2463,
+      "step": 653
+    },
+    {
+      "epoch": 0.13401639344262295,
+      "grad_norm": 7.520400524139404,
+      "learning_rate": 1.943906816548555e-05,
+      "loss": 1.3863,
+      "step": 654
+    },
+    {
+      "epoch": 0.13422131147540983,
+      "grad_norm": 5.84237003326416,
+      "learning_rate": 1.9436874260005504e-05,
+      "loss": 1.4107,
+      "step": 655
+    },
+    {
+      "epoch": 0.13442622950819672,
+      "grad_norm": 6.52760124206543,
+      "learning_rate": 1.9434676196808806e-05,
+      "loss": 1.2194,
+      "step": 656
+    },
+    {
+      "epoch": 0.1346311475409836,
+      "grad_norm": 5.527923583984375,
+      "learning_rate": 1.9432473976863884e-05,
+      "loss": 1.1536,
+      "step": 657
+    },
+    {
+      "epoch": 0.13483606557377048,
+      "grad_norm": 7.988602638244629,
+      "learning_rate": 1.9430267601140996e-05,
+      "loss": 0.8814,
+      "step": 658
+    },
+    {
+      "epoch": 0.13504098360655736,
+      "grad_norm": 5.899420261383057,
+      "learning_rate": 1.9428057070612224e-05,
+      "loss": 1.1816,
+      "step": 659
+    },
+    {
+      "epoch": 0.13524590163934427,
+      "grad_norm": 4.336621284484863,
+      "learning_rate": 1.94258423862515e-05,
+      "loss": 1.217,
+      "step": 660
+    },
+    {
+      "epoch": 0.13545081967213116,
+      "grad_norm": 10.756132125854492,
+      "learning_rate": 1.942362354903457e-05,
+      "loss": 1.2228,
+      "step": 661
+    },
+    {
+      "epoch": 0.13565573770491804,
+      "grad_norm": 5.198432922363281,
+      "learning_rate": 1.942140055993901e-05,
+      "loss": 1.2076,
+      "step": 662
+    },
+    {
+      "epoch": 0.13586065573770492,
+      "grad_norm": 6.024538516998291,
+      "learning_rate": 1.941917341994423e-05,
+      "loss": 1.3753,
+      "step": 663
+    },
+    {
+      "epoch": 0.1360655737704918,
+      "grad_norm": 6.440499782562256,
+      "learning_rate": 1.941694213003147e-05,
+      "loss": 1.4918,
+      "step": 664
+    },
+    {
+      "epoch": 0.1362704918032787,
+      "grad_norm": 9.738688468933105,
+      "learning_rate": 1.941470669118379e-05,
+      "loss": 1.123,
+      "step": 665
+    },
+    {
+      "epoch": 0.13647540983606557,
+      "grad_norm": 4.544641494750977,
+      "learning_rate": 1.9412467104386095e-05,
+      "loss": 1.0774,
+      "step": 666
+    },
+    {
+      "epoch": 0.13668032786885245,
+      "grad_norm": 6.209960460662842,
+      "learning_rate": 1.9410223370625096e-05,
+      "loss": 1.1265,
+      "step": 667
+    },
+    {
+      "epoch": 0.13688524590163934,
+      "grad_norm": 4.609410762786865,
+      "learning_rate": 1.940797549088935e-05,
+      "loss": 1.2667,
+      "step": 668
+    },
+    {
+      "epoch": 0.13709016393442622,
+      "grad_norm": 3.2782175540924072,
+      "learning_rate": 1.9405723466169223e-05,
+      "loss": 1.2499,
+      "step": 669
+    },
+    {
+      "epoch": 0.13729508196721313,
+      "grad_norm": 4.4499592781066895,
+      "learning_rate": 1.9403467297456925e-05,
+      "loss": 1.0527,
+      "step": 670
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 5.380029201507568,
+      "learning_rate": 1.940120698574648e-05,
+      "loss": 1.2761,
+      "step": 671
+    },
+    {
+      "epoch": 0.1377049180327869,
+      "grad_norm": 5.868254661560059,
+      "learning_rate": 1.9398942532033744e-05,
+      "loss": 0.9742,
+      "step": 672
+    },
+    {
+      "epoch": 0.13790983606557378,
+      "grad_norm": 4.638972759246826,
+      "learning_rate": 1.9396673937316386e-05,
+      "loss": 0.9768,
+      "step": 673
+    },
+    {
+      "epoch": 0.13811475409836066,
+      "grad_norm": 4.893986701965332,
+      "learning_rate": 1.9394401202593914e-05,
+      "loss": 0.9007,
+      "step": 674
+    },
+    {
+      "epoch": 0.13831967213114754,
+      "grad_norm": 6.20233154296875,
+      "learning_rate": 1.9392124328867653e-05,
+      "loss": 1.2089,
+      "step": 675
+    },
+    {
+      "epoch": 0.13852459016393442,
+      "grad_norm": 6.816092491149902,
+      "learning_rate": 1.9389843317140756e-05,
+      "loss": 0.9916,
+      "step": 676
+    },
+    {
+      "epoch": 0.1387295081967213,
+      "grad_norm": 6.451143264770508,
+      "learning_rate": 1.938755816841819e-05,
+      "loss": 1.3118,
+      "step": 677
+    },
+    {
+      "epoch": 0.1389344262295082,
+      "grad_norm": 3.7300808429718018,
+      "learning_rate": 1.9385268883706748e-05,
+      "loss": 1.0678,
+      "step": 678
+    },
+    {
+      "epoch": 0.13913934426229507,
+      "grad_norm": 3.8243231773376465,
+      "learning_rate": 1.9382975464015056e-05,
+      "loss": 1.2544,
+      "step": 679
+    },
+    {
+      "epoch": 0.13934426229508196,
+      "grad_norm": 5.900877952575684,
+      "learning_rate": 1.938067791035355e-05,
+      "loss": 1.2534,
+      "step": 680
+    },
+    {
+      "epoch": 0.13954918032786887,
+      "grad_norm": 3.9708058834075928,
+      "learning_rate": 1.9378376223734488e-05,
+      "loss": 1.2791,
+      "step": 681
+    },
+    {
+      "epoch": 0.13975409836065575,
+      "grad_norm": 3.7767293453216553,
+      "learning_rate": 1.9376070405171953e-05,
+      "loss": 1.2245,
+      "step": 682
+    },
+    {
+      "epoch": 0.13995901639344263,
+      "grad_norm": 9.243813514709473,
+      "learning_rate": 1.937376045568185e-05,
+      "loss": 1.165,
+      "step": 683
+    },
+    {
+      "epoch": 0.14016393442622951,
+      "grad_norm": 5.044832229614258,
+      "learning_rate": 1.93714463762819e-05,
+      "loss": 1.2471,
+      "step": 684
+    },
+    {
+      "epoch": 0.1403688524590164,
+      "grad_norm": 7.774890899658203,
+      "learning_rate": 1.936912816799164e-05,
+      "loss": 1.0306,
+      "step": 685
+    },
+    {
+      "epoch": 0.14057377049180328,
+      "grad_norm": 6.710658550262451,
+      "learning_rate": 1.936680583183244e-05,
+      "loss": 1.2935,
+      "step": 686
+    },
+    {
+      "epoch": 0.14077868852459016,
+      "grad_norm": 4.8215012550354,
+      "learning_rate": 1.936447936882747e-05,
+      "loss": 1.0818,
+      "step": 687
+    },
+    {
+      "epoch": 0.14098360655737704,
+      "grad_norm": 8.35384464263916,
+      "learning_rate": 1.9362148780001728e-05,
+      "loss": 1.3218,
+      "step": 688
+    },
+    {
+      "epoch": 0.14118852459016393,
+      "grad_norm": 6.55037260055542,
+      "learning_rate": 1.9359814066382035e-05,
+      "loss": 1.0965,
+      "step": 689
+    },
+    {
+      "epoch": 0.1413934426229508,
+      "grad_norm": 7.319297790527344,
+      "learning_rate": 1.9357475228997025e-05,
+      "loss": 1.121,
+      "step": 690
+    },
+    {
+      "epoch": 0.1415983606557377,
+      "grad_norm": 4.545801639556885,
+      "learning_rate": 1.9355132268877143e-05,
+      "loss": 1.2525,
+      "step": 691
+    },
+    {
+      "epoch": 0.1418032786885246,
+      "grad_norm": 5.578004360198975,
+      "learning_rate": 1.9352785187054654e-05,
+      "loss": 1.2062,
+      "step": 692
+    },
+    {
+      "epoch": 0.14200819672131149,
+      "grad_norm": 7.595458507537842,
+      "learning_rate": 1.9350433984563644e-05,
+      "loss": 1.163,
+      "step": 693
+    },
+    {
+      "epoch": 0.14221311475409837,
+      "grad_norm": 4.10064697265625,
+      "learning_rate": 1.9348078662440008e-05,
+      "loss": 0.9718,
+      "step": 694
+    },
+    {
+      "epoch": 0.14241803278688525,
+      "grad_norm": 3.226483106613159,
+      "learning_rate": 1.9345719221721457e-05,
+      "loss": 1.2588,
+      "step": 695
+    },
+    {
+      "epoch": 0.14262295081967213,
+      "grad_norm": 7.793419361114502,
+      "learning_rate": 1.9343355663447523e-05,
+      "loss": 1.221,
+      "step": 696
+    },
+    {
+      "epoch": 0.14282786885245902,
+      "grad_norm": 6.413252353668213,
+      "learning_rate": 1.934098798865954e-05,
+      "loss": 1.1071,
+      "step": 697
+    },
+    {
+      "epoch": 0.1430327868852459,
+      "grad_norm": 3.9249987602233887,
+      "learning_rate": 1.933861619840067e-05,
+      "loss": 1.271,
+      "step": 698
+    },
+    {
+      "epoch": 0.14323770491803278,
+      "grad_norm": 3.9033305644989014,
+      "learning_rate": 1.933624029371588e-05,
+      "loss": 0.9357,
+      "step": 699
+    },
+    {
+      "epoch": 0.14344262295081966,
+      "grad_norm": 4.667611598968506,
+      "learning_rate": 1.9333860275651947e-05,
+      "loss": 1.08,
+      "step": 700
+    },
+    {
+      "epoch": 0.14364754098360655,
+      "grad_norm": 5.5451860427856445,
+      "learning_rate": 1.9331476145257464e-05,
+      "loss": 1.042,
+      "step": 701
+    },
+    {
+      "epoch": 0.14385245901639343,
+      "grad_norm": 3.7335076332092285,
+      "learning_rate": 1.932908790358284e-05,
+      "loss": 1.0609,
+      "step": 702
+    },
+    {
+      "epoch": 0.14405737704918034,
+      "grad_norm": 3.421173095703125,
+      "learning_rate": 1.932669555168029e-05,
+      "loss": 0.9439,
+      "step": 703
+    },
+    {
+      "epoch": 0.14426229508196722,
+      "grad_norm": 7.05974006652832,
+      "learning_rate": 1.9324299090603836e-05,
+      "loss": 1.0342,
+      "step": 704
+    },
+    {
+      "epoch": 0.1444672131147541,
+      "grad_norm": 6.879289627075195,
+      "learning_rate": 1.932189852140932e-05,
+      "loss": 1.0289,
+      "step": 705
+    },
+    {
+      "epoch": 0.144672131147541,
+      "grad_norm": 4.777533531188965,
+      "learning_rate": 1.9319493845154392e-05,
+      "loss": 1.2211,
+      "step": 706
+    },
+    {
+      "epoch": 0.14487704918032787,
+      "grad_norm": 4.701208114624023,
+      "learning_rate": 1.9317085062898502e-05,
+      "loss": 1.2994,
+      "step": 707
+    },
+    {
+      "epoch": 0.14508196721311475,
+      "grad_norm": 5.547554016113281,
+      "learning_rate": 1.9314672175702924e-05,
+      "loss": 1.0109,
+      "step": 708
+    },
+    {
+      "epoch": 0.14528688524590164,
+      "grad_norm": 4.533079624176025,
+      "learning_rate": 1.9312255184630728e-05,
+      "loss": 1.1621,
+      "step": 709
+    },
+    {
+      "epoch": 0.14549180327868852,
+      "grad_norm": 6.742550849914551,
+      "learning_rate": 1.9309834090746796e-05,
+      "loss": 1.1638,
+      "step": 710
+    },
+    {
+      "epoch": 0.1456967213114754,
+      "grad_norm": 4.084738254547119,
+      "learning_rate": 1.9307408895117825e-05,
+      "loss": 0.9659,
+      "step": 711
+    },
+    {
+      "epoch": 0.14590163934426228,
+      "grad_norm": 6.984925746917725,
+      "learning_rate": 1.93049795988123e-05,
+      "loss": 1.2202,
+      "step": 712
+    },
+    {
+      "epoch": 0.14610655737704917,
+      "grad_norm": 3.6619210243225098,
+      "learning_rate": 1.9302546202900535e-05,
+      "loss": 0.8422,
+      "step": 713
+    },
+    {
+      "epoch": 0.14631147540983608,
+      "grad_norm": 5.66560173034668,
+      "learning_rate": 1.9300108708454637e-05,
+      "loss": 1.0603,
+      "step": 714
+    },
+    {
+      "epoch": 0.14651639344262296,
+      "grad_norm": 2.7827391624450684,
+      "learning_rate": 1.9297667116548525e-05,
+      "loss": 0.8322,
+      "step": 715
+    },
+    {
+      "epoch": 0.14672131147540984,
+      "grad_norm": 4.755606174468994,
+      "learning_rate": 1.9295221428257912e-05,
+      "loss": 0.9649,
+      "step": 716
+    },
+    {
+      "epoch": 0.14692622950819673,
+      "grad_norm": 4.07484245300293,
+      "learning_rate": 1.9292771644660337e-05,
+      "loss": 1.3952,
+      "step": 717
+    },
+    {
+      "epoch": 0.1471311475409836,
+      "grad_norm": 5.5521416664123535,
+      "learning_rate": 1.929031776683512e-05,
+      "loss": 0.912,
+      "step": 718
+    },
+    {
+      "epoch": 0.1473360655737705,
+      "grad_norm": 7.3150858879089355,
+      "learning_rate": 1.9287859795863402e-05,
+      "loss": 1.3561,
+      "step": 719
+    },
+    {
+      "epoch": 0.14754098360655737,
+      "grad_norm": 5.739613056182861,
+      "learning_rate": 1.9285397732828117e-05,
+      "loss": 1.0096,
+      "step": 720
+    },
+    {
+      "epoch": 0.14774590163934426,
+      "grad_norm": 5.016364097595215,
+      "learning_rate": 1.9282931578814004e-05,
+      "loss": 1.1148,
+      "step": 721
+    },
+    {
+      "epoch": 0.14795081967213114,
+      "grad_norm": 4.174158096313477,
+      "learning_rate": 1.9280461334907614e-05,
+      "loss": 1.0742,
+      "step": 722
+    },
+    {
+      "epoch": 0.14815573770491802,
+      "grad_norm": 4.95504903793335,
+      "learning_rate": 1.9277987002197283e-05,
+      "loss": 1.2665,
+      "step": 723
+    },
+    {
+      "epoch": 0.1483606557377049,
+      "grad_norm": 11.968973159790039,
+      "learning_rate": 1.927550858177316e-05,
+      "loss": 1.084,
+      "step": 724
+    },
+    {
+      "epoch": 0.14856557377049182,
+      "grad_norm": 3.329929828643799,
+      "learning_rate": 1.9273026074727192e-05,
+      "loss": 1.2745,
+      "step": 725
+    },
+    {
+      "epoch": 0.1487704918032787,
+      "grad_norm": 2.9051711559295654,
+      "learning_rate": 1.927053948215313e-05,
+      "loss": 1.1149,
+      "step": 726
+    },
+    {
+      "epoch": 0.14897540983606558,
+      "grad_norm": 5.434474945068359,
+      "learning_rate": 1.9268048805146518e-05,
+      "loss": 1.2975,
+      "step": 727
+    },
+    {
+      "epoch": 0.14918032786885246,
+      "grad_norm": 4.804509162902832,
+      "learning_rate": 1.9265554044804707e-05,
+      "loss": 0.9512,
+      "step": 728
+    },
+    {
+      "epoch": 0.14938524590163935,
+      "grad_norm": 4.699665546417236,
+      "learning_rate": 1.926305520222684e-05,
+      "loss": 1.0468,
+      "step": 729
+    },
+    {
+      "epoch": 0.14959016393442623,
+      "grad_norm": 9.989665985107422,
+      "learning_rate": 1.9260552278513868e-05,
+      "loss": 0.9745,
+      "step": 730
+    },
+    {
+      "epoch": 0.1497950819672131,
+      "grad_norm": 6.6286702156066895,
+      "learning_rate": 1.9258045274768524e-05,
+      "loss": 1.1453,
+      "step": 731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.909843921661377,
+      "learning_rate": 1.9255534192095355e-05,
+      "loss": 1.1517,
+      "step": 732
+    },
+    {
+      "epoch": 0.15020491803278688,
+      "grad_norm": 5.781076908111572,
+      "learning_rate": 1.92530190316007e-05,
+      "loss": 1.1248,
+      "step": 733
+    },
+    {
+      "epoch": 0.15040983606557376,
+      "grad_norm": 4.554798603057861,
+      "learning_rate": 1.925049979439269e-05,
+      "loss": 1.1628,
+      "step": 734
+    },
+    {
+      "epoch": 0.15061475409836064,
+      "grad_norm": 7.83950662612915,
+      "learning_rate": 1.924797648158126e-05,
+      "loss": 1.0138,
+      "step": 735
+    },
+    {
+      "epoch": 0.15081967213114755,
+      "grad_norm": 5.327502250671387,
+      "learning_rate": 1.924544909427813e-05,
+      "loss": 0.9196,
+      "step": 736
+    },
+    {
+      "epoch": 0.15102459016393444,
+      "grad_norm": 5.589282989501953,
+      "learning_rate": 1.9242917633596824e-05,
+      "loss": 0.9805,
+      "step": 737
+    },
+    {
+      "epoch": 0.15122950819672132,
+      "grad_norm": 5.69791316986084,
+      "learning_rate": 1.9240382100652658e-05,
+      "loss": 1.1691,
+      "step": 738
+    },
+    {
+      "epoch": 0.1514344262295082,
+      "grad_norm": 4.0950703620910645,
+      "learning_rate": 1.9237842496562744e-05,
+      "loss": 1.2091,
+      "step": 739
+    },
+    {
+      "epoch": 0.15163934426229508,
+      "grad_norm": 4.809014797210693,
+      "learning_rate": 1.9235298822445983e-05,
+      "loss": 1.0742,
+      "step": 740
+    },
+    {
+      "epoch": 0.15184426229508197,
+      "grad_norm": 2.8001630306243896,
+      "learning_rate": 1.9232751079423073e-05,
+      "loss": 1.0205,
+      "step": 741
+    },
+    {
+      "epoch": 0.15204918032786885,
+      "grad_norm": 6.461826324462891,
+      "learning_rate": 1.9230199268616504e-05,
+      "loss": 1.1884,
+      "step": 742
+    },
+    {
+      "epoch": 0.15225409836065573,
+      "grad_norm": 5.973479270935059,
+      "learning_rate": 1.9227643391150558e-05,
+      "loss": 1.1657,
+      "step": 743
+    },
+    {
+      "epoch": 0.15245901639344261,
+      "grad_norm": 6.243319988250732,
+      "learning_rate": 1.922508344815131e-05,
+      "loss": 1.0861,
+      "step": 744
+    },
+    {
+      "epoch": 0.1526639344262295,
+      "grad_norm": 3.1543004512786865,
+      "learning_rate": 1.9222519440746618e-05,
+      "loss": 1.1831,
+      "step": 745
+    },
+    {
+      "epoch": 0.15286885245901638,
+      "grad_norm": 4.410238265991211,
+      "learning_rate": 1.9219951370066145e-05,
+      "loss": 1.2276,
+      "step": 746
+    },
+    {
+      "epoch": 0.1530737704918033,
+      "grad_norm": 9.389508247375488,
+      "learning_rate": 1.9217379237241334e-05,
+      "loss": 1.1559,
+      "step": 747
+    },
+    {
+      "epoch": 0.15327868852459017,
+      "grad_norm": 6.480701446533203,
+      "learning_rate": 1.921480304340542e-05,
+      "loss": 1.3389,
+      "step": 748
+    },
+    {
+      "epoch": 0.15348360655737706,
+      "grad_norm": 5.2343363761901855,
+      "learning_rate": 1.9212222789693427e-05,
+      "loss": 1.2587,
+      "step": 749
+    },
+    {
+      "epoch": 0.15368852459016394,
+      "grad_norm": 3.340850591659546,
+      "learning_rate": 1.9209638477242168e-05,
+      "loss": 0.9687,
+      "step": 750
+    },
+    {
+      "epoch": 0.15389344262295082,
+      "grad_norm": 4.081146717071533,
+      "learning_rate": 1.920705010719025e-05,
+      "loss": 1.0939,
+      "step": 751
+    },
+    {
+      "epoch": 0.1540983606557377,
+      "grad_norm": 4.059302806854248,
+      "learning_rate": 1.9204457680678053e-05,
+      "loss": 1.0833,
+      "step": 752
+    },
+    {
+      "epoch": 0.1543032786885246,
+      "grad_norm": 5.792212963104248,
+      "learning_rate": 1.9201861198847757e-05,
+      "loss": 1.0553,
+      "step": 753
+    },
+    {
+      "epoch": 0.15450819672131147,
+      "grad_norm": 6.079855918884277,
+      "learning_rate": 1.919926066284333e-05,
+      "loss": 1.1042,
+      "step": 754
+    },
+    {
+      "epoch": 0.15471311475409835,
+      "grad_norm": 3.5828325748443604,
+      "learning_rate": 1.9196656073810517e-05,
+      "loss": 0.9969,
+      "step": 755
+    },
+    {
+      "epoch": 0.15491803278688523,
+      "grad_norm": 6.7016472816467285,
+      "learning_rate": 1.9194047432896853e-05,
+      "loss": 1.1052,
+      "step": 756
+    },
+    {
+      "epoch": 0.15512295081967212,
+      "grad_norm": 4.614019870758057,
+      "learning_rate": 1.9191434741251655e-05,
+      "loss": 1.1383,
+      "step": 757
+    },
+    {
+      "epoch": 0.15532786885245903,
+      "grad_norm": 4.094242572784424,
+      "learning_rate": 1.918881800002604e-05,
+      "loss": 1.3388,
+      "step": 758
+    },
+    {
+      "epoch": 0.1555327868852459,
+      "grad_norm": 5.61146354675293,
+      "learning_rate": 1.918619721037288e-05,
+      "loss": 0.844,
+      "step": 759
+    },
+    {
+      "epoch": 0.1557377049180328,
+      "grad_norm": 3.6409482955932617,
+      "learning_rate": 1.9183572373446856e-05,
+      "loss": 0.8187,
+      "step": 760
+    },
+    {
+      "epoch": 0.15594262295081968,
+      "grad_norm": 5.699113845825195,
+      "learning_rate": 1.9180943490404428e-05,
+      "loss": 0.994,
+      "step": 761
+    },
+    {
+      "epoch": 0.15614754098360656,
+      "grad_norm": 3.9078662395477295,
+      "learning_rate": 1.917831056240383e-05,
+      "loss": 1.0565,
+      "step": 762
+    },
+    {
+      "epoch": 0.15635245901639344,
+      "grad_norm": 6.257782936096191,
+      "learning_rate": 1.917567359060508e-05,
+      "loss": 1.1897,
+      "step": 763
+    },
+    {
+      "epoch": 0.15655737704918032,
+      "grad_norm": 4.143136501312256,
+      "learning_rate": 1.9173032576169986e-05,
+      "loss": 1.0006,
+      "step": 764
+    },
+    {
+      "epoch": 0.1567622950819672,
+      "grad_norm": 5.7340168952941895,
+      "learning_rate": 1.9170387520262127e-05,
+      "loss": 1.3242,
+      "step": 765
+    },
+    {
+      "epoch": 0.1569672131147541,
+      "grad_norm": 3.783449649810791,
+      "learning_rate": 1.9167738424046865e-05,
+      "loss": 0.9966,
+      "step": 766
+    },
+    {
+      "epoch": 0.15717213114754097,
+      "grad_norm": 3.5381367206573486,
+      "learning_rate": 1.9165085288691356e-05,
+      "loss": 0.8799,
+      "step": 767
+    },
+    {
+      "epoch": 0.15737704918032788,
+      "grad_norm": 7.452739715576172,
+      "learning_rate": 1.916242811536451e-05,
+      "loss": 1.2124,
+      "step": 768
+    },
+    {
+      "epoch": 0.15758196721311477,
+      "grad_norm": 7.143852710723877,
+      "learning_rate": 1.9159766905237032e-05,
+      "loss": 1.1894,
+      "step": 769
+    },
+    {
+      "epoch": 0.15778688524590165,
+      "grad_norm": 4.078629493713379,
+      "learning_rate": 1.915710165948141e-05,
+      "loss": 1.2699,
+      "step": 770
+    },
+    {
+      "epoch": 0.15799180327868853,
+      "grad_norm": 4.973592281341553,
+      "learning_rate": 1.91544323792719e-05,
+      "loss": 1.1809,
+      "step": 771
+    },
+    {
+      "epoch": 0.1581967213114754,
+      "grad_norm": 5.4835429191589355,
+      "learning_rate": 1.9151759065784536e-05,
+      "loss": 1.3398,
+      "step": 772
+    },
+    {
+      "epoch": 0.1584016393442623,
+      "grad_norm": 3.0752012729644775,
+      "learning_rate": 1.914908172019713e-05,
+      "loss": 1.0608,
+      "step": 773
+    },
+    {
+      "epoch": 0.15860655737704918,
+      "grad_norm": 4.3416852951049805,
+      "learning_rate": 1.914640034368928e-05,
+      "loss": 1.0594,
+      "step": 774
+    },
+    {
+      "epoch": 0.15881147540983606,
+      "grad_norm": 6.57863712310791,
+      "learning_rate": 1.9143714937442347e-05,
+      "loss": 1.1707,
+      "step": 775
+    },
+    {
+      "epoch": 0.15901639344262294,
+      "grad_norm": 5.314719200134277,
+      "learning_rate": 1.9141025502639475e-05,
+      "loss": 1.0621,
+      "step": 776
+    },
+    {
+      "epoch": 0.15922131147540983,
+      "grad_norm": 5.544409275054932,
+      "learning_rate": 1.9138332040465576e-05,
+      "loss": 1.1435,
+      "step": 777
+    },
+    {
+      "epoch": 0.1594262295081967,
+      "grad_norm": 5.680751800537109,
+      "learning_rate": 1.9135634552107347e-05,
+      "loss": 1.1446,
+      "step": 778
+    },
+    {
+      "epoch": 0.15963114754098362,
+      "grad_norm": 5.399108409881592,
+      "learning_rate": 1.9132933038753246e-05,
+      "loss": 1.1598,
+      "step": 779
+    },
+    {
+      "epoch": 0.1598360655737705,
+      "grad_norm": 6.340021133422852,
+      "learning_rate": 1.913022750159352e-05,
+      "loss": 1.224,
+      "step": 780
+    },
+    {
+      "epoch": 0.16004098360655739,
+      "grad_norm": 5.792757511138916,
+      "learning_rate": 1.9127517941820172e-05,
+      "loss": 1.0773,
+      "step": 781
+    },
+    {
+      "epoch": 0.16024590163934427,
+      "grad_norm": 9.777632713317871,
+      "learning_rate": 1.9124804360626986e-05,
+      "loss": 1.2867,
+      "step": 782
+    },
+    {
+      "epoch": 0.16045081967213115,
+      "grad_norm": 6.881640911102295,
+      "learning_rate": 1.9122086759209518e-05,
+      "loss": 1.2301,
+      "step": 783
+    },
+    {
+      "epoch": 0.16065573770491803,
+      "grad_norm": 4.8798699378967285,
+      "learning_rate": 1.9119365138765098e-05,
+      "loss": 0.9718,
+      "step": 784
+    },
+    {
+      "epoch": 0.16086065573770492,
+      "grad_norm": 4.430053234100342,
+      "learning_rate": 1.9116639500492818e-05,
+      "loss": 0.87,
+      "step": 785
+    },
+    {
+      "epoch": 0.1610655737704918,
+      "grad_norm": 4.801238059997559,
+      "learning_rate": 1.9113909845593548e-05,
+      "loss": 1.2342,
+      "step": 786
+    },
+    {
+      "epoch": 0.16127049180327868,
+      "grad_norm": 4.009771347045898,
+      "learning_rate": 1.9111176175269923e-05,
+      "loss": 1.2447,
+      "step": 787
+    },
+    {
+      "epoch": 0.16147540983606556,
+      "grad_norm": 4.669938564300537,
+      "learning_rate": 1.9108438490726348e-05,
+      "loss": 1.0929,
+      "step": 788
+    },
+    {
+      "epoch": 0.16168032786885245,
+      "grad_norm": 6.88905668258667,
+      "learning_rate": 1.9105696793169002e-05,
+      "loss": 1.3473,
+      "step": 789
+    },
+    {
+      "epoch": 0.16188524590163936,
+      "grad_norm": 4.569355010986328,
+      "learning_rate": 1.9102951083805822e-05,
+      "loss": 1.0415,
+      "step": 790
+    },
+    {
+      "epoch": 0.16209016393442624,
+      "grad_norm": 9.68996524810791,
+      "learning_rate": 1.9100201363846523e-05,
+      "loss": 1.2485,
+      "step": 791
+    },
+    {
+      "epoch": 0.16229508196721312,
+      "grad_norm": 6.543159484863281,
+      "learning_rate": 1.909744763450258e-05,
+      "loss": 0.9613,
+      "step": 792
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 5.662082195281982,
+      "learning_rate": 1.9094689896987232e-05,
+      "loss": 1.1336,
+      "step": 793
+    },
+    {
+      "epoch": 0.1627049180327869,
+      "grad_norm": 6.278080463409424,
+      "learning_rate": 1.909192815251549e-05,
+      "loss": 1.0304,
+      "step": 794
+    },
+    {
+      "epoch": 0.16290983606557377,
+      "grad_norm": 3.4243741035461426,
+      "learning_rate": 1.908916240230414e-05,
+      "loss": 0.9589,
+      "step": 795
+    },
+    {
+      "epoch": 0.16311475409836065,
+      "grad_norm": 6.927191257476807,
+      "learning_rate": 1.9086392647571707e-05,
+      "loss": 1.0673,
+      "step": 796
+    },
+    {
+      "epoch": 0.16331967213114754,
+      "grad_norm": 7.489114284515381,
+      "learning_rate": 1.90836188895385e-05,
+      "loss": 1.2201,
+      "step": 797
+    },
+    {
+      "epoch": 0.16352459016393442,
+      "grad_norm": 4.3211517333984375,
+      "learning_rate": 1.908084112942659e-05,
+      "loss": 1.1649,
+      "step": 798
+    },
+    {
+      "epoch": 0.1637295081967213,
+      "grad_norm": 16.77748680114746,
+      "learning_rate": 1.9078059368459807e-05,
+      "loss": 1.3231,
+      "step": 799
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 4.358295917510986,
+      "learning_rate": 1.907527360786374e-05,
+      "loss": 1.1525,
+      "step": 800
+    },
+    {
+      "epoch": 0.1641393442622951,
+      "grad_norm": 6.60353422164917,
+      "learning_rate": 1.907248384886575e-05,
+      "loss": 1.1931,
+      "step": 801
+    },
+    {
+      "epoch": 0.16434426229508198,
+      "grad_norm": 5.945642471313477,
+      "learning_rate": 1.9069690092694956e-05,
+      "loss": 1.1117,
+      "step": 802
+    },
+    {
+      "epoch": 0.16454918032786886,
+      "grad_norm": 7.507762908935547,
+      "learning_rate": 1.906689234058223e-05,
+      "loss": 1.1731,
+      "step": 803
+    },
+    {
+      "epoch": 0.16475409836065574,
+      "grad_norm": 6.143866539001465,
+      "learning_rate": 1.9064090593760215e-05,
+      "loss": 1.2503,
+      "step": 804
+    },
+    {
+      "epoch": 0.16495901639344263,
+      "grad_norm": 4.217045783996582,
+      "learning_rate": 1.9061284853463313e-05,
+      "loss": 1.0861,
+      "step": 805
+    },
+    {
+      "epoch": 0.1651639344262295,
+      "grad_norm": 4.465440273284912,
+      "learning_rate": 1.9058475120927674e-05,
+      "loss": 1.2834,
+      "step": 806
+    },
+    {
+      "epoch": 0.1653688524590164,
+      "grad_norm": 4.464051723480225,
+      "learning_rate": 1.9055661397391224e-05,
+      "loss": 1.1486,
+      "step": 807
+    },
+    {
+      "epoch": 0.16557377049180327,
+      "grad_norm": 6.9499735832214355,
+      "learning_rate": 1.9052843684093633e-05,
+      "loss": 1.2463,
+      "step": 808
+    },
+    {
+      "epoch": 0.16577868852459016,
+      "grad_norm": 6.590907096862793,
+      "learning_rate": 1.905002198227634e-05,
+      "loss": 1.1071,
+      "step": 809
+    },
+    {
+      "epoch": 0.16598360655737704,
+      "grad_norm": 2.9030160903930664,
+      "learning_rate": 1.9047196293182533e-05,
+      "loss": 1.0347,
+      "step": 810
+    },
+    {
+      "epoch": 0.16618852459016392,
+      "grad_norm": 4.754637241363525,
+      "learning_rate": 1.9044366618057157e-05,
+      "loss": 1.3194,
+      "step": 811
+    },
+    {
+      "epoch": 0.16639344262295083,
+      "grad_norm": 4.793728828430176,
+      "learning_rate": 1.9041532958146924e-05,
+      "loss": 1.1023,
+      "step": 812
+    },
+    {
+      "epoch": 0.16659836065573772,
+      "grad_norm": 3.5127248764038086,
+      "learning_rate": 1.9038695314700283e-05,
+      "loss": 1.1229,
+      "step": 813
+    },
+    {
+      "epoch": 0.1668032786885246,
+      "grad_norm": 12.221125602722168,
+      "learning_rate": 1.9035853688967455e-05,
+      "loss": 1.3662,
+      "step": 814
+    },
+    {
+      "epoch": 0.16700819672131148,
+      "grad_norm": 6.597712516784668,
+      "learning_rate": 1.903300808220041e-05,
+      "loss": 1.2629,
+      "step": 815
+    },
+    {
+      "epoch": 0.16721311475409836,
+      "grad_norm": 7.545158386230469,
+      "learning_rate": 1.9030158495652867e-05,
+      "loss": 1.0816,
+      "step": 816
+    },
+    {
+      "epoch": 0.16741803278688525,
+      "grad_norm": 4.376674652099609,
+      "learning_rate": 1.9027304930580307e-05,
+      "loss": 1.1442,
+      "step": 817
+    },
+    {
+      "epoch": 0.16762295081967213,
+      "grad_norm": 8.037652015686035,
+      "learning_rate": 1.9024447388239956e-05,
+      "loss": 1.1605,
+      "step": 818
+    },
+    {
+      "epoch": 0.167827868852459,
+      "grad_norm": 8.78840446472168,
+      "learning_rate": 1.9021585869890792e-05,
+      "loss": 1.4342,
+      "step": 819
+    },
+    {
+      "epoch": 0.1680327868852459,
+      "grad_norm": 2.639348030090332,
+      "learning_rate": 1.9018720376793556e-05,
+      "loss": 0.8361,
+      "step": 820
+    },
+    {
+      "epoch": 0.16823770491803278,
+      "grad_norm": 3.479290246963501,
+      "learning_rate": 1.9015850910210728e-05,
+      "loss": 1.1162,
+      "step": 821
+    },
+    {
+      "epoch": 0.16844262295081966,
+      "grad_norm": 5.278480529785156,
+      "learning_rate": 1.9012977471406547e-05,
+      "loss": 1.3684,
+      "step": 822
+    },
+    {
+      "epoch": 0.16864754098360657,
+      "grad_norm": 3.4766647815704346,
+      "learning_rate": 1.9010100061646993e-05,
+      "loss": 1.0745,
+      "step": 823
+    },
+    {
+      "epoch": 0.16885245901639345,
+      "grad_norm": 3.7447385787963867,
+      "learning_rate": 1.900721868219981e-05,
+      "loss": 0.9131,
+      "step": 824
+    },
+    {
+      "epoch": 0.16905737704918034,
+      "grad_norm": 11.556915283203125,
+      "learning_rate": 1.900433333433447e-05,
+      "loss": 1.0828,
+      "step": 825
+    },
+    {
+      "epoch": 0.16926229508196722,
+      "grad_norm": 4.173906326293945,
+      "learning_rate": 1.9001444019322212e-05,
+      "loss": 1.1787,
+      "step": 826
+    },
+    {
+      "epoch": 0.1694672131147541,
+      "grad_norm": 5.1573991775512695,
+      "learning_rate": 1.8998550738436013e-05,
+      "loss": 1.2056,
+      "step": 827
+    },
+    {
+      "epoch": 0.16967213114754098,
+      "grad_norm": 4.877800464630127,
+      "learning_rate": 1.8995653492950603e-05,
+      "loss": 0.8765,
+      "step": 828
+    },
+    {
+      "epoch": 0.16987704918032787,
+      "grad_norm": 5.339267730712891,
+      "learning_rate": 1.8992752284142458e-05,
+      "loss": 1.1746,
+      "step": 829
+    },
+    {
+      "epoch": 0.17008196721311475,
+      "grad_norm": 5.412646770477295,
+      "learning_rate": 1.8989847113289794e-05,
+      "loss": 1.1628,
+      "step": 830
+    },
+    {
+      "epoch": 0.17028688524590163,
+      "grad_norm": 3.8684234619140625,
+      "learning_rate": 1.8986937981672575e-05,
+      "loss": 1.0646,
+      "step": 831
+    },
+    {
+      "epoch": 0.17049180327868851,
+      "grad_norm": 4.94945764541626,
+      "learning_rate": 1.898402489057252e-05,
+      "loss": 0.9952,
+      "step": 832
+    },
+    {
+      "epoch": 0.1706967213114754,
+      "grad_norm": 4.9288554191589355,
+      "learning_rate": 1.8981107841273082e-05,
+      "loss": 1.1283,
+      "step": 833
+    },
+    {
+      "epoch": 0.1709016393442623,
+      "grad_norm": 3.4554831981658936,
+      "learning_rate": 1.8978186835059455e-05,
+      "loss": 1.0823,
+      "step": 834
+    },
+    {
+      "epoch": 0.1711065573770492,
+      "grad_norm": 5.88681697845459,
+      "learning_rate": 1.8975261873218582e-05,
+      "loss": 1.273,
+      "step": 835
+    },
+    {
+      "epoch": 0.17131147540983607,
+      "grad_norm": 9.030847549438477,
+      "learning_rate": 1.8972332957039158e-05,
+      "loss": 1.0632,
+      "step": 836
+    },
+    {
+      "epoch": 0.17151639344262296,
+      "grad_norm": 3.826826333999634,
+      "learning_rate": 1.89694000878116e-05,
+      "loss": 0.8355,
+      "step": 837
+    },
+    {
+      "epoch": 0.17172131147540984,
+      "grad_norm": 3.708608627319336,
+      "learning_rate": 1.896646326682808e-05,
+      "loss": 0.8625,
+      "step": 838
+    },
+    {
+      "epoch": 0.17192622950819672,
+      "grad_norm": 11.499753952026367,
+      "learning_rate": 1.8963522495382512e-05,
+      "loss": 1.1701,
+      "step": 839
+    },
+    {
+      "epoch": 0.1721311475409836,
+      "grad_norm": 4.342697620391846,
+      "learning_rate": 1.8960577774770544e-05,
+      "loss": 0.791,
+      "step": 840
+    },
+    {
+      "epoch": 0.17233606557377049,
+      "grad_norm": 9.236817359924316,
+      "learning_rate": 1.895762910628957e-05,
+      "loss": 0.8294,
+      "step": 841
+    },
+    {
+      "epoch": 0.17254098360655737,
+      "grad_norm": 6.659608364105225,
+      "learning_rate": 1.8954676491238714e-05,
+      "loss": 1.0778,
+      "step": 842
+    },
+    {
+      "epoch": 0.17274590163934425,
+      "grad_norm": 6.934392929077148,
+      "learning_rate": 1.8951719930918847e-05,
+      "loss": 1.1735,
+      "step": 843
+    },
+    {
+      "epoch": 0.17295081967213113,
+      "grad_norm": 7.306548595428467,
+      "learning_rate": 1.894875942663258e-05,
+      "loss": 0.9503,
+      "step": 844
+    },
+    {
+      "epoch": 0.17315573770491804,
+      "grad_norm": 4.932034492492676,
+      "learning_rate": 1.894579497968426e-05,
+      "loss": 1.1869,
+      "step": 845
+    },
+    {
+      "epoch": 0.17336065573770493,
+      "grad_norm": 5.586042404174805,
+      "learning_rate": 1.8942826591379958e-05,
+      "loss": 0.9453,
+      "step": 846
+    },
+    {
+      "epoch": 0.1735655737704918,
+      "grad_norm": 11.414579391479492,
+      "learning_rate": 1.8939854263027504e-05,
+      "loss": 1.2409,
+      "step": 847
+    },
+    {
+      "epoch": 0.1737704918032787,
+      "grad_norm": 7.1155877113342285,
+      "learning_rate": 1.8936877995936442e-05,
+      "loss": 1.0056,
+      "step": 848
+    },
+    {
+      "epoch": 0.17397540983606558,
+      "grad_norm": 6.435999870300293,
+      "learning_rate": 1.8933897791418072e-05,
+      "loss": 1.0871,
+      "step": 849
+    },
+    {
+      "epoch": 0.17418032786885246,
+      "grad_norm": 3.8874731063842773,
+      "learning_rate": 1.893091365078541e-05,
+      "loss": 1.0291,
+      "step": 850
+    },
+    {
+      "epoch": 0.17438524590163934,
+      "grad_norm": 4.376993656158447,
+      "learning_rate": 1.8927925575353215e-05,
+      "loss": 1.253,
+      "step": 851
+    },
+    {
+      "epoch": 0.17459016393442622,
+      "grad_norm": 6.062700271606445,
+      "learning_rate": 1.8924933566437985e-05,
+      "loss": 1.0067,
+      "step": 852
+    },
+    {
+      "epoch": 0.1747950819672131,
+      "grad_norm": 5.092137813568115,
+      "learning_rate": 1.8921937625357944e-05,
+      "loss": 1.2825,
+      "step": 853
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 5.802750110626221,
+      "learning_rate": 1.8918937753433046e-05,
+      "loss": 1.2491,
+      "step": 854
+    },
+    {
+      "epoch": 0.17520491803278687,
+      "grad_norm": 6.454592227935791,
+      "learning_rate": 1.8915933951984978e-05,
+      "loss": 1.1827,
+      "step": 855
+    },
+    {
+      "epoch": 0.17540983606557378,
+      "grad_norm": 4.921342372894287,
+      "learning_rate": 1.8912926222337168e-05,
+      "loss": 1.1993,
+      "step": 856
+    },
+    {
+      "epoch": 0.17561475409836066,
+      "grad_norm": 4.026766300201416,
+      "learning_rate": 1.8909914565814765e-05,
+      "loss": 1.0195,
+      "step": 857
+    },
+    {
+      "epoch": 0.17581967213114755,
+      "grad_norm": 10.060113906860352,
+      "learning_rate": 1.890689898374465e-05,
+      "loss": 1.4534,
+      "step": 858
+    },
+    {
+      "epoch": 0.17602459016393443,
+      "grad_norm": 6.2751898765563965,
+      "learning_rate": 1.890387947745543e-05,
+      "loss": 1.1788,
+      "step": 859
+    },
+    {
+      "epoch": 0.1762295081967213,
+      "grad_norm": 3.9526219367980957,
+      "learning_rate": 1.8900856048277453e-05,
+      "loss": 1.1567,
+      "step": 860
+    },
+    {
+      "epoch": 0.1764344262295082,
+      "grad_norm": 3.568697214126587,
+      "learning_rate": 1.8897828697542783e-05,
+      "loss": 0.9182,
+      "step": 861
+    },
+    {
+      "epoch": 0.17663934426229508,
+      "grad_norm": 3.1825404167175293,
+      "learning_rate": 1.8894797426585216e-05,
+      "loss": 1.1595,
+      "step": 862
+    },
+    {
+      "epoch": 0.17684426229508196,
+      "grad_norm": 5.4135260581970215,
+      "learning_rate": 1.8891762236740278e-05,
+      "loss": 0.9991,
+      "step": 863
+    },
+    {
+      "epoch": 0.17704918032786884,
+      "grad_norm": 4.676265239715576,
+      "learning_rate": 1.888872312934521e-05,
+      "loss": 0.9895,
+      "step": 864
+    },
+    {
+      "epoch": 0.17725409836065573,
+      "grad_norm": 3.7680716514587402,
+      "learning_rate": 1.8885680105739002e-05,
+      "loss": 1.1171,
+      "step": 865
+    },
+    {
+      "epoch": 0.17745901639344264,
+      "grad_norm": 6.586954116821289,
+      "learning_rate": 1.8882633167262344e-05,
+      "loss": 0.8914,
+      "step": 866
+    },
+    {
+      "epoch": 0.17766393442622952,
+      "grad_norm": 5.360543251037598,
+      "learning_rate": 1.887958231525767e-05,
+      "loss": 1.2682,
+      "step": 867
+    },
+    {
+      "epoch": 0.1778688524590164,
+      "grad_norm": 5.0391693115234375,
+      "learning_rate": 1.887652755106912e-05,
+      "loss": 1.1863,
+      "step": 868
+    },
+    {
+      "epoch": 0.17807377049180328,
+      "grad_norm": 5.029898643493652,
+      "learning_rate": 1.8873468876042576e-05,
+      "loss": 1.3448,
+      "step": 869
+    },
+    {
+      "epoch": 0.17827868852459017,
+      "grad_norm": 5.439546585083008,
+      "learning_rate": 1.8870406291525638e-05,
+      "loss": 1.1326,
+      "step": 870
+    },
+    {
+      "epoch": 0.17848360655737705,
+      "grad_norm": 6.989493370056152,
+      "learning_rate": 1.886733979886762e-05,
+      "loss": 1.2194,
+      "step": 871
+    },
+    {
+      "epoch": 0.17868852459016393,
+      "grad_norm": 7.108112335205078,
+      "learning_rate": 1.8864269399419562e-05,
+      "loss": 1.0819,
+      "step": 872
+    },
+    {
+      "epoch": 0.17889344262295082,
+      "grad_norm": 5.812595844268799,
+      "learning_rate": 1.886119509453423e-05,
+      "loss": 1.177,
+      "step": 873
+    },
+    {
+      "epoch": 0.1790983606557377,
+      "grad_norm": 5.955722808837891,
+      "learning_rate": 1.8858116885566108e-05,
+      "loss": 1.0257,
+      "step": 874
+    },
+    {
+      "epoch": 0.17930327868852458,
+      "grad_norm": 6.998990058898926,
+      "learning_rate": 1.88550347738714e-05,
+      "loss": 1.2001,
+      "step": 875
+    },
+    {
+      "epoch": 0.17950819672131146,
+      "grad_norm": 2.98923921585083,
+      "learning_rate": 1.8851948760808022e-05,
+      "loss": 0.9724,
+      "step": 876
+    },
+    {
+      "epoch": 0.17971311475409837,
+      "grad_norm": 4.63785982131958,
+      "learning_rate": 1.8848858847735622e-05,
+      "loss": 1.2002,
+      "step": 877
+    },
+    {
+      "epoch": 0.17991803278688526,
+      "grad_norm": 4.762745380401611,
+      "learning_rate": 1.884576503601556e-05,
+      "loss": 0.9372,
+      "step": 878
+    },
+    {
+      "epoch": 0.18012295081967214,
+      "grad_norm": 3.9985239505767822,
+      "learning_rate": 1.8842667327010908e-05,
+      "loss": 1.0848,
+      "step": 879
+    },
+    {
+      "epoch": 0.18032786885245902,
+      "grad_norm": 16.560516357421875,
+      "learning_rate": 1.883956572208647e-05,
+      "loss": 1.1748,
+      "step": 880
+    },
+    {
+      "epoch": 0.1805327868852459,
+      "grad_norm": 2.9951374530792236,
+      "learning_rate": 1.8836460222608748e-05,
+      "loss": 1.1309,
+      "step": 881
+    },
+    {
+      "epoch": 0.1807377049180328,
+      "grad_norm": 4.132318019866943,
+      "learning_rate": 1.8833350829945977e-05,
+      "loss": 1.1431,
+      "step": 882
+    },
+    {
+      "epoch": 0.18094262295081967,
+      "grad_norm": 9.6427640914917,
+      "learning_rate": 1.8830237545468092e-05,
+      "loss": 1.073,
+      "step": 883
+    },
+    {
+      "epoch": 0.18114754098360655,
+      "grad_norm": 6.989070415496826,
+      "learning_rate": 1.8827120370546753e-05,
+      "loss": 1.1558,
+      "step": 884
+    },
+    {
+      "epoch": 0.18135245901639344,
+      "grad_norm": 5.4212117195129395,
+      "learning_rate": 1.8823999306555334e-05,
+      "loss": 1.2187,
+      "step": 885
+    },
+    {
+      "epoch": 0.18155737704918032,
+      "grad_norm": 4.593606948852539,
+      "learning_rate": 1.8820874354868916e-05,
+      "loss": 1.1365,
+      "step": 886
+    },
+    {
+      "epoch": 0.1817622950819672,
+      "grad_norm": 6.507501602172852,
+      "learning_rate": 1.8817745516864296e-05,
+      "loss": 1.0331,
+      "step": 887
+    },
+    {
+      "epoch": 0.1819672131147541,
+      "grad_norm": 9.231318473815918,
+      "learning_rate": 1.8814612793919987e-05,
+      "loss": 0.9408,
+      "step": 888
+    },
+    {
+      "epoch": 0.182172131147541,
+      "grad_norm": 11.905319213867188,
+      "learning_rate": 1.8811476187416206e-05,
+      "loss": 0.9628,
+      "step": 889
+    },
+    {
+      "epoch": 0.18237704918032788,
+      "grad_norm": 8.971048355102539,
+      "learning_rate": 1.880833569873489e-05,
+      "loss": 0.9458,
+      "step": 890
+    },
+    {
+      "epoch": 0.18258196721311476,
+      "grad_norm": 5.020455837249756,
+      "learning_rate": 1.8805191329259677e-05,
+      "loss": 1.0833,
+      "step": 891
+    },
+    {
+      "epoch": 0.18278688524590164,
+      "grad_norm": 4.823147773742676,
+      "learning_rate": 1.8802043080375924e-05,
+      "loss": 1.2789,
+      "step": 892
+    },
+    {
+      "epoch": 0.18299180327868853,
+      "grad_norm": 9.952731132507324,
+      "learning_rate": 1.879889095347069e-05,
+      "loss": 1.3852,
+      "step": 893
+    },
+    {
+      "epoch": 0.1831967213114754,
+      "grad_norm": 6.098225116729736,
+      "learning_rate": 1.8795734949932746e-05,
+      "loss": 1.2838,
+      "step": 894
+    },
+    {
+      "epoch": 0.1834016393442623,
+      "grad_norm": 14.882206916809082,
+      "learning_rate": 1.8792575071152565e-05,
+      "loss": 1.2162,
+      "step": 895
+    },
+    {
+      "epoch": 0.18360655737704917,
+      "grad_norm": 4.606672763824463,
+      "learning_rate": 1.8789411318522342e-05,
+      "loss": 0.9339,
+      "step": 896
+    },
+    {
+      "epoch": 0.18381147540983606,
+      "grad_norm": 4.507289409637451,
+      "learning_rate": 1.8786243693435968e-05,
+      "loss": 1.0832,
+      "step": 897
+    },
+    {
+      "epoch": 0.18401639344262294,
+      "grad_norm": 2.9086811542510986,
+      "learning_rate": 1.878307219728903e-05,
+      "loss": 1.1505,
+      "step": 898
+    },
+    {
+      "epoch": 0.18422131147540985,
+      "grad_norm": 4.466919898986816,
+      "learning_rate": 1.877989683147885e-05,
+      "loss": 1.0416,
+      "step": 899
+    },
+    {
+      "epoch": 0.18442622950819673,
+      "grad_norm": 7.517127990722656,
+      "learning_rate": 1.8776717597404422e-05,
+      "loss": 1.1137,
+      "step": 900
+    },
+    {
+      "epoch": 0.18463114754098361,
+      "grad_norm": 7.294081687927246,
+      "learning_rate": 1.8773534496466463e-05,
+      "loss": 1.1474,
+      "step": 901
+    },
+    {
+      "epoch": 0.1848360655737705,
+      "grad_norm": 4.6593523025512695,
+      "learning_rate": 1.8770347530067395e-05,
+      "loss": 1.393,
+      "step": 902
+    },
+    {
+      "epoch": 0.18504098360655738,
+      "grad_norm": 4.931035995483398,
+      "learning_rate": 1.8767156699611334e-05,
+      "loss": 1.2091,
+      "step": 903
+    },
+    {
+      "epoch": 0.18524590163934426,
+      "grad_norm": 4.628993988037109,
+      "learning_rate": 1.87639620065041e-05,
+      "loss": 0.9824,
+      "step": 904
+    },
+    {
+      "epoch": 0.18545081967213115,
+      "grad_norm": 3.6848201751708984,
+      "learning_rate": 1.876076345215322e-05,
+      "loss": 1.0909,
+      "step": 905
+    },
+    {
+      "epoch": 0.18565573770491803,
+      "grad_norm": 9.283903121948242,
+      "learning_rate": 1.875756103796792e-05,
+      "loss": 1.3532,
+      "step": 906
+    },
+    {
+      "epoch": 0.1858606557377049,
+      "grad_norm": 4.0193939208984375,
+      "learning_rate": 1.8754354765359132e-05,
+      "loss": 0.7952,
+      "step": 907
+    },
+    {
+      "epoch": 0.1860655737704918,
+      "grad_norm": 4.324032306671143,
+      "learning_rate": 1.875114463573947e-05,
+      "loss": 0.8929,
+      "step": 908
+    },
+    {
+      "epoch": 0.18627049180327868,
+      "grad_norm": 4.387557506561279,
+      "learning_rate": 1.8747930650523264e-05,
+      "loss": 0.9764,
+      "step": 909
+    },
+    {
+      "epoch": 0.1864754098360656,
+      "grad_norm": 3.8864264488220215,
+      "learning_rate": 1.8744712811126536e-05,
+      "loss": 1.1177,
+      "step": 910
+    },
+    {
+      "epoch": 0.18668032786885247,
+      "grad_norm": 3.459482431411743,
+      "learning_rate": 1.8741491118967013e-05,
+      "loss": 1.0199,
+      "step": 911
+    },
+    {
+      "epoch": 0.18688524590163935,
+      "grad_norm": 6.227849960327148,
+      "learning_rate": 1.8738265575464112e-05,
+      "loss": 1.0805,
+      "step": 912
+    },
+    {
+      "epoch": 0.18709016393442623,
+      "grad_norm": 8.29771614074707,
+      "learning_rate": 1.873503618203895e-05,
+      "loss": 1.0716,
+      "step": 913
+    },
+    {
+      "epoch": 0.18729508196721312,
+      "grad_norm": 5.90207052230835,
+      "learning_rate": 1.8731802940114334e-05,
+      "loss": 1.2958,
+      "step": 914
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 9.033169746398926,
+      "learning_rate": 1.8728565851114782e-05,
+      "loss": 1.0476,
+      "step": 915
+    },
+    {
+      "epoch": 0.18770491803278688,
+      "grad_norm": 5.49434232711792,
+      "learning_rate": 1.872532491646649e-05,
+      "loss": 0.8182,
+      "step": 916
+    },
+    {
+      "epoch": 0.18790983606557377,
+      "grad_norm": 6.806762218475342,
+      "learning_rate": 1.8722080137597355e-05,
+      "loss": 1.2385,
+      "step": 917
+    },
+    {
+      "epoch": 0.18811475409836065,
+      "grad_norm": 5.79287576675415,
+      "learning_rate": 1.8718831515936973e-05,
+      "loss": 1.193,
+      "step": 918
+    },
+    {
+      "epoch": 0.18831967213114753,
+      "grad_norm": 5.1449432373046875,
+      "learning_rate": 1.8715579052916625e-05,
+      "loss": 1.1021,
+      "step": 919
+    },
+    {
+      "epoch": 0.1885245901639344,
+      "grad_norm": 3.8027307987213135,
+      "learning_rate": 1.8712322749969288e-05,
+      "loss": 0.8738,
+      "step": 920
+    },
+    {
+      "epoch": 0.18872950819672132,
+      "grad_norm": 4.78309965133667,
+      "learning_rate": 1.8709062608529628e-05,
+      "loss": 1.0409,
+      "step": 921
+    },
+    {
+      "epoch": 0.1889344262295082,
+      "grad_norm": 4.148694038391113,
+      "learning_rate": 1.870579863003401e-05,
+      "loss": 1.0385,
+      "step": 922
+    },
+    {
+      "epoch": 0.1891393442622951,
+      "grad_norm": 5.1571526527404785,
+      "learning_rate": 1.870253081592048e-05,
+      "loss": 1.3171,
+      "step": 923
+    },
+    {
+      "epoch": 0.18934426229508197,
+      "grad_norm": 5.088181972503662,
+      "learning_rate": 1.869925916762878e-05,
+      "loss": 1.176,
+      "step": 924
+    },
+    {
+      "epoch": 0.18954918032786885,
+      "grad_norm": 5.626553535461426,
+      "learning_rate": 1.8695983686600335e-05,
+      "loss": 1.0919,
+      "step": 925
+    },
+    {
+      "epoch": 0.18975409836065574,
+      "grad_norm": 3.7373616695404053,
+      "learning_rate": 1.8692704374278268e-05,
+      "loss": 1.118,
+      "step": 926
+    },
+    {
+      "epoch": 0.18995901639344262,
+      "grad_norm": 5.423557758331299,
+      "learning_rate": 1.8689421232107382e-05,
+      "loss": 0.9634,
+      "step": 927
+    },
+    {
+      "epoch": 0.1901639344262295,
+      "grad_norm": 5.080060958862305,
+      "learning_rate": 1.8686134261534176e-05,
+      "loss": 1.1579,
+      "step": 928
+    },
+    {
+      "epoch": 0.19036885245901639,
+      "grad_norm": 2.6996910572052,
+      "learning_rate": 1.868284346400682e-05,
+      "loss": 0.9437,
+      "step": 929
+    },
+    {
+      "epoch": 0.19057377049180327,
+      "grad_norm": 4.541398048400879,
+      "learning_rate": 1.8679548840975185e-05,
+      "loss": 1.0877,
+      "step": 930
+    },
+    {
+      "epoch": 0.19077868852459015,
+      "grad_norm": 4.3583083152771,
+      "learning_rate": 1.8676250393890828e-05,
+      "loss": 1.192,
+      "step": 931
+    },
+    {
+      "epoch": 0.19098360655737706,
+      "grad_norm": 4.629337787628174,
+      "learning_rate": 1.8672948124206973e-05,
+      "loss": 1.2183,
+      "step": 932
+    },
+    {
+      "epoch": 0.19118852459016394,
+      "grad_norm": 5.796934604644775,
+      "learning_rate": 1.866964203337855e-05,
+      "loss": 1.2396,
+      "step": 933
+    },
+    {
+      "epoch": 0.19139344262295083,
+      "grad_norm": 5.7569427490234375,
+      "learning_rate": 1.866633212286216e-05,
+      "loss": 1.0984,
+      "step": 934
+    },
+    {
+      "epoch": 0.1915983606557377,
+      "grad_norm": 6.626437187194824,
+      "learning_rate": 1.8663018394116092e-05,
+      "loss": 0.8947,
+      "step": 935
+    },
+    {
+      "epoch": 0.1918032786885246,
+      "grad_norm": 3.421056032180786,
+      "learning_rate": 1.8659700848600314e-05,
+      "loss": 1.0812,
+      "step": 936
+    },
+    {
+      "epoch": 0.19200819672131147,
+      "grad_norm": 3.9736626148223877,
+      "learning_rate": 1.8656379487776478e-05,
+      "loss": 0.9621,
+      "step": 937
+    },
+    {
+      "epoch": 0.19221311475409836,
+      "grad_norm": 4.799187660217285,
+      "learning_rate": 1.865305431310791e-05,
+      "loss": 0.9606,
+      "step": 938
+    },
+    {
+      "epoch": 0.19241803278688524,
+      "grad_norm": 3.855839252471924,
+      "learning_rate": 1.8649725326059628e-05,
+      "loss": 1.1451,
+      "step": 939
+    },
+    {
+      "epoch": 0.19262295081967212,
+      "grad_norm": 7.347762107849121,
+      "learning_rate": 1.864639252809832e-05,
+      "loss": 1.172,
+      "step": 940
+    },
+    {
+      "epoch": 0.192827868852459,
+      "grad_norm": 3.9868974685668945,
+      "learning_rate": 1.864305592069236e-05,
+      "loss": 0.9056,
+      "step": 941
+    },
+    {
+      "epoch": 0.1930327868852459,
+      "grad_norm": 4.6992106437683105,
+      "learning_rate": 1.8639715505311794e-05,
+      "loss": 1.1692,
+      "step": 942
+    },
+    {
+      "epoch": 0.1932377049180328,
+      "grad_norm": 7.615385055541992,
+      "learning_rate": 1.8636371283428355e-05,
+      "loss": 1.2867,
+      "step": 943
+    },
+    {
+      "epoch": 0.19344262295081968,
+      "grad_norm": 4.132473945617676,
+      "learning_rate": 1.8633023256515436e-05,
+      "loss": 1.1275,
+      "step": 944
+    },
+    {
+      "epoch": 0.19364754098360656,
+      "grad_norm": 6.131743907928467,
+      "learning_rate": 1.8629671426048126e-05,
+      "loss": 1.1506,
+      "step": 945
+    },
+    {
+      "epoch": 0.19385245901639345,
+      "grad_norm": 7.8305768966674805,
+      "learning_rate": 1.8626315793503175e-05,
+      "loss": 1.1478,
+      "step": 946
+    },
+    {
+      "epoch": 0.19405737704918033,
+      "grad_norm": 4.490309238433838,
+      "learning_rate": 1.8622956360359025e-05,
+      "loss": 1.1731,
+      "step": 947
+    },
+    {
+      "epoch": 0.1942622950819672,
+      "grad_norm": 6.566570281982422,
+      "learning_rate": 1.8619593128095767e-05,
+      "loss": 0.7689,
+      "step": 948
+    },
+    {
+      "epoch": 0.1944672131147541,
+      "grad_norm": 4.5851240158081055,
+      "learning_rate": 1.8616226098195192e-05,
+      "loss": 1.0727,
+      "step": 949
+    },
+    {
+      "epoch": 0.19467213114754098,
+      "grad_norm": 4.646297931671143,
+      "learning_rate": 1.861285527214075e-05,
+      "loss": 0.9511,
+      "step": 950
+    },
+    {
+      "epoch": 0.19487704918032786,
+      "grad_norm": 3.1957550048828125,
+      "learning_rate": 1.860948065141756e-05,
+      "loss": 0.9663,
+      "step": 951
+    },
+    {
+      "epoch": 0.19508196721311474,
+      "grad_norm": 4.267082214355469,
+      "learning_rate": 1.8606102237512426e-05,
+      "loss": 1.1096,
+      "step": 952
+    },
+    {
+      "epoch": 0.19528688524590163,
+      "grad_norm": 5.666517734527588,
+      "learning_rate": 1.860272003191381e-05,
+      "loss": 1.1606,
+      "step": 953
+    },
+    {
+      "epoch": 0.19549180327868854,
+      "grad_norm": 3.5661869049072266,
+      "learning_rate": 1.859933403611186e-05,
+      "loss": 1.2505,
+      "step": 954
+    },
+    {
+      "epoch": 0.19569672131147542,
+      "grad_norm": 5.89561653137207,
+      "learning_rate": 1.8595944251598377e-05,
+      "loss": 1.1352,
+      "step": 955
+    },
+    {
+      "epoch": 0.1959016393442623,
+      "grad_norm": 4.925987243652344,
+      "learning_rate": 1.8592550679866843e-05,
+      "loss": 1.2031,
+      "step": 956
+    },
+    {
+      "epoch": 0.19610655737704918,
+      "grad_norm": 5.326219081878662,
+      "learning_rate": 1.8589153322412405e-05,
+      "loss": 1.1914,
+      "step": 957
+    },
+    {
+      "epoch": 0.19631147540983607,
+      "grad_norm": 5.051677703857422,
+      "learning_rate": 1.858575218073187e-05,
+      "loss": 1.0164,
+      "step": 958
+    },
+    {
+      "epoch": 0.19651639344262295,
+      "grad_norm": 6.946462631225586,
+      "learning_rate": 1.858234725632373e-05,
+      "loss": 0.8428,
+      "step": 959
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 2.837131977081299,
+      "learning_rate": 1.8578938550688123e-05,
+      "loss": 1.1085,
+      "step": 960
+    },
+    {
+      "epoch": 0.19692622950819672,
+      "grad_norm": 7.7736897468566895,
+      "learning_rate": 1.8575526065326874e-05,
+      "loss": 0.7997,
+      "step": 961
+    },
+    {
+      "epoch": 0.1971311475409836,
+      "grad_norm": 4.748393535614014,
+      "learning_rate": 1.8572109801743453e-05,
+      "loss": 1.1032,
+      "step": 962
+    },
+    {
+      "epoch": 0.19733606557377048,
+      "grad_norm": 3.3190906047821045,
+      "learning_rate": 1.856868976144301e-05,
+      "loss": 0.9178,
+      "step": 963
+    },
+    {
+      "epoch": 0.19754098360655736,
+      "grad_norm": 3.955166816711426,
+      "learning_rate": 1.8565265945932352e-05,
+      "loss": 1.2643,
+      "step": 964
+    },
+    {
+      "epoch": 0.19774590163934427,
+      "grad_norm": 3.7708942890167236,
+      "learning_rate": 1.856183835671995e-05,
+      "loss": 1.1829,
+      "step": 965
+    },
+    {
+      "epoch": 0.19795081967213116,
+      "grad_norm": 3.8316152095794678,
+      "learning_rate": 1.8558406995315932e-05,
+      "loss": 1.2665,
+      "step": 966
+    },
+    {
+      "epoch": 0.19815573770491804,
+      "grad_norm": 6.7828216552734375,
+      "learning_rate": 1.855497186323211e-05,
+      "loss": 1.3665,
+      "step": 967
+    },
+    {
+      "epoch": 0.19836065573770492,
+      "grad_norm": 5.175777912139893,
+      "learning_rate": 1.8551532961981923e-05,
+      "loss": 0.9111,
+      "step": 968
+    },
+    {
+      "epoch": 0.1985655737704918,
+      "grad_norm": 5.224827289581299,
+      "learning_rate": 1.85480902930805e-05,
+      "loss": 1.3265,
+      "step": 969
+    },
+    {
+      "epoch": 0.1987704918032787,
+      "grad_norm": 4.920624732971191,
+      "learning_rate": 1.8544643858044616e-05,
+      "loss": 1.1486,
+      "step": 970
+    },
+    {
+      "epoch": 0.19897540983606557,
+      "grad_norm": 4.034660339355469,
+      "learning_rate": 1.8541193658392705e-05,
+      "loss": 1.0108,
+      "step": 971
+    },
+    {
+      "epoch": 0.19918032786885245,
+      "grad_norm": 4.190369129180908,
+      "learning_rate": 1.853773969564487e-05,
+      "loss": 0.9691,
+      "step": 972
+    },
+    {
+      "epoch": 0.19938524590163934,
+      "grad_norm": 2.8341872692108154,
+      "learning_rate": 1.8534281971322856e-05,
+      "loss": 0.8292,
+      "step": 973
+    },
+    {
+      "epoch": 0.19959016393442622,
+      "grad_norm": 6.46570348739624,
+      "learning_rate": 1.853082048695008e-05,
+      "loss": 0.9376,
+      "step": 974
+    },
+    {
+      "epoch": 0.19979508196721313,
+      "grad_norm": 3.4102466106414795,
+      "learning_rate": 1.8527355244051607e-05,
+      "loss": 1.0734,
+      "step": 975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.306502342224121,
+      "learning_rate": 1.8523886244154163e-05,
+      "loss": 1.3614,
+      "step": 976
+    },
+    {
+      "epoch": 0.2002049180327869,
+      "grad_norm": 5.575007438659668,
+      "learning_rate": 1.8520413488786127e-05,
+      "loss": 0.9646,
+      "step": 977
+    },
+    {
+      "epoch": 0.20040983606557378,
+      "grad_norm": 4.9284234046936035,
+      "learning_rate": 1.8516936979477527e-05,
+      "loss": 0.9399,
+      "step": 978
+    },
+    {
+      "epoch": 0.20061475409836066,
+      "grad_norm": 3.894876480102539,
+      "learning_rate": 1.8513456717760057e-05,
+      "loss": 1.1448,
+      "step": 979
+    },
+    {
+      "epoch": 0.20081967213114754,
+      "grad_norm": 5.445644855499268,
+      "learning_rate": 1.8509972705167052e-05,
+      "loss": 1.0533,
+      "step": 980
+    },
+    {
+      "epoch": 0.20102459016393442,
+      "grad_norm": 3.8068034648895264,
+      "learning_rate": 1.850648494323351e-05,
+      "loss": 1.177,
+      "step": 981
+    },
+    {
+      "epoch": 0.2012295081967213,
+      "grad_norm": 5.057478427886963,
+      "learning_rate": 1.8502993433496073e-05,
+      "loss": 0.9223,
+      "step": 982
+    },
+    {
+      "epoch": 0.2014344262295082,
+      "grad_norm": 5.586034774780273,
+      "learning_rate": 1.8499498177493036e-05,
+      "loss": 1.3301,
+      "step": 983
+    },
+    {
+      "epoch": 0.20163934426229507,
+      "grad_norm": 5.663778305053711,
+      "learning_rate": 1.849599917676435e-05,
+      "loss": 1.0617,
+      "step": 984
+    },
+    {
+      "epoch": 0.20184426229508196,
+      "grad_norm": 10.671714782714844,
+      "learning_rate": 1.849249643285161e-05,
+      "loss": 1.1394,
+      "step": 985
+    },
+    {
+      "epoch": 0.20204918032786887,
+      "grad_norm": 3.895968198776245,
+      "learning_rate": 1.848898994729806e-05,
+      "loss": 1.0295,
+      "step": 986
+    },
+    {
+      "epoch": 0.20225409836065575,
+      "grad_norm": 3.6684226989746094,
+      "learning_rate": 1.8485479721648595e-05,
+      "loss": 0.8435,
+      "step": 987
+    },
+    {
+      "epoch": 0.20245901639344263,
+      "grad_norm": 5.869389057159424,
+      "learning_rate": 1.8481965757449755e-05,
+      "loss": 0.8776,
+      "step": 988
+    },
+    {
+      "epoch": 0.20266393442622951,
+      "grad_norm": 3.3255343437194824,
+      "learning_rate": 1.8478448056249737e-05,
+      "loss": 0.8388,
+      "step": 989
+    },
+    {
+      "epoch": 0.2028688524590164,
+      "grad_norm": 5.290645599365234,
+      "learning_rate": 1.847492661959837e-05,
+      "loss": 0.9912,
+      "step": 990
+    },
+    {
+      "epoch": 0.20307377049180328,
+      "grad_norm": 4.239506244659424,
+      "learning_rate": 1.8471401449047138e-05,
+      "loss": 1.2863,
+      "step": 991
+    },
+    {
+      "epoch": 0.20327868852459016,
+      "grad_norm": 3.7198426723480225,
+      "learning_rate": 1.8467872546149165e-05,
+      "loss": 0.9189,
+      "step": 992
+    },
+    {
+      "epoch": 0.20348360655737704,
+      "grad_norm": 5.304177761077881,
+      "learning_rate": 1.846433991245922e-05,
+      "loss": 1.2152,
+      "step": 993
+    },
+    {
+      "epoch": 0.20368852459016393,
+      "grad_norm": 5.3899617195129395,
+      "learning_rate": 1.846080354953373e-05,
+      "loss": 1.0024,
+      "step": 994
+    },
+    {
+      "epoch": 0.2038934426229508,
+      "grad_norm": 3.7625017166137695,
+      "learning_rate": 1.8457263458930737e-05,
+      "loss": 1.1883,
+      "step": 995
+    },
+    {
+      "epoch": 0.2040983606557377,
+      "grad_norm": 2.850883960723877,
+      "learning_rate": 1.845371964220995e-05,
+      "loss": 1.0271,
+      "step": 996
+    },
+    {
+      "epoch": 0.2043032786885246,
+      "grad_norm": 6.35784912109375,
+      "learning_rate": 1.8450172100932712e-05,
+      "loss": 0.9596,
+      "step": 997
+    },
+    {
+      "epoch": 0.20450819672131149,
+      "grad_norm": 4.048913955688477,
+      "learning_rate": 1.8446620836662e-05,
+      "loss": 1.3884,
+      "step": 998
+    },
+    {
+      "epoch": 0.20471311475409837,
+      "grad_norm": 4.099578857421875,
+      "learning_rate": 1.8443065850962442e-05,
+      "loss": 0.9973,
+      "step": 999
+    },
+    {
+      "epoch": 0.20491803278688525,
+      "grad_norm": 3.3128273487091064,
+      "learning_rate": 1.8439507145400294e-05,
+      "loss": 1.0933,
+      "step": 1000
+    },
+    {
+      "epoch": 0.20512295081967213,
+      "grad_norm": 4.017665386199951,
+      "learning_rate": 1.8435944721543466e-05,
+      "loss": 1.0303,
+      "step": 1001
+    },
+    {
+      "epoch": 0.20532786885245902,
+      "grad_norm": 3.2721312046051025,
+      "learning_rate": 1.843237858096149e-05,
+      "loss": 0.9468,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2055327868852459,
+      "grad_norm": 5.009561061859131,
+      "learning_rate": 1.8428808725225548e-05,
+      "loss": 0.9595,
+      "step": 1003
+    },
+    {
+      "epoch": 0.20573770491803278,
+      "grad_norm": 7.730933666229248,
+      "learning_rate": 1.8425235155908455e-05,
+      "loss": 0.9657,
+      "step": 1004
+    },
+    {
+      "epoch": 0.20594262295081966,
+      "grad_norm": 7.169773101806641,
+      "learning_rate": 1.8421657874584658e-05,
+      "loss": 1.122,
+      "step": 1005
+    },
+    {
+      "epoch": 0.20614754098360655,
+      "grad_norm": 5.160264015197754,
+      "learning_rate": 1.8418076882830244e-05,
+      "loss": 1.1675,
+      "step": 1006
+    },
+    {
+      "epoch": 0.20635245901639343,
+      "grad_norm": 3.8367984294891357,
+      "learning_rate": 1.8414492182222933e-05,
+      "loss": 1.1295,
+      "step": 1007
+    },
+    {
+      "epoch": 0.20655737704918034,
+      "grad_norm": 4.597427845001221,
+      "learning_rate": 1.841090377434208e-05,
+      "loss": 1.1061,
+      "step": 1008
+    },
+    {
+      "epoch": 0.20676229508196722,
+      "grad_norm": 5.315457344055176,
+      "learning_rate": 1.840731166076867e-05,
+      "loss": 0.9015,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2069672131147541,
+      "grad_norm": 4.302783489227295,
+      "learning_rate": 1.8403715843085325e-05,
+      "loss": 1.1394,
+      "step": 1010
+    },
+    {
+      "epoch": 0.207172131147541,
+      "grad_norm": 3.900545835494995,
+      "learning_rate": 1.8400116322876302e-05,
+      "loss": 1.1426,
+      "step": 1011
+    },
+    {
+      "epoch": 0.20737704918032787,
+      "grad_norm": 3.67492938041687,
+      "learning_rate": 1.839651310172748e-05,
+      "loss": 1.2871,
+      "step": 1012
+    },
+    {
+      "epoch": 0.20758196721311475,
+      "grad_norm": 5.586658000946045,
+      "learning_rate": 1.8392906181226376e-05,
+      "loss": 1.0318,
+      "step": 1013
+    },
+    {
+      "epoch": 0.20778688524590164,
+      "grad_norm": 3.6502573490142822,
+      "learning_rate": 1.8389295562962132e-05,
+      "loss": 1.3096,
+      "step": 1014
+    },
+    {
+      "epoch": 0.20799180327868852,
+      "grad_norm": 8.218432426452637,
+      "learning_rate": 1.838568124852552e-05,
+      "loss": 1.1616,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2081967213114754,
+      "grad_norm": 4.253335475921631,
+      "learning_rate": 1.8382063239508944e-05,
+      "loss": 0.8159,
+      "step": 1016
+    },
+    {
+      "epoch": 0.20840163934426228,
+      "grad_norm": 4.5475358963012695,
+      "learning_rate": 1.837844153750643e-05,
+      "loss": 1.1898,
+      "step": 1017
+    },
+    {
+      "epoch": 0.20860655737704917,
+      "grad_norm": 4.777431488037109,
+      "learning_rate": 1.837481614411364e-05,
+      "loss": 1.0007,
+      "step": 1018
+    },
+    {
+      "epoch": 0.20881147540983608,
+      "grad_norm": 3.576352834701538,
+      "learning_rate": 1.8371187060927855e-05,
+      "loss": 1.0648,
+      "step": 1019
+    },
+    {
+      "epoch": 0.20901639344262296,
+      "grad_norm": 4.7599687576293945,
+      "learning_rate": 1.8367554289547985e-05,
+      "loss": 1.2267,
+      "step": 1020
+    },
+    {
+      "epoch": 0.20922131147540984,
+      "grad_norm": 5.029023170471191,
+      "learning_rate": 1.836391783157456e-05,
+      "loss": 1.1115,
+      "step": 1021
+    },
+    {
+      "epoch": 0.20942622950819673,
+      "grad_norm": 10.017075538635254,
+      "learning_rate": 1.836027768860974e-05,
+      "loss": 1.3141,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2096311475409836,
+      "grad_norm": 5.580081462860107,
+      "learning_rate": 1.8356633862257304e-05,
+      "loss": 0.7,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2098360655737705,
+      "grad_norm": 4.607186794281006,
+      "learning_rate": 1.8352986354122663e-05,
+      "loss": 0.7837,
+      "step": 1024
+    },
+    {
+      "epoch": 0.21004098360655737,
+      "grad_norm": 3.5278444290161133,
+      "learning_rate": 1.8349335165812835e-05,
+      "loss": 0.892,
+      "step": 1025
+    },
+    {
+      "epoch": 0.21024590163934426,
+      "grad_norm": 4.4638991355896,
+      "learning_rate": 1.8345680298936472e-05,
+      "loss": 1.2302,
+      "step": 1026
+    },
+    {
+      "epoch": 0.21045081967213114,
+      "grad_norm": 3.2856898307800293,
+      "learning_rate": 1.834202175510384e-05,
+      "loss": 1.1203,
+      "step": 1027
+    },
+    {
+      "epoch": 0.21065573770491802,
+      "grad_norm": 3.161489248275757,
+      "learning_rate": 1.8338359535926834e-05,
+      "loss": 0.8722,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2108606557377049,
+      "grad_norm": 4.518817901611328,
+      "learning_rate": 1.833469364301895e-05,
+      "loss": 1.3453,
+      "step": 1029
+    },
+    {
+      "epoch": 0.21106557377049182,
+      "grad_norm": 4.039306163787842,
+      "learning_rate": 1.8331024077995326e-05,
+      "loss": 1.0224,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2112704918032787,
+      "grad_norm": 4.246673107147217,
+      "learning_rate": 1.83273508424727e-05,
+      "loss": 1.1341,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21147540983606558,
+      "grad_norm": 3.9591758251190186,
+      "learning_rate": 1.8323673938069434e-05,
+      "loss": 0.9456,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21168032786885246,
+      "grad_norm": 3.5740773677825928,
+      "learning_rate": 1.831999336640551e-05,
+      "loss": 1.012,
+      "step": 1033
+    },
+    {
+      "epoch": 0.21188524590163935,
+      "grad_norm": 5.737273216247559,
+      "learning_rate": 1.8316309129102513e-05,
+      "loss": 0.9107,
+      "step": 1034
+    },
+    {
+      "epoch": 0.21209016393442623,
+      "grad_norm": 3.6804699897766113,
+      "learning_rate": 1.831262122778366e-05,
+      "loss": 0.9774,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2122950819672131,
+      "grad_norm": 3.5921149253845215,
+      "learning_rate": 1.830892966407377e-05,
+      "loss": 0.8399,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 3.37103533744812,
+      "learning_rate": 1.8305234439599276e-05,
+      "loss": 0.6994,
+      "step": 1037
+    },
+    {
+      "epoch": 0.21270491803278688,
+      "grad_norm": 10.140278816223145,
+      "learning_rate": 1.8301535555988235e-05,
+      "loss": 1.1235,
+      "step": 1038
+    },
+    {
+      "epoch": 0.21290983606557376,
+      "grad_norm": 7.858144760131836,
+      "learning_rate": 1.8297833014870307e-05,
+      "loss": 0.9847,
+      "step": 1039
+    },
+    {
+      "epoch": 0.21311475409836064,
+      "grad_norm": 6.460100173950195,
+      "learning_rate": 1.8294126817876758e-05,
+      "loss": 1.4031,
+      "step": 1040
+    },
+    {
+      "epoch": 0.21331967213114755,
+      "grad_norm": 5.704457759857178,
+      "learning_rate": 1.8290416966640475e-05,
+      "loss": 0.9448,
+      "step": 1041
+    },
+    {
+      "epoch": 0.21352459016393444,
+      "grad_norm": 7.923317909240723,
+      "learning_rate": 1.8286703462795957e-05,
+      "loss": 1.038,
+      "step": 1042
+    },
+    {
+      "epoch": 0.21372950819672132,
+      "grad_norm": 6.529424667358398,
+      "learning_rate": 1.82829863079793e-05,
+      "loss": 0.981,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2139344262295082,
+      "grad_norm": 3.221997022628784,
+      "learning_rate": 1.8279265503828215e-05,
+      "loss": 1.1681,
+      "step": 1044
+    },
+    {
+      "epoch": 0.21413934426229508,
+      "grad_norm": 5.993236541748047,
+      "learning_rate": 1.827554105198203e-05,
+      "loss": 1.059,
+      "step": 1045
+    },
+    {
+      "epoch": 0.21434426229508197,
+      "grad_norm": 4.100046634674072,
+      "learning_rate": 1.8271812954081662e-05,
+      "loss": 0.813,
+      "step": 1046
+    },
+    {
+      "epoch": 0.21454918032786885,
+      "grad_norm": 3.4920547008514404,
+      "learning_rate": 1.8268081211769645e-05,
+      "loss": 1.1521,
+      "step": 1047
+    },
+    {
+      "epoch": 0.21475409836065573,
+      "grad_norm": 3.6895933151245117,
+      "learning_rate": 1.8264345826690123e-05,
+      "loss": 1.1912,
+      "step": 1048
+    },
+    {
+      "epoch": 0.21495901639344261,
+      "grad_norm": 4.080672264099121,
+      "learning_rate": 1.826060680048883e-05,
+      "loss": 1.2495,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2151639344262295,
+      "grad_norm": 3.503981351852417,
+      "learning_rate": 1.8256864134813126e-05,
+      "loss": 1.3822,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21536885245901638,
+      "grad_norm": 6.306840419769287,
+      "learning_rate": 1.8253117831311954e-05,
+      "loss": 1.1445,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2155737704918033,
+      "grad_norm": 4.832455158233643,
+      "learning_rate": 1.8249367891635864e-05,
+      "loss": 1.0318,
+      "step": 1052
+    },
+    {
+      "epoch": 0.21577868852459017,
+      "grad_norm": 5.371854305267334,
+      "learning_rate": 1.824561431743702e-05,
+      "loss": 1.1165,
+      "step": 1053
+    },
+    {
+      "epoch": 0.21598360655737706,
+      "grad_norm": 6.135509490966797,
+      "learning_rate": 1.824185711036918e-05,
+      "loss": 1.2094,
+      "step": 1054
+    },
+    {
+      "epoch": 0.21618852459016394,
+      "grad_norm": 5.590333938598633,
+      "learning_rate": 1.8238096272087694e-05,
+      "loss": 1.3509,
+      "step": 1055
+    },
+    {
+      "epoch": 0.21639344262295082,
+      "grad_norm": 5.3369364738464355,
+      "learning_rate": 1.823433180424952e-05,
+      "loss": 0.8242,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2165983606557377,
+      "grad_norm": 11.196516990661621,
+      "learning_rate": 1.823056370851322e-05,
+      "loss": 1.3037,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2168032786885246,
+      "grad_norm": 4.091281890869141,
+      "learning_rate": 1.822679198653895e-05,
+      "loss": 1.0546,
+      "step": 1058
+    },
+    {
+      "epoch": 0.21700819672131147,
+      "grad_norm": 3.980672836303711,
+      "learning_rate": 1.8223016639988458e-05,
+      "loss": 1.1307,
+      "step": 1059
+    },
+    {
+      "epoch": 0.21721311475409835,
+      "grad_norm": 3.9089958667755127,
+      "learning_rate": 1.8219237670525092e-05,
+      "loss": 0.8005,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21741803278688523,
+      "grad_norm": 4.796226978302002,
+      "learning_rate": 1.8215455079813806e-05,
+      "loss": 1.0869,
+      "step": 1061
+    },
+    {
+      "epoch": 0.21762295081967212,
+      "grad_norm": 5.142142295837402,
+      "learning_rate": 1.8211668869521136e-05,
+      "loss": 1.1137,
+      "step": 1062
+    },
+    {
+      "epoch": 0.21782786885245903,
+      "grad_norm": 4.119559288024902,
+      "learning_rate": 1.8207879041315216e-05,
+      "loss": 1.2862,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2180327868852459,
+      "grad_norm": 5.553428649902344,
+      "learning_rate": 1.8204085596865778e-05,
+      "loss": 0.8991,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2182377049180328,
+      "grad_norm": 4.018271446228027,
+      "learning_rate": 1.8200288537844146e-05,
+      "loss": 0.9669,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21844262295081968,
+      "grad_norm": 4.089879035949707,
+      "learning_rate": 1.8196487865923234e-05,
+      "loss": 1.0088,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21864754098360656,
+      "grad_norm": 29.53445053100586,
+      "learning_rate": 1.8192683582777553e-05,
+      "loss": 1.1463,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21885245901639344,
+      "grad_norm": 11.05590534210205,
+      "learning_rate": 1.8188875690083197e-05,
+      "loss": 1.1666,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21905737704918032,
+      "grad_norm": 7.182314395904541,
+      "learning_rate": 1.8185064189517857e-05,
+      "loss": 0.8651,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2192622950819672,
+      "grad_norm": 3.443282127380371,
+      "learning_rate": 1.8181249082760813e-05,
+      "loss": 0.9544,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2194672131147541,
+      "grad_norm": 6.974148750305176,
+      "learning_rate": 1.817743037149293e-05,
+      "loss": 1.1762,
+      "step": 1071
+    },
+    {
+      "epoch": 0.21967213114754097,
+      "grad_norm": 4.781474590301514,
+      "learning_rate": 1.8173608057396667e-05,
+      "loss": 0.9528,
+      "step": 1072
+    },
+    {
+      "epoch": 0.21987704918032788,
+      "grad_norm": 5.563875198364258,
+      "learning_rate": 1.816978214215606e-05,
+      "loss": 1.2939,
+      "step": 1073
+    },
+    {
+      "epoch": 0.22008196721311477,
+      "grad_norm": 4.307204246520996,
+      "learning_rate": 1.8165952627456746e-05,
+      "loss": 0.8829,
+      "step": 1074
+    },
+    {
+      "epoch": 0.22028688524590165,
+      "grad_norm": 3.4520773887634277,
+      "learning_rate": 1.8162119514985934e-05,
+      "loss": 0.8456,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22049180327868853,
+      "grad_norm": 4.218603610992432,
+      "learning_rate": 1.8158282806432428e-05,
+      "loss": 1.184,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2206967213114754,
+      "grad_norm": 3.6401944160461426,
+      "learning_rate": 1.8154442503486608e-05,
+      "loss": 1.2513,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2209016393442623,
+      "grad_norm": 4.284875392913818,
+      "learning_rate": 1.8150598607840448e-05,
+      "loss": 1.1927,
+      "step": 1078
+    },
+    {
+      "epoch": 0.22110655737704918,
+      "grad_norm": 6.416600704193115,
+      "learning_rate": 1.8146751121187497e-05,
+      "loss": 1.0943,
+      "step": 1079
+    },
+    {
+      "epoch": 0.22131147540983606,
+      "grad_norm": 7.961711406707764,
+      "learning_rate": 1.8142900045222888e-05,
+      "loss": 1.0829,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22151639344262294,
+      "grad_norm": 4.150201797485352,
+      "learning_rate": 1.8139045381643332e-05,
+      "loss": 1.084,
+      "step": 1081
+    },
+    {
+      "epoch": 0.22172131147540983,
+      "grad_norm": 3.901859760284424,
+      "learning_rate": 1.813518713214713e-05,
+      "loss": 1.1772,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2219262295081967,
+      "grad_norm": 2.397149085998535,
+      "learning_rate": 1.813132529843415e-05,
+      "loss": 0.8969,
+      "step": 1083
+    },
+    {
+      "epoch": 0.22213114754098362,
+      "grad_norm": 6.706818580627441,
+      "learning_rate": 1.812745988220586e-05,
+      "loss": 0.9278,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2223360655737705,
+      "grad_norm": 3.903796434402466,
+      "learning_rate": 1.812359088516528e-05,
+      "loss": 1.2008,
+      "step": 1085
+    },
+    {
+      "epoch": 0.22254098360655739,
+      "grad_norm": 5.661761283874512,
+      "learning_rate": 1.811971830901702e-05,
+      "loss": 1.1924,
+      "step": 1086
+    },
+    {
+      "epoch": 0.22274590163934427,
+      "grad_norm": 4.0636396408081055,
+      "learning_rate": 1.8115842155467275e-05,
+      "loss": 1.0178,
+      "step": 1087
+    },
+    {
+      "epoch": 0.22295081967213115,
+      "grad_norm": 7.642810344696045,
+      "learning_rate": 1.8111962426223802e-05,
+      "loss": 0.9631,
+      "step": 1088
+    },
+    {
+      "epoch": 0.22315573770491803,
+      "grad_norm": 13.062255859375,
+      "learning_rate": 1.8108079122995946e-05,
+      "loss": 1.3786,
+      "step": 1089
+    },
+    {
+      "epoch": 0.22336065573770492,
+      "grad_norm": 3.8819994926452637,
+      "learning_rate": 1.8104192247494618e-05,
+      "loss": 1.1141,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2235655737704918,
+      "grad_norm": 4.028090000152588,
+      "learning_rate": 1.81003018014323e-05,
+      "loss": 1.0254,
+      "step": 1091
+    },
+    {
+      "epoch": 0.22377049180327868,
+      "grad_norm": 2.365708827972412,
+      "learning_rate": 1.809640778652306e-05,
+      "loss": 1.0355,
+      "step": 1092
+    },
+    {
+      "epoch": 0.22397540983606556,
+      "grad_norm": 5.757355213165283,
+      "learning_rate": 1.8092510204482526e-05,
+      "loss": 1.0787,
+      "step": 1093
+    },
+    {
+      "epoch": 0.22418032786885245,
+      "grad_norm": 4.617683410644531,
+      "learning_rate": 1.8088609057027902e-05,
+      "loss": 1.0603,
+      "step": 1094
+    },
+    {
+      "epoch": 0.22438524590163936,
+      "grad_norm": 4.553032398223877,
+      "learning_rate": 1.8084704345877966e-05,
+      "loss": 0.9187,
+      "step": 1095
+    },
+    {
+      "epoch": 0.22459016393442624,
+      "grad_norm": 9.410490036010742,
+      "learning_rate": 1.8080796072753063e-05,
+      "loss": 0.9816,
+      "step": 1096
+    },
+    {
+      "epoch": 0.22479508196721312,
+      "grad_norm": 4.9451751708984375,
+      "learning_rate": 1.807688423937511e-05,
+      "loss": 1.2767,
+      "step": 1097
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 6.538898944854736,
+      "learning_rate": 1.807296884746758e-05,
+      "loss": 0.8631,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2252049180327869,
+      "grad_norm": 5.494630813598633,
+      "learning_rate": 1.8069049898755534e-05,
+      "loss": 1.1116,
+      "step": 1099
+    },
+    {
+      "epoch": 0.22540983606557377,
+      "grad_norm": 6.514915943145752,
+      "learning_rate": 1.806512739496559e-05,
+      "loss": 1.3968,
+      "step": 1100
+    },
+    {
+      "epoch": 0.22561475409836065,
+      "grad_norm": 7.082265377044678,
+      "learning_rate": 1.8061201337825927e-05,
+      "loss": 1.0573,
+      "step": 1101
+    },
+    {
+      "epoch": 0.22581967213114754,
+      "grad_norm": 7.936633586883545,
+      "learning_rate": 1.8057271729066297e-05,
+      "loss": 1.1772,
+      "step": 1102
+    },
+    {
+      "epoch": 0.22602459016393442,
+      "grad_norm": 4.701104164123535,
+      "learning_rate": 1.8053338570418017e-05,
+      "loss": 1.2081,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2262295081967213,
+      "grad_norm": 4.764162063598633,
+      "learning_rate": 1.8049401863613966e-05,
+      "loss": 1.1426,
+      "step": 1104
+    },
+    {
+      "epoch": 0.22643442622950818,
+      "grad_norm": 4.456353664398193,
+      "learning_rate": 1.8045461610388583e-05,
+      "loss": 1.1851,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2266393442622951,
+      "grad_norm": 4.390126705169678,
+      "learning_rate": 1.8041517812477875e-05,
+      "loss": 1.1327,
+      "step": 1106
+    },
+    {
+      "epoch": 0.22684426229508198,
+      "grad_norm": 3.2750320434570312,
+      "learning_rate": 1.8037570471619406e-05,
+      "loss": 0.9225,
+      "step": 1107
+    },
+    {
+      "epoch": 0.22704918032786886,
+      "grad_norm": 3.8049402236938477,
+      "learning_rate": 1.8033619589552304e-05,
+      "loss": 1.0642,
+      "step": 1108
+    },
+    {
+      "epoch": 0.22725409836065574,
+      "grad_norm": 3.7341670989990234,
+      "learning_rate": 1.8029665168017258e-05,
+      "loss": 1.0995,
+      "step": 1109
+    },
+    {
+      "epoch": 0.22745901639344263,
+      "grad_norm": 6.466542720794678,
+      "learning_rate": 1.802570720875651e-05,
+      "loss": 0.9796,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2276639344262295,
+      "grad_norm": 4.552248477935791,
+      "learning_rate": 1.8021745713513877e-05,
+      "loss": 1.0531,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2278688524590164,
+      "grad_norm": 7.346588611602783,
+      "learning_rate": 1.801778068403471e-05,
+      "loss": 1.463,
+      "step": 1112
+    },
+    {
+      "epoch": 0.22807377049180327,
+      "grad_norm": 10.799627304077148,
+      "learning_rate": 1.801381212206593e-05,
+      "loss": 1.1685,
+      "step": 1113
+    },
+    {
+      "epoch": 0.22827868852459016,
+      "grad_norm": 5.398428440093994,
+      "learning_rate": 1.800984002935602e-05,
+      "loss": 1.13,
+      "step": 1114
+    },
+    {
+      "epoch": 0.22848360655737704,
+      "grad_norm": 4.510843753814697,
+      "learning_rate": 1.8005864407655014e-05,
+      "loss": 0.9751,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22868852459016392,
+      "grad_norm": 4.843355655670166,
+      "learning_rate": 1.8001885258714494e-05,
+      "loss": 1.2295,
+      "step": 1116
+    },
+    {
+      "epoch": 0.22889344262295083,
+      "grad_norm": 7.293311595916748,
+      "learning_rate": 1.79979025842876e-05,
+      "loss": 1.3154,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22909836065573772,
+      "grad_norm": 4.251345634460449,
+      "learning_rate": 1.7993916386129033e-05,
+      "loss": 0.9182,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2293032786885246,
+      "grad_norm": 5.643980503082275,
+      "learning_rate": 1.7989926665995033e-05,
+      "loss": 1.2859,
+      "step": 1119
+    },
+    {
+      "epoch": 0.22950819672131148,
+      "grad_norm": 4.701449394226074,
+      "learning_rate": 1.7985933425643404e-05,
+      "loss": 0.8496,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22971311475409836,
+      "grad_norm": 4.486889839172363,
+      "learning_rate": 1.7981936666833494e-05,
+      "loss": 1.0978,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22991803278688525,
+      "grad_norm": 3.8645007610321045,
+      "learning_rate": 1.7977936391326204e-05,
+      "loss": 1.0646,
+      "step": 1122
+    },
+    {
+      "epoch": 0.23012295081967213,
+      "grad_norm": 5.873363018035889,
+      "learning_rate": 1.7973932600883982e-05,
+      "loss": 1.3031,
+      "step": 1123
+    },
+    {
+      "epoch": 0.230327868852459,
+      "grad_norm": 2.949512243270874,
+      "learning_rate": 1.796992529727082e-05,
+      "loss": 1.164,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2305327868852459,
+      "grad_norm": 5.568406105041504,
+      "learning_rate": 1.7965914482252274e-05,
+      "loss": 1.1558,
+      "step": 1125
+    },
+    {
+      "epoch": 0.23073770491803278,
+      "grad_norm": 7.000124931335449,
+      "learning_rate": 1.7961900157595435e-05,
+      "loss": 1.2479,
+      "step": 1126
+    },
+    {
+      "epoch": 0.23094262295081966,
+      "grad_norm": 3.3897554874420166,
+      "learning_rate": 1.7957882325068938e-05,
+      "loss": 0.9713,
+      "step": 1127
+    },
+    {
+      "epoch": 0.23114754098360657,
+      "grad_norm": 4.562893867492676,
+      "learning_rate": 1.7953860986442964e-05,
+      "loss": 1.137,
+      "step": 1128
+    },
+    {
+      "epoch": 0.23135245901639345,
+      "grad_norm": 4.2387189865112305,
+      "learning_rate": 1.7949836143489256e-05,
+      "loss": 1.1993,
+      "step": 1129
+    },
+    {
+      "epoch": 0.23155737704918034,
+      "grad_norm": 6.8403706550598145,
+      "learning_rate": 1.794580779798107e-05,
+      "loss": 1.2397,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23176229508196722,
+      "grad_norm": 2.8030765056610107,
+      "learning_rate": 1.7941775951693236e-05,
+      "loss": 0.6926,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2319672131147541,
+      "grad_norm": 3.48711895942688,
+      "learning_rate": 1.7937740606402104e-05,
+      "loss": 0.8808,
+      "step": 1132
+    },
+    {
+      "epoch": 0.23217213114754098,
+      "grad_norm": 9.551231384277344,
+      "learning_rate": 1.793370176388558e-05,
+      "loss": 1.4043,
+      "step": 1133
+    },
+    {
+      "epoch": 0.23237704918032787,
+      "grad_norm": 3.9807522296905518,
+      "learning_rate": 1.7929659425923104e-05,
+      "loss": 0.8657,
+      "step": 1134
+    },
+    {
+      "epoch": 0.23258196721311475,
+      "grad_norm": 4.706490516662598,
+      "learning_rate": 1.7925613594295655e-05,
+      "loss": 1.274,
+      "step": 1135
+    },
+    {
+      "epoch": 0.23278688524590163,
+      "grad_norm": 5.316752910614014,
+      "learning_rate": 1.7921564270785757e-05,
+      "loss": 1.1143,
+      "step": 1136
+    },
+    {
+      "epoch": 0.23299180327868851,
+      "grad_norm": 4.617867946624756,
+      "learning_rate": 1.7917511457177464e-05,
+      "loss": 0.9651,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2331967213114754,
+      "grad_norm": 5.244766712188721,
+      "learning_rate": 1.791345515525638e-05,
+      "loss": 0.9721,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2334016393442623,
+      "grad_norm": 4.282528400421143,
+      "learning_rate": 1.7909395366809632e-05,
+      "loss": 1.0727,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2336065573770492,
+      "grad_norm": 7.695476055145264,
+      "learning_rate": 1.7905332093625893e-05,
+      "loss": 0.982,
+      "step": 1140
+    },
+    {
+      "epoch": 0.23381147540983607,
+      "grad_norm": 5.313815116882324,
+      "learning_rate": 1.790126533749536e-05,
+      "loss": 1.0264,
+      "step": 1141
+    },
+    {
+      "epoch": 0.23401639344262296,
+      "grad_norm": 5.612570762634277,
+      "learning_rate": 1.7897195100209783e-05,
+      "loss": 1.277,
+      "step": 1142
+    },
+    {
+      "epoch": 0.23422131147540984,
+      "grad_norm": 6.332525253295898,
+      "learning_rate": 1.789312138356243e-05,
+      "loss": 1.2056,
+      "step": 1143
+    },
+    {
+      "epoch": 0.23442622950819672,
+      "grad_norm": 3.912306785583496,
+      "learning_rate": 1.788904418934811e-05,
+      "loss": 0.8944,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2346311475409836,
+      "grad_norm": 5.1158766746521,
+      "learning_rate": 1.7884963519363162e-05,
+      "loss": 0.9411,
+      "step": 1145
+    },
+    {
+      "epoch": 0.23483606557377049,
+      "grad_norm": 2.1139020919799805,
+      "learning_rate": 1.788087937540545e-05,
+      "loss": 0.8096,
+      "step": 1146
+    },
+    {
+      "epoch": 0.23504098360655737,
+      "grad_norm": 4.309313774108887,
+      "learning_rate": 1.7876791759274378e-05,
+      "loss": 1.2832,
+      "step": 1147
+    },
+    {
+      "epoch": 0.23524590163934425,
+      "grad_norm": 3.4459140300750732,
+      "learning_rate": 1.7872700672770876e-05,
+      "loss": 1.0488,
+      "step": 1148
+    },
+    {
+      "epoch": 0.23545081967213113,
+      "grad_norm": 8.454244613647461,
+      "learning_rate": 1.7868606117697396e-05,
+      "loss": 0.9946,
+      "step": 1149
+    },
+    {
+      "epoch": 0.23565573770491804,
+      "grad_norm": 5.462462902069092,
+      "learning_rate": 1.786450809585794e-05,
+      "loss": 1.152,
+      "step": 1150
+    },
+    {
+      "epoch": 0.23586065573770493,
+      "grad_norm": 2.7730815410614014,
+      "learning_rate": 1.7860406609058006e-05,
+      "loss": 0.9234,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2360655737704918,
+      "grad_norm": 5.486767768859863,
+      "learning_rate": 1.7856301659104647e-05,
+      "loss": 1.0965,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2362704918032787,
+      "grad_norm": 4.576313495635986,
+      "learning_rate": 1.7852193247806427e-05,
+      "loss": 1.3088,
+      "step": 1153
+    },
+    {
+      "epoch": 0.23647540983606558,
+      "grad_norm": 3.779038429260254,
+      "learning_rate": 1.7848081376973437e-05,
+      "loss": 0.9548,
+      "step": 1154
+    },
+    {
+      "epoch": 0.23668032786885246,
+      "grad_norm": 6.896630764007568,
+      "learning_rate": 1.7843966048417292e-05,
+      "loss": 1.2986,
+      "step": 1155
+    },
+    {
+      "epoch": 0.23688524590163934,
+      "grad_norm": 3.7893500328063965,
+      "learning_rate": 1.783984726395113e-05,
+      "loss": 0.7857,
+      "step": 1156
+    },
+    {
+      "epoch": 0.23709016393442622,
+      "grad_norm": 9.258204460144043,
+      "learning_rate": 1.7835725025389617e-05,
+      "loss": 0.9218,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2372950819672131,
+      "grad_norm": 12.809342384338379,
+      "learning_rate": 1.783159933454894e-05,
+      "loss": 1.1068,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 4.174959182739258,
+      "learning_rate": 1.7827470193246796e-05,
+      "loss": 1.2523,
+      "step": 1159
+    },
+    {
+      "epoch": 0.23770491803278687,
+      "grad_norm": 6.717686653137207,
+      "learning_rate": 1.7823337603302412e-05,
+      "loss": 0.8364,
+      "step": 1160
+    },
+    {
+      "epoch": 0.23790983606557378,
+      "grad_norm": 2.375969171524048,
+      "learning_rate": 1.781920156653654e-05,
+      "loss": 0.9131,
+      "step": 1161
+    },
+    {
+      "epoch": 0.23811475409836066,
+      "grad_norm": 5.470546722412109,
+      "learning_rate": 1.7815062084771434e-05,
+      "loss": 1.3088,
+      "step": 1162
+    },
+    {
+      "epoch": 0.23831967213114755,
+      "grad_norm": 7.928076267242432,
+      "learning_rate": 1.781091915983088e-05,
+      "loss": 0.8841,
+      "step": 1163
+    },
+    {
+      "epoch": 0.23852459016393443,
+      "grad_norm": 4.569458961486816,
+      "learning_rate": 1.780677279354017e-05,
+      "loss": 0.982,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2387295081967213,
+      "grad_norm": 4.667491912841797,
+      "learning_rate": 1.780262298772613e-05,
+      "loss": 0.9785,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2389344262295082,
+      "grad_norm": 3.142251491546631,
+      "learning_rate": 1.779846974421708e-05,
+      "loss": 0.8769,
+      "step": 1166
+    },
+    {
+      "epoch": 0.23913934426229508,
+      "grad_norm": 5.201872825622559,
+      "learning_rate": 1.7794313064842868e-05,
+      "loss": 1.0411,
+      "step": 1167
+    },
+    {
+      "epoch": 0.23934426229508196,
+      "grad_norm": 4.407444000244141,
+      "learning_rate": 1.779015295143485e-05,
+      "loss": 1.0673,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23954918032786884,
+      "grad_norm": 4.581703186035156,
+      "learning_rate": 1.7785989405825898e-05,
+      "loss": 1.0448,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23975409836065573,
+      "grad_norm": 4.342859268188477,
+      "learning_rate": 1.7781822429850396e-05,
+      "loss": 1.0484,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23995901639344264,
+      "grad_norm": 3.099581718444824,
+      "learning_rate": 1.7777652025344235e-05,
+      "loss": 1.1876,
+      "step": 1171
+    },
+    {
+      "epoch": 0.24016393442622952,
+      "grad_norm": 4.016264915466309,
+      "learning_rate": 1.7773478194144825e-05,
+      "loss": 0.9056,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2403688524590164,
+      "grad_norm": 4.305540561676025,
+      "learning_rate": 1.7769300938091076e-05,
+      "loss": 0.9448,
+      "step": 1173
+    },
+    {
+      "epoch": 0.24057377049180328,
+      "grad_norm": 3.946990489959717,
+      "learning_rate": 1.7765120259023416e-05,
+      "loss": 0.9147,
+      "step": 1174
+    },
+    {
+      "epoch": 0.24077868852459017,
+      "grad_norm": 2.8008649349212646,
+      "learning_rate": 1.7760936158783773e-05,
+      "loss": 1.0168,
+      "step": 1175
+    },
+    {
+      "epoch": 0.24098360655737705,
+      "grad_norm": 4.236630439758301,
+      "learning_rate": 1.775674863921559e-05,
+      "loss": 1.0782,
+      "step": 1176
+    },
+    {
+      "epoch": 0.24118852459016393,
+      "grad_norm": 2.90096378326416,
+      "learning_rate": 1.7752557702163807e-05,
+      "loss": 1.1972,
+      "step": 1177
+    },
+    {
+      "epoch": 0.24139344262295082,
+      "grad_norm": 4.5724358558654785,
+      "learning_rate": 1.774836334947488e-05,
+      "loss": 1.2068,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2415983606557377,
+      "grad_norm": 5.0162034034729,
+      "learning_rate": 1.7744165582996768e-05,
+      "loss": 1.3327,
+      "step": 1179
+    },
+    {
+      "epoch": 0.24180327868852458,
+      "grad_norm": 6.548060894012451,
+      "learning_rate": 1.7739964404578925e-05,
+      "loss": 1.1263,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24200819672131146,
+      "grad_norm": 28.511938095092773,
+      "learning_rate": 1.7735759816072312e-05,
+      "loss": 1.1605,
+      "step": 1181
+    },
+    {
+      "epoch": 0.24221311475409837,
+      "grad_norm": 3.9331748485565186,
+      "learning_rate": 1.7731551819329407e-05,
+      "loss": 0.7986,
+      "step": 1182
+    },
+    {
+      "epoch": 0.24241803278688526,
+      "grad_norm": 3.7204480171203613,
+      "learning_rate": 1.7727340416204165e-05,
+      "loss": 0.9838,
+      "step": 1183
+    },
+    {
+      "epoch": 0.24262295081967214,
+      "grad_norm": 4.327337741851807,
+      "learning_rate": 1.7723125608552063e-05,
+      "loss": 1.1717,
+      "step": 1184
+    },
+    {
+      "epoch": 0.24282786885245902,
+      "grad_norm": 3.6376969814300537,
+      "learning_rate": 1.7718907398230067e-05,
+      "loss": 0.9976,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2430327868852459,
+      "grad_norm": 4.748467445373535,
+      "learning_rate": 1.7714685787096636e-05,
+      "loss": 1.1316,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2432377049180328,
+      "grad_norm": 2.9113354682922363,
+      "learning_rate": 1.7710460777011745e-05,
+      "loss": 1.0758,
+      "step": 1187
+    },
+    {
+      "epoch": 0.24344262295081967,
+      "grad_norm": 6.221397876739502,
+      "learning_rate": 1.770623236983686e-05,
+      "loss": 0.9916,
+      "step": 1188
+    },
+    {
+      "epoch": 0.24364754098360655,
+      "grad_norm": 2.613605260848999,
+      "learning_rate": 1.770200056743493e-05,
+      "loss": 0.8974,
+      "step": 1189
+    },
+    {
+      "epoch": 0.24385245901639344,
+      "grad_norm": 7.1223602294921875,
+      "learning_rate": 1.769776537167042e-05,
+      "loss": 1.0797,
+      "step": 1190
+    },
+    {
+      "epoch": 0.24405737704918032,
+      "grad_norm": 7.631997585296631,
+      "learning_rate": 1.7693526784409278e-05,
+      "loss": 0.7838,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2442622950819672,
+      "grad_norm": 4.083502292633057,
+      "learning_rate": 1.7689284807518948e-05,
+      "loss": 0.7836,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2444672131147541,
+      "grad_norm": 9.48933219909668,
+      "learning_rate": 1.768503944286837e-05,
+      "loss": 1.1491,
+      "step": 1193
+    },
+    {
+      "epoch": 0.244672131147541,
+      "grad_norm": 3.949674129486084,
+      "learning_rate": 1.7680790692327974e-05,
+      "loss": 0.9412,
+      "step": 1194
+    },
+    {
+      "epoch": 0.24487704918032788,
+      "grad_norm": 2.862210512161255,
+      "learning_rate": 1.7676538557769685e-05,
+      "loss": 1.2048,
+      "step": 1195
+    },
+    {
+      "epoch": 0.24508196721311476,
+      "grad_norm": 2.6499650478363037,
+      "learning_rate": 1.7672283041066912e-05,
+      "loss": 0.9905,
+      "step": 1196
+    },
+    {
+      "epoch": 0.24528688524590164,
+      "grad_norm": 5.196450710296631,
+      "learning_rate": 1.7668024144094565e-05,
+      "loss": 1.1722,
+      "step": 1197
+    },
+    {
+      "epoch": 0.24549180327868853,
+      "grad_norm": 7.550540447235107,
+      "learning_rate": 1.7663761868729034e-05,
+      "loss": 1.1983,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2456967213114754,
+      "grad_norm": 4.352987289428711,
+      "learning_rate": 1.7659496216848202e-05,
+      "loss": 0.8276,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2459016393442623,
+      "grad_norm": 7.579164505004883,
+      "learning_rate": 1.7655227190331434e-05,
+      "loss": 1.116,
+      "step": 1200
+    },
+    {
+      "epoch": 0.24610655737704917,
+      "grad_norm": 5.2537078857421875,
+      "learning_rate": 1.7650954791059592e-05,
+      "loss": 1.1669,
+      "step": 1201
+    },
+    {
+      "epoch": 0.24631147540983606,
+      "grad_norm": 5.008403778076172,
+      "learning_rate": 1.7646679020915013e-05,
+      "loss": 1.1928,
+      "step": 1202
+    },
+    {
+      "epoch": 0.24651639344262294,
+      "grad_norm": 4.6789751052856445,
+      "learning_rate": 1.764239988178153e-05,
+      "loss": 1.1253,
+      "step": 1203
+    },
+    {
+      "epoch": 0.24672131147540985,
+      "grad_norm": 4.318830966949463,
+      "learning_rate": 1.7638117375544454e-05,
+      "loss": 1.076,
+      "step": 1204
+    },
+    {
+      "epoch": 0.24692622950819673,
+      "grad_norm": 3.2706451416015625,
+      "learning_rate": 1.763383150409057e-05,
+      "loss": 1.1679,
+      "step": 1205
+    },
+    {
+      "epoch": 0.24713114754098361,
+      "grad_norm": 8.470789909362793,
+      "learning_rate": 1.7629542269308165e-05,
+      "loss": 1.0179,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2473360655737705,
+      "grad_norm": 7.4565653800964355,
+      "learning_rate": 1.7625249673087e-05,
+      "loss": 1.3824,
+      "step": 1207
+    },
+    {
+      "epoch": 0.24754098360655738,
+      "grad_norm": 5.823420524597168,
+      "learning_rate": 1.762095371731831e-05,
+      "loss": 1.1537,
+      "step": 1208
+    },
+    {
+      "epoch": 0.24774590163934426,
+      "grad_norm": 4.383430004119873,
+      "learning_rate": 1.7616654403894818e-05,
+      "loss": 0.9493,
+      "step": 1209
+    },
+    {
+      "epoch": 0.24795081967213115,
+      "grad_norm": 3.1374683380126953,
+      "learning_rate": 1.7612351734710724e-05,
+      "loss": 0.8421,
+      "step": 1210
+    },
+    {
+      "epoch": 0.24815573770491803,
+      "grad_norm": 4.945880889892578,
+      "learning_rate": 1.7608045711661708e-05,
+      "loss": 1.128,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2483606557377049,
+      "grad_norm": 5.466212272644043,
+      "learning_rate": 1.760373633664492e-05,
+      "loss": 0.9218,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2485655737704918,
+      "grad_norm": 5.267118453979492,
+      "learning_rate": 1.7599423611559e-05,
+      "loss": 1.3664,
+      "step": 1213
+    },
+    {
+      "epoch": 0.24877049180327868,
+      "grad_norm": 5.102695465087891,
+      "learning_rate": 1.7595107538304054e-05,
+      "loss": 1.1033,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2489754098360656,
+      "grad_norm": 5.2802276611328125,
+      "learning_rate": 1.7590788118781668e-05,
+      "loss": 1.3009,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24918032786885247,
+      "grad_norm": 3.3143105506896973,
+      "learning_rate": 1.7586465354894895e-05,
+      "loss": 1.094,
+      "step": 1216
+    },
+    {
+      "epoch": 0.24938524590163935,
+      "grad_norm": 5.117283344268799,
+      "learning_rate": 1.758213924854827e-05,
+      "loss": 1.1248,
+      "step": 1217
+    },
+    {
+      "epoch": 0.24959016393442623,
+      "grad_norm": 5.151920795440674,
+      "learning_rate": 1.75778098016478e-05,
+      "loss": 1.1884,
+      "step": 1218
+    },
+    {
+      "epoch": 0.24979508196721312,
+      "grad_norm": 3.7192063331604004,
+      "learning_rate": 1.7573477016100962e-05,
+      "loss": 1.1192,
+      "step": 1219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.654855728149414,
+      "learning_rate": 1.7569140893816695e-05,
+      "loss": 1.0711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2502049180327869,
+      "grad_norm": 2.8077921867370605,
+      "learning_rate": 1.7564801436705425e-05,
+      "loss": 1.1036,
+      "step": 1221
+    },
+    {
+      "epoch": 0.25040983606557377,
+      "grad_norm": 6.334467887878418,
+      "learning_rate": 1.7560458646679034e-05,
+      "loss": 1.4896,
+      "step": 1222
+    },
+    {
+      "epoch": 0.25061475409836065,
+      "grad_norm": 3.975597858428955,
+      "learning_rate": 1.7556112525650875e-05,
+      "loss": 1.0504,
+      "step": 1223
+    },
+    {
+      "epoch": 0.25081967213114753,
+      "grad_norm": 7.000563144683838,
+      "learning_rate": 1.7551763075535778e-05,
+      "loss": 1.053,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2510245901639344,
+      "grad_norm": 4.710223197937012,
+      "learning_rate": 1.7547410298250026e-05,
+      "loss": 1.0624,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2512295081967213,
+      "grad_norm": 5.727853775024414,
+      "learning_rate": 1.754305419571138e-05,
+      "loss": 1.3079,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2514344262295082,
+      "grad_norm": 6.848883152008057,
+      "learning_rate": 1.7538694769839052e-05,
+      "loss": 1.0677,
+      "step": 1227
+    },
+    {
+      "epoch": 0.25163934426229506,
+      "grad_norm": 3.2617993354797363,
+      "learning_rate": 1.7534332022553733e-05,
+      "loss": 1.0539,
+      "step": 1228
+    },
+    {
+      "epoch": 0.25184426229508194,
+      "grad_norm": 3.7519922256469727,
+      "learning_rate": 1.752996595577757e-05,
+      "loss": 1.0218,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2520491803278688,
+      "grad_norm": 6.843014240264893,
+      "learning_rate": 1.752559657143417e-05,
+      "loss": 1.033,
+      "step": 1230
+    },
+    {
+      "epoch": 0.25225409836065577,
+      "grad_norm": 4.349466800689697,
+      "learning_rate": 1.7521223871448608e-05,
+      "loss": 1.1148,
+      "step": 1231
+    },
+    {
+      "epoch": 0.25245901639344265,
+      "grad_norm": 5.769502639770508,
+      "learning_rate": 1.7516847857747413e-05,
+      "loss": 1.0994,
+      "step": 1232
+    },
+    {
+      "epoch": 0.25266393442622953,
+      "grad_norm": 6.852553367614746,
+      "learning_rate": 1.751246853225858e-05,
+      "loss": 0.9163,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2528688524590164,
+      "grad_norm": 3.8027589321136475,
+      "learning_rate": 1.7508085896911565e-05,
+      "loss": 1.1333,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2530737704918033,
+      "grad_norm": 3.6916449069976807,
+      "learning_rate": 1.7503699953637272e-05,
+      "loss": 0.6908,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2532786885245902,
+      "grad_norm": 4.181401252746582,
+      "learning_rate": 1.7499310704368073e-05,
+      "loss": 0.991,
+      "step": 1236
+    },
+    {
+      "epoch": 0.25348360655737706,
+      "grad_norm": 7.227137565612793,
+      "learning_rate": 1.749491815103779e-05,
+      "loss": 1.0174,
+      "step": 1237
+    },
+    {
+      "epoch": 0.25368852459016394,
+      "grad_norm": 13.63884162902832,
+      "learning_rate": 1.7490522295581703e-05,
+      "loss": 1.0868,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2538934426229508,
+      "grad_norm": 6.987456321716309,
+      "learning_rate": 1.7486123139936543e-05,
+      "loss": 1.0436,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2540983606557377,
+      "grad_norm": 3.5601565837860107,
+      "learning_rate": 1.7481720686040506e-05,
+      "loss": 1.0974,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2543032786885246,
+      "grad_norm": 4.271510601043701,
+      "learning_rate": 1.7477314935833226e-05,
+      "loss": 1.1791,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2545081967213115,
+      "grad_norm": 6.030939102172852,
+      "learning_rate": 1.7472905891255806e-05,
+      "loss": 0.9026,
+      "step": 1242
+    },
+    {
+      "epoch": 0.25471311475409836,
+      "grad_norm": 3.549834728240967,
+      "learning_rate": 1.746849355425078e-05,
+      "loss": 1.0746,
+      "step": 1243
+    },
+    {
+      "epoch": 0.25491803278688524,
+      "grad_norm": 5.294729709625244,
+      "learning_rate": 1.7464077926762154e-05,
+      "loss": 1.0752,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2551229508196721,
+      "grad_norm": 10.35051155090332,
+      "learning_rate": 1.7459659010735376e-05,
+      "loss": 1.2263,
+      "step": 1245
+    },
+    {
+      "epoch": 0.255327868852459,
+      "grad_norm": 3.6949241161346436,
+      "learning_rate": 1.745523680811733e-05,
+      "loss": 0.9872,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2555327868852459,
+      "grad_norm": 7.128984451293945,
+      "learning_rate": 1.745081132085636e-05,
+      "loss": 1.0401,
+      "step": 1247
+    },
+    {
+      "epoch": 0.25573770491803277,
+      "grad_norm": 3.7170896530151367,
+      "learning_rate": 1.744638255090227e-05,
+      "loss": 1.158,
+      "step": 1248
+    },
+    {
+      "epoch": 0.25594262295081965,
+      "grad_norm": 3.8893070220947266,
+      "learning_rate": 1.744195050020628e-05,
+      "loss": 1.1189,
+      "step": 1249
+    },
+    {
+      "epoch": 0.25614754098360654,
+      "grad_norm": 5.831721305847168,
+      "learning_rate": 1.743751517072108e-05,
+      "loss": 0.9925,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2563524590163934,
+      "grad_norm": 5.768498420715332,
+      "learning_rate": 1.743307656440079e-05,
+      "loss": 1.0338,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2565573770491803,
+      "grad_norm": 5.331440448760986,
+      "learning_rate": 1.7428634683200987e-05,
+      "loss": 1.0034,
+      "step": 1252
+    },
+    {
+      "epoch": 0.25676229508196724,
+      "grad_norm": 4.463216781616211,
+      "learning_rate": 1.742418952907868e-05,
+      "loss": 1.2271,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2569672131147541,
+      "grad_norm": 6.627896785736084,
+      "learning_rate": 1.7419741103992326e-05,
+      "loss": 1.03,
+      "step": 1254
+    },
+    {
+      "epoch": 0.257172131147541,
+      "grad_norm": 4.272292137145996,
+      "learning_rate": 1.7415289409901817e-05,
+      "loss": 1.1066,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2573770491803279,
+      "grad_norm": 2.7570319175720215,
+      "learning_rate": 1.7410834448768493e-05,
+      "loss": 1.0684,
+      "step": 1256
+    },
+    {
+      "epoch": 0.25758196721311477,
+      "grad_norm": 3.619246482849121,
+      "learning_rate": 1.740637622255513e-05,
+      "loss": 0.9708,
+      "step": 1257
+    },
+    {
+      "epoch": 0.25778688524590165,
+      "grad_norm": 3.72554874420166,
+      "learning_rate": 1.7401914733225934e-05,
+      "loss": 0.9641,
+      "step": 1258
+    },
+    {
+      "epoch": 0.25799180327868854,
+      "grad_norm": 4.547085285186768,
+      "learning_rate": 1.739744998274657e-05,
+      "loss": 1.1108,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2581967213114754,
+      "grad_norm": 4.801657676696777,
+      "learning_rate": 1.7392981973084114e-05,
+      "loss": 1.2315,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2584016393442623,
+      "grad_norm": 5.536099433898926,
+      "learning_rate": 1.7388510706207098e-05,
+      "loss": 1.0656,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2586065573770492,
+      "grad_norm": 4.888221263885498,
+      "learning_rate": 1.738403618408548e-05,
+      "loss": 1.2632,
+      "step": 1262
+    },
+    {
+      "epoch": 0.25881147540983607,
+      "grad_norm": 3.3921926021575928,
+      "learning_rate": 1.737955840869065e-05,
+      "loss": 0.9164,
+      "step": 1263
+    },
+    {
+      "epoch": 0.25901639344262295,
+      "grad_norm": 3.838977336883545,
+      "learning_rate": 1.7375077381995443e-05,
+      "loss": 1.1455,
+      "step": 1264
+    },
+    {
+      "epoch": 0.25922131147540983,
+      "grad_norm": 2.8051953315734863,
+      "learning_rate": 1.7370593105974115e-05,
+      "loss": 1.0404,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2594262295081967,
+      "grad_norm": 4.853449821472168,
+      "learning_rate": 1.7366105582602353e-05,
+      "loss": 1.0194,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2596311475409836,
+      "grad_norm": 3.8788604736328125,
+      "learning_rate": 1.736161481385728e-05,
+      "loss": 1.2716,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2598360655737705,
+      "grad_norm": 5.106436252593994,
+      "learning_rate": 1.7357120801717452e-05,
+      "loss": 1.1892,
+      "step": 1268
+    },
+    {
+      "epoch": 0.26004098360655736,
+      "grad_norm": 7.339359283447266,
+      "learning_rate": 1.7352623548162847e-05,
+      "loss": 1.0315,
+      "step": 1269
+    },
+    {
+      "epoch": 0.26024590163934425,
+      "grad_norm": 3.8432071208953857,
+      "learning_rate": 1.7348123055174876e-05,
+      "loss": 1.2457,
+      "step": 1270
+    },
+    {
+      "epoch": 0.26045081967213113,
+      "grad_norm": 4.685242652893066,
+      "learning_rate": 1.7343619324736372e-05,
+      "loss": 0.9389,
+      "step": 1271
+    },
+    {
+      "epoch": 0.260655737704918,
+      "grad_norm": 17.182331085205078,
+      "learning_rate": 1.7339112358831597e-05,
+      "loss": 1.0847,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2608606557377049,
+      "grad_norm": 3.06630802154541,
+      "learning_rate": 1.7334602159446245e-05,
+      "loss": 1.2664,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2610655737704918,
+      "grad_norm": 5.024033546447754,
+      "learning_rate": 1.733008872856742e-05,
+      "loss": 1.0268,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2612704918032787,
+      "grad_norm": 2.827383041381836,
+      "learning_rate": 1.7325572068183664e-05,
+      "loss": 0.5614,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2614754098360656,
+      "grad_norm": 3.1347169876098633,
+      "learning_rate": 1.7321052180284937e-05,
+      "loss": 1.0809,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2616803278688525,
+      "grad_norm": 4.463098526000977,
+      "learning_rate": 1.7316529066862614e-05,
+      "loss": 1.1584,
+      "step": 1277
+    },
+    {
+      "epoch": 0.26188524590163936,
+      "grad_norm": 4.095867156982422,
+      "learning_rate": 1.7312002729909502e-05,
+      "loss": 1.3598,
+      "step": 1278
+    },
+    {
+      "epoch": 0.26209016393442625,
+      "grad_norm": 3.1553380489349365,
+      "learning_rate": 1.7307473171419823e-05,
+      "loss": 1.0807,
+      "step": 1279
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 5.3828959465026855,
+      "learning_rate": 1.730294039338922e-05,
+      "loss": 0.9884,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 5.089024543762207,
+      "learning_rate": 1.729840439781475e-05,
+      "loss": 0.97,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2627049180327869,
+      "grad_norm": 5.139216423034668,
+      "learning_rate": 1.7293865186694893e-05,
+      "loss": 1.0153,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2629098360655738,
+      "grad_norm": 2.6963984966278076,
+      "learning_rate": 1.728932276202954e-05,
+      "loss": 0.7773,
+      "step": 1283
+    },
+    {
+      "epoch": 0.26311475409836066,
+      "grad_norm": 5.1209516525268555,
+      "learning_rate": 1.7284777125820007e-05,
+      "loss": 0.992,
+      "step": 1284
+    },
+    {
+      "epoch": 0.26331967213114754,
+      "grad_norm": 3.1723945140838623,
+      "learning_rate": 1.7280228280069016e-05,
+      "loss": 1.2297,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2635245901639344,
+      "grad_norm": 5.78580379486084,
+      "learning_rate": 1.7275676226780707e-05,
+      "loss": 1.2583,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2637295081967213,
+      "grad_norm": 3.3825337886810303,
+      "learning_rate": 1.7271120967960632e-05,
+      "loss": 0.9569,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2639344262295082,
+      "grad_norm": 7.8138556480407715,
+      "learning_rate": 1.726656250561576e-05,
+      "loss": 1.3797,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2641393442622951,
+      "grad_norm": 2.5900895595550537,
+      "learning_rate": 1.7262000841754463e-05,
+      "loss": 1.0141,
+      "step": 1289
+    },
+    {
+      "epoch": 0.26434426229508196,
+      "grad_norm": 3.999483108520508,
+      "learning_rate": 1.7257435978386525e-05,
+      "loss": 1.0606,
+      "step": 1290
+    },
+    {
+      "epoch": 0.26454918032786884,
+      "grad_norm": 3.2367422580718994,
+      "learning_rate": 1.725286791752315e-05,
+      "loss": 1.0393,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2647540983606557,
+      "grad_norm": 8.1745023727417,
+      "learning_rate": 1.724829666117694e-05,
+      "loss": 1.0697,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2649590163934426,
+      "grad_norm": 5.052765846252441,
+      "learning_rate": 1.7243722211361902e-05,
+      "loss": 1.2757,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2651639344262295,
+      "grad_norm": 4.89489221572876,
+      "learning_rate": 1.7239144570093468e-05,
+      "loss": 1.1885,
+      "step": 1294
+    },
+    {
+      "epoch": 0.26536885245901637,
+      "grad_norm": 3.033461809158325,
+      "learning_rate": 1.7234563739388454e-05,
+      "loss": 0.813,
+      "step": 1295
+    },
+    {
+      "epoch": 0.26557377049180325,
+      "grad_norm": 2.8310492038726807,
+      "learning_rate": 1.72299797212651e-05,
+      "loss": 1.1015,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2657786885245902,
+      "grad_norm": 3.9985873699188232,
+      "learning_rate": 1.7225392517743035e-05,
+      "loss": 1.2153,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2659836065573771,
+      "grad_norm": 5.432483196258545,
+      "learning_rate": 1.72208021308433e-05,
+      "loss": 1.0101,
+      "step": 1298
+    },
+    {
+      "epoch": 0.26618852459016396,
+      "grad_norm": 4.7979416847229,
+      "learning_rate": 1.7216208562588335e-05,
+      "loss": 0.9378,
+      "step": 1299
+    },
+    {
+      "epoch": 0.26639344262295084,
+      "grad_norm": 5.303640842437744,
+      "learning_rate": 1.7211611815001986e-05,
+      "loss": 0.9613,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2665983606557377,
+      "grad_norm": 5.183971881866455,
+      "learning_rate": 1.72070118901095e-05,
+      "loss": 1.3127,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2668032786885246,
+      "grad_norm": 7.270858287811279,
+      "learning_rate": 1.7202408789937514e-05,
+      "loss": 1.0278,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2670081967213115,
+      "grad_norm": 8.736547470092773,
+      "learning_rate": 1.7197802516514076e-05,
+      "loss": 1.2309,
+      "step": 1303
+    },
+    {
+      "epoch": 0.26721311475409837,
+      "grad_norm": 4.2764153480529785,
+      "learning_rate": 1.7193193071868622e-05,
+      "loss": 0.9436,
+      "step": 1304
+    },
+    {
+      "epoch": 0.26741803278688525,
+      "grad_norm": 5.553186416625977,
+      "learning_rate": 1.7188580458032e-05,
+      "loss": 0.8933,
+      "step": 1305
+    },
+    {
+      "epoch": 0.26762295081967213,
+      "grad_norm": 3.1828882694244385,
+      "learning_rate": 1.7183964677036435e-05,
+      "loss": 1.0993,
+      "step": 1306
+    },
+    {
+      "epoch": 0.267827868852459,
+      "grad_norm": 5.5201735496521,
+      "learning_rate": 1.717934573091556e-05,
+      "loss": 1.4974,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2680327868852459,
+      "grad_norm": 4.221625804901123,
+      "learning_rate": 1.71747236217044e-05,
+      "loss": 1.1179,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2682377049180328,
+      "grad_norm": 5.322704792022705,
+      "learning_rate": 1.717009835143937e-05,
+      "loss": 0.885,
+      "step": 1309
+    },
+    {
+      "epoch": 0.26844262295081966,
+      "grad_norm": 4.825213432312012,
+      "learning_rate": 1.7165469922158284e-05,
+      "loss": 1.1611,
+      "step": 1310
+    },
+    {
+      "epoch": 0.26864754098360655,
+      "grad_norm": 3.7115633487701416,
+      "learning_rate": 1.7160838335900342e-05,
+      "loss": 0.6452,
+      "step": 1311
+    },
+    {
+      "epoch": 0.26885245901639343,
+      "grad_norm": 5.407562255859375,
+      "learning_rate": 1.7156203594706144e-05,
+      "loss": 0.9528,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2690573770491803,
+      "grad_norm": 8.982230186462402,
+      "learning_rate": 1.7151565700617662e-05,
+      "loss": 1.0905,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2692622950819672,
+      "grad_norm": 9.62136173248291,
+      "learning_rate": 1.714692465567828e-05,
+      "loss": 0.9802,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2694672131147541,
+      "grad_norm": 6.021239757537842,
+      "learning_rate": 1.714228046193275e-05,
+      "loss": 1.0682,
+      "step": 1315
+    },
+    {
+      "epoch": 0.26967213114754096,
+      "grad_norm": 4.527925968170166,
+      "learning_rate": 1.7137633121427225e-05,
+      "loss": 1.1769,
+      "step": 1316
+    },
+    {
+      "epoch": 0.26987704918032784,
+      "grad_norm": 4.250114917755127,
+      "learning_rate": 1.713298263620924e-05,
+      "loss": 1.3,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2700819672131147,
+      "grad_norm": 3.30751633644104,
+      "learning_rate": 1.7128329008327712e-05,
+      "loss": 1.056,
+      "step": 1318
+    },
+    {
+      "epoch": 0.27028688524590166,
+      "grad_norm": 5.6617865562438965,
+      "learning_rate": 1.7123672239832947e-05,
+      "loss": 0.911,
+      "step": 1319
+    },
+    {
+      "epoch": 0.27049180327868855,
+      "grad_norm": 9.998952865600586,
+      "learning_rate": 1.7119012332776632e-05,
+      "loss": 1.1109,
+      "step": 1320
+    },
+    {
+      "epoch": 0.27069672131147543,
+      "grad_norm": 6.66757345199585,
+      "learning_rate": 1.7114349289211838e-05,
+      "loss": 0.9634,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2709016393442623,
+      "grad_norm": 8.314507484436035,
+      "learning_rate": 1.7109683111193017e-05,
+      "loss": 0.9539,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2711065573770492,
+      "grad_norm": 4.732038974761963,
+      "learning_rate": 1.710501380077601e-05,
+      "loss": 0.9214,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2713114754098361,
+      "grad_norm": 5.667189121246338,
+      "learning_rate": 1.7100341360018024e-05,
+      "loss": 1.1487,
+      "step": 1324
+    },
+    {
+      "epoch": 0.27151639344262296,
+      "grad_norm": 5.735222339630127,
+      "learning_rate": 1.709566579097765e-05,
+      "loss": 1.1807,
+      "step": 1325
+    },
+    {
+      "epoch": 0.27172131147540984,
+      "grad_norm": 4.4173359870910645,
+      "learning_rate": 1.7090987095714863e-05,
+      "loss": 1.0533,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2719262295081967,
+      "grad_norm": 6.613245964050293,
+      "learning_rate": 1.7086305276291017e-05,
+      "loss": 1.1999,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2721311475409836,
+      "grad_norm": 5.432023525238037,
+      "learning_rate": 1.7081620334768828e-05,
+      "loss": 0.7152,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2723360655737705,
+      "grad_norm": 2.882643222808838,
+      "learning_rate": 1.70769322732124e-05,
+      "loss": 1.0239,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2725409836065574,
+      "grad_norm": 2.501784086227417,
+      "learning_rate": 1.7072241093687215e-05,
+      "loss": 0.7946,
+      "step": 1330
+    },
+    {
+      "epoch": 0.27274590163934426,
+      "grad_norm": 6.113258361816406,
+      "learning_rate": 1.7067546798260114e-05,
+      "loss": 0.9243,
+      "step": 1331
+    },
+    {
+      "epoch": 0.27295081967213114,
+      "grad_norm": 4.5418477058410645,
+      "learning_rate": 1.7062849388999318e-05,
+      "loss": 1.2083,
+      "step": 1332
+    },
+    {
+      "epoch": 0.273155737704918,
+      "grad_norm": 5.047703742980957,
+      "learning_rate": 1.7058148867974427e-05,
+      "loss": 1.2407,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2733606557377049,
+      "grad_norm": 4.4364399909973145,
+      "learning_rate": 1.7053445237256403e-05,
+      "loss": 0.935,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2735655737704918,
+      "grad_norm": 5.32100248336792,
+      "learning_rate": 1.704873849891758e-05,
+      "loss": 1.0408,
+      "step": 1335
+    },
+    {
+      "epoch": 0.27377049180327867,
+      "grad_norm": 10.243492126464844,
+      "learning_rate": 1.7044028655031667e-05,
+      "loss": 0.8732,
+      "step": 1336
+    },
+    {
+      "epoch": 0.27397540983606555,
+      "grad_norm": 3.0960004329681396,
+      "learning_rate": 1.7039315707673726e-05,
+      "loss": 1.1912,
+      "step": 1337
+    },
+    {
+      "epoch": 0.27418032786885244,
+      "grad_norm": 4.346772193908691,
+      "learning_rate": 1.7034599658920206e-05,
+      "loss": 0.9664,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2743852459016393,
+      "grad_norm": 4.438989639282227,
+      "learning_rate": 1.702988051084891e-05,
+      "loss": 1.2471,
+      "step": 1339
+    },
+    {
+      "epoch": 0.27459016393442626,
+      "grad_norm": 4.474316596984863,
+      "learning_rate": 1.702515826553901e-05,
+      "loss": 1.1196,
+      "step": 1340
+    },
+    {
+      "epoch": 0.27479508196721314,
+      "grad_norm": 4.5468525886535645,
+      "learning_rate": 1.702043292507104e-05,
+      "loss": 0.8334,
+      "step": 1341
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 4.670716285705566,
+      "learning_rate": 1.7015704491526902e-05,
+      "loss": 1.0592,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2752049180327869,
+      "grad_norm": 7.507452964782715,
+      "learning_rate": 1.701097296698986e-05,
+      "loss": 1.1103,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2754098360655738,
+      "grad_norm": 3.8601083755493164,
+      "learning_rate": 1.7006238353544536e-05,
+      "loss": 1.0842,
+      "step": 1344
+    },
+    {
+      "epoch": 0.27561475409836067,
+      "grad_norm": 3.6346490383148193,
+      "learning_rate": 1.700150065327691e-05,
+      "loss": 0.8967,
+      "step": 1345
+    },
+    {
+      "epoch": 0.27581967213114755,
+      "grad_norm": 3.8468222618103027,
+      "learning_rate": 1.699675986827434e-05,
+      "loss": 1.0899,
+      "step": 1346
+    },
+    {
+      "epoch": 0.27602459016393444,
+      "grad_norm": 5.668501377105713,
+      "learning_rate": 1.6992016000625523e-05,
+      "loss": 1.0981,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2762295081967213,
+      "grad_norm": 4.882596015930176,
+      "learning_rate": 1.698726905242052e-05,
+      "loss": 1.1523,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2764344262295082,
+      "grad_norm": 5.116253852844238,
+      "learning_rate": 1.698251902575075e-05,
+      "loss": 1.0151,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2766393442622951,
+      "grad_norm": 5.111154079437256,
+      "learning_rate": 1.6977765922708995e-05,
+      "loss": 1.0001,
+      "step": 1350
+    },
+    {
+      "epoch": 0.27684426229508197,
+      "grad_norm": 4.156217098236084,
+      "learning_rate": 1.6973009745389383e-05,
+      "loss": 0.8912,
+      "step": 1351
+    },
+    {
+      "epoch": 0.27704918032786885,
+      "grad_norm": 6.405979633331299,
+      "learning_rate": 1.69682504958874e-05,
+      "loss": 1.2749,
+      "step": 1352
+    },
+    {
+      "epoch": 0.27725409836065573,
+      "grad_norm": 6.193970203399658,
+      "learning_rate": 1.6963488176299888e-05,
+      "loss": 1.0543,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2774590163934426,
+      "grad_norm": 7.5982985496521,
+      "learning_rate": 1.6958722788725037e-05,
+      "loss": 0.8066,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2776639344262295,
+      "grad_norm": 2.7278056144714355,
+      "learning_rate": 1.6953954335262384e-05,
+      "loss": 1.086,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2778688524590164,
+      "grad_norm": 3.673245906829834,
+      "learning_rate": 1.6949182818012837e-05,
+      "loss": 0.6419,
+      "step": 1356
+    },
+    {
+      "epoch": 0.27807377049180326,
+      "grad_norm": 7.027377128601074,
+      "learning_rate": 1.6944408239078632e-05,
+      "loss": 1.23,
+      "step": 1357
+    },
+    {
+      "epoch": 0.27827868852459015,
+      "grad_norm": 3.6625473499298096,
+      "learning_rate": 1.6939630600563366e-05,
+      "loss": 1.2823,
+      "step": 1358
+    },
+    {
+      "epoch": 0.27848360655737703,
+      "grad_norm": 6.815012454986572,
+      "learning_rate": 1.6934849904571977e-05,
+      "loss": 1.0588,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2786885245901639,
+      "grad_norm": 3.9845244884490967,
+      "learning_rate": 1.6930066153210756e-05,
+      "loss": 1.0478,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2788934426229508,
+      "grad_norm": 2.895371675491333,
+      "learning_rate": 1.6925279348587337e-05,
+      "loss": 0.8456,
+      "step": 1361
+    },
+    {
+      "epoch": 0.27909836065573773,
+      "grad_norm": 4.102571487426758,
+      "learning_rate": 1.6920489492810702e-05,
+      "loss": 1.1926,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2793032786885246,
+      "grad_norm": 3.2833166122436523,
+      "learning_rate": 1.6915696587991167e-05,
+      "loss": 1.0291,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2795081967213115,
+      "grad_norm": 8.0424165725708,
+      "learning_rate": 1.6910900636240407e-05,
+      "loss": 0.8889,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2797131147540984,
+      "grad_norm": 4.53961706161499,
+      "learning_rate": 1.690610163967143e-05,
+      "loss": 0.9379,
+      "step": 1365
+    },
+    {
+      "epoch": 0.27991803278688526,
+      "grad_norm": 3.578051805496216,
+      "learning_rate": 1.690129960039859e-05,
+      "loss": 1.0203,
+      "step": 1366
+    },
+    {
+      "epoch": 0.28012295081967215,
+      "grad_norm": 3.6766607761383057,
+      "learning_rate": 1.6896494520537576e-05,
+      "loss": 0.9164,
+      "step": 1367
+    },
+    {
+      "epoch": 0.28032786885245903,
+      "grad_norm": 2.8674488067626953,
+      "learning_rate": 1.689168640220542e-05,
+      "loss": 1.0397,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2805327868852459,
+      "grad_norm": 5.320888042449951,
+      "learning_rate": 1.688687524752049e-05,
+      "loss": 1.1155,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2807377049180328,
+      "grad_norm": 3.3699545860290527,
+      "learning_rate": 1.68820610586025e-05,
+      "loss": 0.8335,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2809426229508197,
+      "grad_norm": 5.336819648742676,
+      "learning_rate": 1.6877243837572486e-05,
+      "loss": 0.9464,
+      "step": 1371
+    },
+    {
+      "epoch": 0.28114754098360656,
+      "grad_norm": 4.018170356750488,
+      "learning_rate": 1.6872423586552837e-05,
+      "loss": 1.0759,
+      "step": 1372
+    },
+    {
+      "epoch": 0.28135245901639344,
+      "grad_norm": 5.378224849700928,
+      "learning_rate": 1.6867600307667263e-05,
+      "loss": 0.981,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2815573770491803,
+      "grad_norm": 15.710500717163086,
+      "learning_rate": 1.6862774003040817e-05,
+      "loss": 1.3669,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2817622950819672,
+      "grad_norm": 6.35200309753418,
+      "learning_rate": 1.685794467479988e-05,
+      "loss": 1.1271,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2819672131147541,
+      "grad_norm": 30.737777709960938,
+      "learning_rate": 1.6853112325072172e-05,
+      "loss": 1.1217,
+      "step": 1376
+    },
+    {
+      "epoch": 0.28217213114754097,
+      "grad_norm": 4.332432270050049,
+      "learning_rate": 1.684827695598673e-05,
+      "loss": 1.3833,
+      "step": 1377
+    },
+    {
+      "epoch": 0.28237704918032785,
+      "grad_norm": 2.628124713897705,
+      "learning_rate": 1.684343856967394e-05,
+      "loss": 0.9186,
+      "step": 1378
+    },
+    {
+      "epoch": 0.28258196721311474,
+      "grad_norm": 5.450994491577148,
+      "learning_rate": 1.6838597168265497e-05,
+      "loss": 1.2084,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2827868852459016,
+      "grad_norm": 7.592185974121094,
+      "learning_rate": 1.6833752753894444e-05,
+      "loss": 1.2881,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2829918032786885,
+      "grad_norm": 4.731937885284424,
+      "learning_rate": 1.6828905328695144e-05,
+      "loss": 1.0113,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2831967213114754,
+      "grad_norm": 4.027661323547363,
+      "learning_rate": 1.682405489480328e-05,
+      "loss": 0.99,
+      "step": 1382
+    },
+    {
+      "epoch": 0.28340163934426227,
+      "grad_norm": 5.051592826843262,
+      "learning_rate": 1.6819201454355865e-05,
+      "loss": 0.9464,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2836065573770492,
+      "grad_norm": 3.8045828342437744,
+      "learning_rate": 1.681434500949124e-05,
+      "loss": 1.0388,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2838114754098361,
+      "grad_norm": 2.8553664684295654,
+      "learning_rate": 1.6809485562349064e-05,
+      "loss": 0.5913,
+      "step": 1385
+    },
+    {
+      "epoch": 0.28401639344262297,
+      "grad_norm": 3.5638301372528076,
+      "learning_rate": 1.6804623115070325e-05,
+      "loss": 1.2464,
+      "step": 1386
+    },
+    {
+      "epoch": 0.28422131147540985,
+      "grad_norm": 3.4604363441467285,
+      "learning_rate": 1.679975766979733e-05,
+      "loss": 1.0894,
+      "step": 1387
+    },
+    {
+      "epoch": 0.28442622950819674,
+      "grad_norm": 4.8250322341918945,
+      "learning_rate": 1.6794889228673706e-05,
+      "loss": 1.0411,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2846311475409836,
+      "grad_norm": 3.625021457672119,
+      "learning_rate": 1.6790017793844398e-05,
+      "loss": 1.0638,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2848360655737705,
+      "grad_norm": 4.9896674156188965,
+      "learning_rate": 1.6785143367455673e-05,
+      "loss": 1.084,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2850409836065574,
+      "grad_norm": 3.9496371746063232,
+      "learning_rate": 1.6780265951655116e-05,
+      "loss": 1.1298,
+      "step": 1391
+    },
+    {
+      "epoch": 0.28524590163934427,
+      "grad_norm": 7.626348972320557,
+      "learning_rate": 1.6775385548591633e-05,
+      "loss": 1.1482,
+      "step": 1392
+    },
+    {
+      "epoch": 0.28545081967213115,
+      "grad_norm": 8.649357795715332,
+      "learning_rate": 1.6770502160415433e-05,
+      "loss": 1.0688,
+      "step": 1393
+    },
+    {
+      "epoch": 0.28565573770491803,
+      "grad_norm": 5.55466365814209,
+      "learning_rate": 1.6765615789278055e-05,
+      "loss": 1.2032,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2858606557377049,
+      "grad_norm": 4.172555446624756,
+      "learning_rate": 1.6760726437332345e-05,
+      "loss": 1.0446,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2860655737704918,
+      "grad_norm": 10.14751148223877,
+      "learning_rate": 1.675583410673246e-05,
+      "loss": 1.1246,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2862704918032787,
+      "grad_norm": 6.89901065826416,
+      "learning_rate": 1.6750938799633874e-05,
+      "loss": 1.0936,
+      "step": 1397
+    },
+    {
+      "epoch": 0.28647540983606556,
+      "grad_norm": 5.048938751220703,
+      "learning_rate": 1.6746040518193376e-05,
+      "loss": 1.1995,
+      "step": 1398
+    },
+    {
+      "epoch": 0.28668032786885245,
+      "grad_norm": 3.8240716457366943,
+      "learning_rate": 1.6741139264569055e-05,
+      "loss": 1.0714,
+      "step": 1399
+    },
+    {
+      "epoch": 0.28688524590163933,
+      "grad_norm": 3.072120189666748,
+      "learning_rate": 1.6736235040920317e-05,
+      "loss": 1.0344,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2870901639344262,
+      "grad_norm": 4.184927463531494,
+      "learning_rate": 1.6731327849407873e-05,
+      "loss": 1.0667,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2872950819672131,
+      "grad_norm": 2.5344595909118652,
+      "learning_rate": 1.6726417692193745e-05,
+      "loss": 0.9152,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 4.82230281829834,
+      "learning_rate": 1.672150457144126e-05,
+      "loss": 1.106,
+      "step": 1403
+    },
+    {
+      "epoch": 0.28770491803278686,
+      "grad_norm": 5.214466571807861,
+      "learning_rate": 1.6716588489315046e-05,
+      "loss": 1.4609,
+      "step": 1404
+    },
+    {
+      "epoch": 0.28790983606557374,
+      "grad_norm": 5.115179538726807,
+      "learning_rate": 1.6711669447981048e-05,
+      "loss": 0.7038,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2881147540983607,
+      "grad_norm": 3.925933361053467,
+      "learning_rate": 1.6706747449606497e-05,
+      "loss": 1.1982,
+      "step": 1406
+    },
+    {
+      "epoch": 0.28831967213114756,
+      "grad_norm": 4.628397464752197,
+      "learning_rate": 1.670182249635995e-05,
+      "loss": 1.0949,
+      "step": 1407
+    },
+    {
+      "epoch": 0.28852459016393445,
+      "grad_norm": 4.943192481994629,
+      "learning_rate": 1.6696894590411237e-05,
+      "loss": 0.9567,
+      "step": 1408
+    },
+    {
+      "epoch": 0.28872950819672133,
+      "grad_norm": 4.787381172180176,
+      "learning_rate": 1.6691963733931515e-05,
+      "loss": 0.8382,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2889344262295082,
+      "grad_norm": 4.902907371520996,
+      "learning_rate": 1.668702992909323e-05,
+      "loss": 1.1778,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2891393442622951,
+      "grad_norm": 4.018610000610352,
+      "learning_rate": 1.6682093178070122e-05,
+      "loss": 0.9663,
+      "step": 1411
+    },
+    {
+      "epoch": 0.289344262295082,
+      "grad_norm": 4.672762870788574,
+      "learning_rate": 1.667715348303724e-05,
+      "loss": 0.9208,
+      "step": 1412
+    },
+    {
+      "epoch": 0.28954918032786886,
+      "grad_norm": 5.474822044372559,
+      "learning_rate": 1.6672210846170922e-05,
+      "loss": 1.3264,
+      "step": 1413
+    },
+    {
+      "epoch": 0.28975409836065574,
+      "grad_norm": 3.579280138015747,
+      "learning_rate": 1.6667265269648805e-05,
+      "loss": 0.7421,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2899590163934426,
+      "grad_norm": 7.362536907196045,
+      "learning_rate": 1.666231675564982e-05,
+      "loss": 1.1925,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2901639344262295,
+      "grad_norm": 10.942028999328613,
+      "learning_rate": 1.665736530635419e-05,
+      "loss": 0.8575,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2903688524590164,
+      "grad_norm": 8.34606647491455,
+      "learning_rate": 1.6652410923943448e-05,
+      "loss": 1.1292,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2905737704918033,
+      "grad_norm": 3.8088505268096924,
+      "learning_rate": 1.6647453610600387e-05,
+      "loss": 0.9817,
+      "step": 1418
+    },
+    {
+      "epoch": 0.29077868852459016,
+      "grad_norm": 4.814629077911377,
+      "learning_rate": 1.664249336850912e-05,
+      "loss": 1.1978,
+      "step": 1419
+    },
+    {
+      "epoch": 0.29098360655737704,
+      "grad_norm": 6.303809642791748,
+      "learning_rate": 1.6637530199855042e-05,
+      "loss": 1.1773,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2911885245901639,
+      "grad_norm": 3.8096065521240234,
+      "learning_rate": 1.663256410682483e-05,
+      "loss": 1.1375,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2913934426229508,
+      "grad_norm": 4.899345874786377,
+      "learning_rate": 1.662759509160646e-05,
+      "loss": 1.0853,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2915983606557377,
+      "grad_norm": 6.021592617034912,
+      "learning_rate": 1.6622623156389186e-05,
+      "loss": 0.9953,
+      "step": 1423
+    },
+    {
+      "epoch": 0.29180327868852457,
+      "grad_norm": 7.4281229972839355,
+      "learning_rate": 1.6617648303363555e-05,
+      "loss": 0.8531,
+      "step": 1424
+    },
+    {
+      "epoch": 0.29200819672131145,
+      "grad_norm": 3.209414482116699,
+      "learning_rate": 1.66126705347214e-05,
+      "loss": 1.0459,
+      "step": 1425
+    },
+    {
+      "epoch": 0.29221311475409834,
+      "grad_norm": 7.388914585113525,
+      "learning_rate": 1.660768985265583e-05,
+      "loss": 1.0843,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2924180327868853,
+      "grad_norm": 6.163972854614258,
+      "learning_rate": 1.660270625936125e-05,
+      "loss": 1.3072,
+      "step": 1427
+    },
+    {
+      "epoch": 0.29262295081967216,
+      "grad_norm": 6.782332897186279,
+      "learning_rate": 1.6597719757033336e-05,
+      "loss": 0.9218,
+      "step": 1428
+    },
+    {
+      "epoch": 0.29282786885245904,
+      "grad_norm": 7.164936065673828,
+      "learning_rate": 1.6592730347869055e-05,
+      "loss": 0.9401,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2930327868852459,
+      "grad_norm": 2.7654645442962646,
+      "learning_rate": 1.6587738034066653e-05,
+      "loss": 0.75,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2932377049180328,
+      "grad_norm": 9.380576133728027,
+      "learning_rate": 1.6582742817825646e-05,
+      "loss": 0.8796,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2934426229508197,
+      "grad_norm": 8.308199882507324,
+      "learning_rate": 1.657774470134684e-05,
+      "loss": 1.091,
+      "step": 1432
+    },
+    {
+      "epoch": 0.29364754098360657,
+      "grad_norm": 3.630551815032959,
+      "learning_rate": 1.6572743686832317e-05,
+      "loss": 0.6222,
+      "step": 1433
+    },
+    {
+      "epoch": 0.29385245901639345,
+      "grad_norm": 5.9973039627075195,
+      "learning_rate": 1.656773977648543e-05,
+      "loss": 1.2812,
+      "step": 1434
+    },
+    {
+      "epoch": 0.29405737704918034,
+      "grad_norm": 2.761718988418579,
+      "learning_rate": 1.6562732972510816e-05,
+      "loss": 0.9128,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2942622950819672,
+      "grad_norm": 7.670491695404053,
+      "learning_rate": 1.655772327711438e-05,
+      "loss": 0.943,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2944672131147541,
+      "grad_norm": 5.98268985748291,
+      "learning_rate": 1.65527106925033e-05,
+      "loss": 1.3429,
+      "step": 1437
+    },
+    {
+      "epoch": 0.294672131147541,
+      "grad_norm": 5.4197587966918945,
+      "learning_rate": 1.654769522088604e-05,
+      "loss": 0.9657,
+      "step": 1438
+    },
+    {
+      "epoch": 0.29487704918032787,
+      "grad_norm": 3.461679220199585,
+      "learning_rate": 1.6542676864472317e-05,
+      "loss": 0.8999,
+      "step": 1439
+    },
+    {
+      "epoch": 0.29508196721311475,
+      "grad_norm": 8.12448501586914,
+      "learning_rate": 1.6537655625473133e-05,
+      "loss": 0.8832,
+      "step": 1440
+    },
+    {
+      "epoch": 0.29528688524590163,
+      "grad_norm": 7.60206937789917,
+      "learning_rate": 1.6532631506100753e-05,
+      "loss": 1.1601,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2954918032786885,
+      "grad_norm": 7.946483612060547,
+      "learning_rate": 1.6527604508568714e-05,
+      "loss": 1.0177,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2956967213114754,
+      "grad_norm": 5.50120210647583,
+      "learning_rate": 1.6522574635091818e-05,
+      "loss": 0.7557,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2959016393442623,
+      "grad_norm": 6.890280723571777,
+      "learning_rate": 1.6517541887886147e-05,
+      "loss": 1.0912,
+      "step": 1444
+    },
+    {
+      "epoch": 0.29610655737704916,
+      "grad_norm": 6.498577117919922,
+      "learning_rate": 1.651250626916903e-05,
+      "loss": 1.0175,
+      "step": 1445
+    },
+    {
+      "epoch": 0.29631147540983604,
+      "grad_norm": 7.339698314666748,
+      "learning_rate": 1.6507467781159067e-05,
+      "loss": 0.9537,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2965163934426229,
+      "grad_norm": 8.501199722290039,
+      "learning_rate": 1.6502426426076134e-05,
+      "loss": 0.9291,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2967213114754098,
+      "grad_norm": 5.909817218780518,
+      "learning_rate": 1.649738220614135e-05,
+      "loss": 1.19,
+      "step": 1448
+    },
+    {
+      "epoch": 0.29692622950819675,
+      "grad_norm": 5.385056972503662,
+      "learning_rate": 1.6492335123577117e-05,
+      "loss": 0.7117,
+      "step": 1449
+    },
+    {
+      "epoch": 0.29713114754098363,
+      "grad_norm": 8.374762535095215,
+      "learning_rate": 1.6487285180607085e-05,
+      "loss": 0.8529,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2973360655737705,
+      "grad_norm": 2.4098870754241943,
+      "learning_rate": 1.6482232379456168e-05,
+      "loss": 1.1404,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2975409836065574,
+      "grad_norm": 3.4348599910736084,
+      "learning_rate": 1.647717672235054e-05,
+      "loss": 1.0015,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2977459016393443,
+      "grad_norm": 8.167287826538086,
+      "learning_rate": 1.6472118211517632e-05,
+      "loss": 1.1166,
+      "step": 1453
+    },
+    {
+      "epoch": 0.29795081967213116,
+      "grad_norm": 5.00312614440918,
+      "learning_rate": 1.646705684918613e-05,
+      "loss": 1.139,
+      "step": 1454
+    },
+    {
+      "epoch": 0.29815573770491804,
+      "grad_norm": 5.405128479003906,
+      "learning_rate": 1.6461992637585983e-05,
+      "loss": 1.0365,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2983606557377049,
+      "grad_norm": 3.6091434955596924,
+      "learning_rate": 1.6456925578948388e-05,
+      "loss": 1.1891,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2985655737704918,
+      "grad_norm": 4.759099960327148,
+      "learning_rate": 1.6451855675505802e-05,
+      "loss": 1.199,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2987704918032787,
+      "grad_norm": 2.682908296585083,
+      "learning_rate": 1.6446782929491934e-05,
+      "loss": 1.1696,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2989754098360656,
+      "grad_norm": 7.235214710235596,
+      "learning_rate": 1.644170734314174e-05,
+      "loss": 1.2345,
+      "step": 1459
+    },
+    {
+      "epoch": 0.29918032786885246,
+      "grad_norm": 4.574580669403076,
+      "learning_rate": 1.6436628918691434e-05,
+      "loss": 0.9524,
+      "step": 1460
+    },
+    {
+      "epoch": 0.29938524590163934,
+      "grad_norm": 3.058997869491577,
+      "learning_rate": 1.6431547658378478e-05,
+      "loss": 0.7876,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2995901639344262,
+      "grad_norm": 5.858450889587402,
+      "learning_rate": 1.6426463564441585e-05,
+      "loss": 1.2167,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2997950819672131,
+      "grad_norm": 3.4082489013671875,
+      "learning_rate": 1.6421376639120717e-05,
+      "loss": 1.2555,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3898441791534424,
+      "learning_rate": 1.641628688465708e-05,
+      "loss": 1.1177,
+      "step": 1464
+    },
+    {
+      "epoch": 0.30020491803278687,
+      "grad_norm": 3.0979738235473633,
+      "learning_rate": 1.641119430329312e-05,
+      "loss": 0.9873,
+      "step": 1465
+    },
+    {
+      "epoch": 0.30040983606557375,
+      "grad_norm": 4.333409786224365,
+      "learning_rate": 1.640609889727255e-05,
+      "loss": 0.968,
+      "step": 1466
+    },
+    {
+      "epoch": 0.30061475409836064,
+      "grad_norm": 5.049163818359375,
+      "learning_rate": 1.6401000668840306e-05,
+      "loss": 0.9298,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3008196721311475,
+      "grad_norm": 3.625288963317871,
+      "learning_rate": 1.6395899620242577e-05,
+      "loss": 1.1523,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3010245901639344,
+      "grad_norm": 4.728281497955322,
+      "learning_rate": 1.639079575372679e-05,
+      "loss": 1.0657,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3012295081967213,
+      "grad_norm": 5.57958459854126,
+      "learning_rate": 1.6385689071541627e-05,
+      "loss": 0.9182,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3014344262295082,
+      "grad_norm": 5.307872295379639,
+      "learning_rate": 1.638057957593699e-05,
+      "loss": 1.0954,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3016393442622951,
+      "grad_norm": 4.240473747253418,
+      "learning_rate": 1.637546726916403e-05,
+      "loss": 1.272,
+      "step": 1472
+    },
+    {
+      "epoch": 0.301844262295082,
+      "grad_norm": 6.792049884796143,
+      "learning_rate": 1.6370352153475147e-05,
+      "loss": 1.0829,
+      "step": 1473
+    },
+    {
+      "epoch": 0.30204918032786887,
+      "grad_norm": 3.341514825820923,
+      "learning_rate": 1.6365234231123955e-05,
+      "loss": 0.8475,
+      "step": 1474
+    },
+    {
+      "epoch": 0.30225409836065575,
+      "grad_norm": 4.690948486328125,
+      "learning_rate": 1.6360113504365333e-05,
+      "loss": 0.8639,
+      "step": 1475
+    },
+    {
+      "epoch": 0.30245901639344264,
+      "grad_norm": 3.0989930629730225,
+      "learning_rate": 1.6354989975455368e-05,
+      "loss": 1.2244,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3026639344262295,
+      "grad_norm": 7.885980606079102,
+      "learning_rate": 1.63498636466514e-05,
+      "loss": 1.1628,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3028688524590164,
+      "grad_norm": 4.565046310424805,
+      "learning_rate": 1.6344734520212e-05,
+      "loss": 0.8678,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3030737704918033,
+      "grad_norm": 4.29588508605957,
+      "learning_rate": 1.6339602598396967e-05,
+      "loss": 1.0472,
+      "step": 1479
+    },
+    {
+      "epoch": 0.30327868852459017,
+      "grad_norm": 2.399726629257202,
+      "learning_rate": 1.633446788346733e-05,
+      "loss": 0.6357,
+      "step": 1480
+    },
+    {
+      "epoch": 0.30348360655737705,
+      "grad_norm": 3.7031455039978027,
+      "learning_rate": 1.6329330377685356e-05,
+      "loss": 0.6939,
+      "step": 1481
+    },
+    {
+      "epoch": 0.30368852459016393,
+      "grad_norm": 4.108364582061768,
+      "learning_rate": 1.632419008331453e-05,
+      "loss": 1.045,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3038934426229508,
+      "grad_norm": 6.093892574310303,
+      "learning_rate": 1.6319047002619584e-05,
+      "loss": 1.3108,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3040983606557377,
+      "grad_norm": 3.7725772857666016,
+      "learning_rate": 1.6313901137866456e-05,
+      "loss": 1.0294,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3043032786885246,
+      "grad_norm": 4.6957783699035645,
+      "learning_rate": 1.630875249132233e-05,
+      "loss": 1.0028,
+      "step": 1485
+    },
+    {
+      "epoch": 0.30450819672131146,
+      "grad_norm": 5.200054168701172,
+      "learning_rate": 1.63036010652556e-05,
+      "loss": 0.9181,
+      "step": 1486
+    },
+    {
+      "epoch": 0.30471311475409835,
+      "grad_norm": 7.854465961456299,
+      "learning_rate": 1.6298446861935898e-05,
+      "loss": 0.9993,
+      "step": 1487
+    },
+    {
+      "epoch": 0.30491803278688523,
+      "grad_norm": 2.941171169281006,
+      "learning_rate": 1.6293289883634066e-05,
+      "loss": 1.1155,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3051229508196721,
+      "grad_norm": 5.786971092224121,
+      "learning_rate": 1.628813013262218e-05,
+      "loss": 1.4728,
+      "step": 1489
+    },
+    {
+      "epoch": 0.305327868852459,
+      "grad_norm": 3.9484386444091797,
+      "learning_rate": 1.6282967611173536e-05,
+      "loss": 0.7511,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3055327868852459,
+      "grad_norm": 3.9436614513397217,
+      "learning_rate": 1.6277802321562642e-05,
+      "loss": 1.2232,
+      "step": 1491
+    },
+    {
+      "epoch": 0.30573770491803276,
+      "grad_norm": 5.202317714691162,
+      "learning_rate": 1.627263426606523e-05,
+      "loss": 1.1565,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3059426229508197,
+      "grad_norm": 2.704188823699951,
+      "learning_rate": 1.626746344695826e-05,
+      "loss": 0.7545,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3061475409836066,
+      "grad_norm": 2.8272249698638916,
+      "learning_rate": 1.6262289866519897e-05,
+      "loss": 1.1963,
+      "step": 1494
+    },
+    {
+      "epoch": 0.30635245901639346,
+      "grad_norm": 4.482245922088623,
+      "learning_rate": 1.6257113527029527e-05,
+      "loss": 0.8387,
+      "step": 1495
+    },
+    {
+      "epoch": 0.30655737704918035,
+      "grad_norm": 3.2814037799835205,
+      "learning_rate": 1.6251934430767752e-05,
+      "loss": 0.9106,
+      "step": 1496
+    },
+    {
+      "epoch": 0.30676229508196723,
+      "grad_norm": 2.223787784576416,
+      "learning_rate": 1.624675258001639e-05,
+      "loss": 0.8855,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3069672131147541,
+      "grad_norm": 4.34966516494751,
+      "learning_rate": 1.624156797705847e-05,
+      "loss": 1.2336,
+      "step": 1498
+    },
+    {
+      "epoch": 0.307172131147541,
+      "grad_norm": 6.714766025543213,
+      "learning_rate": 1.623638062417823e-05,
+      "loss": 1.2082,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3073770491803279,
+      "grad_norm": 3.4543662071228027,
+      "learning_rate": 1.623119052366114e-05,
+      "loss": 1.0287,
+      "step": 1500
+    },
+    {
+      "epoch": 0.30758196721311476,
+      "grad_norm": 7.2339768409729,
+      "learning_rate": 1.6225997677793842e-05,
+      "loss": 0.9308,
+      "step": 1501
+    },
+    {
+      "epoch": 0.30778688524590164,
+      "grad_norm": 3.6243691444396973,
+      "learning_rate": 1.6220802088864227e-05,
+      "loss": 1.2738,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3079918032786885,
+      "grad_norm": 7.669528007507324,
+      "learning_rate": 1.6215603759161372e-05,
+      "loss": 0.962,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3081967213114754,
+      "grad_norm": 6.140476703643799,
+      "learning_rate": 1.6210402690975565e-05,
+      "loss": 1.0663,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3084016393442623,
+      "grad_norm": 5.582521915435791,
+      "learning_rate": 1.620519888659831e-05,
+      "loss": 1.0742,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3086065573770492,
+      "grad_norm": 3.901350259780884,
+      "learning_rate": 1.6199992348322305e-05,
+      "loss": 0.9826,
+      "step": 1506
+    },
+    {
+      "epoch": 0.30881147540983606,
+      "grad_norm": 4.535983085632324,
+      "learning_rate": 1.6194783078441458e-05,
+      "loss": 1.0895,
+      "step": 1507
+    },
+    {
+      "epoch": 0.30901639344262294,
+      "grad_norm": 3.56119966506958,
+      "learning_rate": 1.6189571079250875e-05,
+      "loss": 0.7785,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3092213114754098,
+      "grad_norm": 6.940043926239014,
+      "learning_rate": 1.6184356353046873e-05,
+      "loss": 1.2064,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3094262295081967,
+      "grad_norm": 4.471426963806152,
+      "learning_rate": 1.6179138902126968e-05,
+      "loss": 1.0183,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3096311475409836,
+      "grad_norm": 6.458265781402588,
+      "learning_rate": 1.617391872878987e-05,
+      "loss": 1.0406,
+      "step": 1511
+    },
+    {
+      "epoch": 0.30983606557377047,
+      "grad_norm": 7.136669158935547,
+      "learning_rate": 1.6168695835335496e-05,
+      "loss": 1.432,
+      "step": 1512
+    },
+    {
+      "epoch": 0.31004098360655735,
+      "grad_norm": 5.144681930541992,
+      "learning_rate": 1.6163470224064958e-05,
+      "loss": 1.2814,
+      "step": 1513
+    },
+    {
+      "epoch": 0.31024590163934423,
+      "grad_norm": 8.28431510925293,
+      "learning_rate": 1.6158241897280564e-05,
+      "loss": 1.1555,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3104508196721312,
+      "grad_norm": 4.177076816558838,
+      "learning_rate": 1.615301085728583e-05,
+      "loss": 0.7929,
+      "step": 1515
+    },
+    {
+      "epoch": 0.31065573770491806,
+      "grad_norm": 4.331387519836426,
+      "learning_rate": 1.6147777106385446e-05,
+      "loss": 0.9688,
+      "step": 1516
+    },
+    {
+      "epoch": 0.31086065573770494,
+      "grad_norm": 6.7804951667785645,
+      "learning_rate": 1.6142540646885316e-05,
+      "loss": 0.8127,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3110655737704918,
+      "grad_norm": 3.6099889278411865,
+      "learning_rate": 1.6137301481092523e-05,
+      "loss": 1.3071,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3112704918032787,
+      "grad_norm": 5.460268020629883,
+      "learning_rate": 1.6132059611315352e-05,
+      "loss": 1.1635,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3114754098360656,
+      "grad_norm": 3.7679054737091064,
+      "learning_rate": 1.6126815039863282e-05,
+      "loss": 1.3164,
+      "step": 1520
+    },
+    {
+      "epoch": 0.31168032786885247,
+      "grad_norm": 16.212350845336914,
+      "learning_rate": 1.612156776904697e-05,
+      "loss": 0.973,
+      "step": 1521
+    },
+    {
+      "epoch": 0.31188524590163935,
+      "grad_norm": 4.7931809425354,
+      "learning_rate": 1.6116317801178273e-05,
+      "loss": 1.0667,
+      "step": 1522
+    },
+    {
+      "epoch": 0.31209016393442623,
+      "grad_norm": 5.537519454956055,
+      "learning_rate": 1.6111065138570223e-05,
+      "loss": 1.0779,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3122950819672131,
+      "grad_norm": 5.860342025756836,
+      "learning_rate": 1.6105809783537063e-05,
+      "loss": 1.1347,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 2.4572460651397705,
+      "learning_rate": 1.6100551738394194e-05,
+      "loss": 0.8091,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3127049180327869,
+      "grad_norm": 3.9189109802246094,
+      "learning_rate": 1.6095291005458223e-05,
+      "loss": 1.1946,
+      "step": 1526
+    },
+    {
+      "epoch": 0.31290983606557377,
+      "grad_norm": 3.646432638168335,
+      "learning_rate": 1.6090027587046935e-05,
+      "loss": 1.21,
+      "step": 1527
+    },
+    {
+      "epoch": 0.31311475409836065,
+      "grad_norm": 3.816833734512329,
+      "learning_rate": 1.6084761485479297e-05,
+      "loss": 0.9061,
+      "step": 1528
+    },
+    {
+      "epoch": 0.31331967213114753,
+      "grad_norm": 6.913815975189209,
+      "learning_rate": 1.6079492703075455e-05,
+      "loss": 0.8795,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3135245901639344,
+      "grad_norm": 5.8984856605529785,
+      "learning_rate": 1.607422124215674e-05,
+      "loss": 1.2641,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3137295081967213,
+      "grad_norm": 4.375982761383057,
+      "learning_rate": 1.606894710504566e-05,
+      "loss": 1.1054,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3139344262295082,
+      "grad_norm": 4.296959400177002,
+      "learning_rate": 1.6063670294065914e-05,
+      "loss": 1.2035,
+      "step": 1532
+    },
+    {
+      "epoch": 0.31413934426229506,
+      "grad_norm": 4.847105979919434,
+      "learning_rate": 1.6058390811542363e-05,
+      "loss": 0.9165,
+      "step": 1533
+    },
+    {
+      "epoch": 0.31434426229508194,
+      "grad_norm": 3.7259321212768555,
+      "learning_rate": 1.6053108659801053e-05,
+      "loss": 1.0118,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3145491803278688,
+      "grad_norm": 4.420511722564697,
+      "learning_rate": 1.6047823841169204e-05,
+      "loss": 1.1301,
+      "step": 1535
+    },
+    {
+      "epoch": 0.31475409836065577,
+      "grad_norm": 3.4586246013641357,
+      "learning_rate": 1.6042536357975215e-05,
+      "loss": 1.1734,
+      "step": 1536
+    },
+    {
+      "epoch": 0.31495901639344265,
+      "grad_norm": 4.04399299621582,
+      "learning_rate": 1.6037246212548653e-05,
+      "loss": 0.9543,
+      "step": 1537
+    },
+    {
+      "epoch": 0.31516393442622953,
+      "grad_norm": 3.43122935295105,
+      "learning_rate": 1.6031953407220257e-05,
+      "loss": 0.9199,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3153688524590164,
+      "grad_norm": 3.4176876544952393,
+      "learning_rate": 1.6026657944321948e-05,
+      "loss": 1.0061,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3155737704918033,
+      "grad_norm": 6.385382175445557,
+      "learning_rate": 1.6021359826186812e-05,
+      "loss": 1.0098,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3157786885245902,
+      "grad_norm": 4.173152923583984,
+      "learning_rate": 1.60160590551491e-05,
+      "loss": 1.2205,
+      "step": 1541
+    },
+    {
+      "epoch": 0.31598360655737706,
+      "grad_norm": 2.601982593536377,
+      "learning_rate": 1.6010755633544235e-05,
+      "loss": 0.912,
+      "step": 1542
+    },
+    {
+      "epoch": 0.31618852459016394,
+      "grad_norm": 3.92033314704895,
+      "learning_rate": 1.6005449563708814e-05,
+      "loss": 0.9169,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3163934426229508,
+      "grad_norm": 6.487717628479004,
+      "learning_rate": 1.6000140847980593e-05,
+      "loss": 1.1648,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3165983606557377,
+      "grad_norm": 4.249373435974121,
+      "learning_rate": 1.59948294886985e-05,
+      "loss": 1.0025,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3168032786885246,
+      "grad_norm": 2.7345590591430664,
+      "learning_rate": 1.5989515488202616e-05,
+      "loss": 0.7966,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3170081967213115,
+      "grad_norm": 7.052485942840576,
+      "learning_rate": 1.59841988488342e-05,
+      "loss": 1.0935,
+      "step": 1547
+    },
+    {
+      "epoch": 0.31721311475409836,
+      "grad_norm": 3.8757991790771484,
+      "learning_rate": 1.5978879572935667e-05,
+      "loss": 1.1786,
+      "step": 1548
+    },
+    {
+      "epoch": 0.31741803278688524,
+      "grad_norm": 2.8870551586151123,
+      "learning_rate": 1.5973557662850594e-05,
+      "loss": 0.9974,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3176229508196721,
+      "grad_norm": 4.774921894073486,
+      "learning_rate": 1.596823312092372e-05,
+      "loss": 1.0259,
+      "step": 1550
+    },
+    {
+      "epoch": 0.317827868852459,
+      "grad_norm": 4.69555139541626,
+      "learning_rate": 1.596290594950093e-05,
+      "loss": 1.0509,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3180327868852459,
+      "grad_norm": 5.341336250305176,
+      "learning_rate": 1.59575761509293e-05,
+      "loss": 1.0934,
+      "step": 1552
+    },
+    {
+      "epoch": 0.31823770491803277,
+      "grad_norm": 4.593174934387207,
+      "learning_rate": 1.595224372755703e-05,
+      "loss": 1.1307,
+      "step": 1553
+    },
+    {
+      "epoch": 0.31844262295081965,
+      "grad_norm": 3.744046211242676,
+      "learning_rate": 1.594690868173349e-05,
+      "loss": 0.7657,
+      "step": 1554
+    },
+    {
+      "epoch": 0.31864754098360654,
+      "grad_norm": 3.535003900527954,
+      "learning_rate": 1.5941571015809207e-05,
+      "loss": 0.9444,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3188524590163934,
+      "grad_norm": 5.941051006317139,
+      "learning_rate": 1.5936230732135866e-05,
+      "loss": 1.1247,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3190573770491803,
+      "grad_norm": 4.069591522216797,
+      "learning_rate": 1.593088783306629e-05,
+      "loss": 1.0507,
+      "step": 1557
+    },
+    {
+      "epoch": 0.31926229508196724,
+      "grad_norm": 3.661802053451538,
+      "learning_rate": 1.592554232095447e-05,
+      "loss": 0.9364,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3194672131147541,
+      "grad_norm": 4.624448776245117,
+      "learning_rate": 1.5920194198155542e-05,
+      "loss": 1.0744,
+      "step": 1559
+    },
+    {
+      "epoch": 0.319672131147541,
+      "grad_norm": 4.251600742340088,
+      "learning_rate": 1.5914843467025788e-05,
+      "loss": 1.204,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3198770491803279,
+      "grad_norm": 7.206779479980469,
+      "learning_rate": 1.590949012992265e-05,
+      "loss": 1.3895,
+      "step": 1561
+    },
+    {
+      "epoch": 0.32008196721311477,
+      "grad_norm": 3.1636006832122803,
+      "learning_rate": 1.5904134189204706e-05,
+      "loss": 1.1693,
+      "step": 1562
+    },
+    {
+      "epoch": 0.32028688524590165,
+      "grad_norm": 4.9473090171813965,
+      "learning_rate": 1.589877564723169e-05,
+      "loss": 1.0163,
+      "step": 1563
+    },
+    {
+      "epoch": 0.32049180327868854,
+      "grad_norm": 3.6467862129211426,
+      "learning_rate": 1.5893414506364483e-05,
+      "loss": 0.912,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3206967213114754,
+      "grad_norm": 3.2556416988372803,
+      "learning_rate": 1.58880507689651e-05,
+      "loss": 1.2952,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3209016393442623,
+      "grad_norm": 10.005574226379395,
+      "learning_rate": 1.5882684437396707e-05,
+      "loss": 1.0921,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3211065573770492,
+      "grad_norm": 5.045826435089111,
+      "learning_rate": 1.587731551402362e-05,
+      "loss": 1.1582,
+      "step": 1567
+    },
+    {
+      "epoch": 0.32131147540983607,
+      "grad_norm": 4.519015312194824,
+      "learning_rate": 1.587194400121128e-05,
+      "loss": 1.1204,
+      "step": 1568
+    },
+    {
+      "epoch": 0.32151639344262295,
+      "grad_norm": 6.881636142730713,
+      "learning_rate": 1.586656990132629e-05,
+      "loss": 1.1205,
+      "step": 1569
+    },
+    {
+      "epoch": 0.32172131147540983,
+      "grad_norm": 5.11054801940918,
+      "learning_rate": 1.5861193216736372e-05,
+      "loss": 1.0007,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3219262295081967,
+      "grad_norm": 3.2122952938079834,
+      "learning_rate": 1.5855813949810397e-05,
+      "loss": 1.077,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3221311475409836,
+      "grad_norm": 5.064713001251221,
+      "learning_rate": 1.5850432102918383e-05,
+      "loss": 1.2241,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3223360655737705,
+      "grad_norm": 4.348764896392822,
+      "learning_rate": 1.5845047678431463e-05,
+      "loss": 0.7306,
+      "step": 1573
+    },
+    {
+      "epoch": 0.32254098360655736,
+      "grad_norm": 3.8575494289398193,
+      "learning_rate": 1.583966067872192e-05,
+      "loss": 1.2201,
+      "step": 1574
+    },
+    {
+      "epoch": 0.32274590163934425,
+      "grad_norm": 6.508906364440918,
+      "learning_rate": 1.583427110616318e-05,
+      "loss": 1.0227,
+      "step": 1575
+    },
+    {
+      "epoch": 0.32295081967213113,
+      "grad_norm": 10.230183601379395,
+      "learning_rate": 1.5828878963129773e-05,
+      "loss": 1.1368,
+      "step": 1576
+    },
+    {
+      "epoch": 0.323155737704918,
+      "grad_norm": 2.2485198974609375,
+      "learning_rate": 1.5823484251997392e-05,
+      "loss": 1.0312,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3233606557377049,
+      "grad_norm": 4.690741062164307,
+      "learning_rate": 1.5818086975142847e-05,
+      "loss": 1.0954,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3235655737704918,
+      "grad_norm": 5.363114356994629,
+      "learning_rate": 1.581268713494408e-05,
+      "loss": 1.0031,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3237704918032787,
+      "grad_norm": 4.422763347625732,
+      "learning_rate": 1.580728473378017e-05,
+      "loss": 1.0433,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3239754098360656,
+      "grad_norm": 9.148633003234863,
+      "learning_rate": 1.5801879774031302e-05,
+      "loss": 0.9592,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3241803278688525,
+      "grad_norm": 5.123961448669434,
+      "learning_rate": 1.579647225807882e-05,
+      "loss": 0.9251,
+      "step": 1582
+    },
+    {
+      "epoch": 0.32438524590163936,
+      "grad_norm": 5.855801582336426,
+      "learning_rate": 1.579106218830517e-05,
+      "loss": 0.8595,
+      "step": 1583
+    },
+    {
+      "epoch": 0.32459016393442625,
+      "grad_norm": 5.355648517608643,
+      "learning_rate": 1.5785649567093933e-05,
+      "loss": 0.7885,
+      "step": 1584
+    },
+    {
+      "epoch": 0.32479508196721313,
+      "grad_norm": 4.017860412597656,
+      "learning_rate": 1.578023439682981e-05,
+      "loss": 1.1038,
+      "step": 1585
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 3.505661964416504,
+      "learning_rate": 1.577481667989863e-05,
+      "loss": 1.1429,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3252049180327869,
+      "grad_norm": 9.889689445495605,
+      "learning_rate": 1.576939641868734e-05,
+      "loss": 1.2028,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3254098360655738,
+      "grad_norm": 4.381469249725342,
+      "learning_rate": 1.576397361558401e-05,
+      "loss": 0.8069,
+      "step": 1588
+    },
+    {
+      "epoch": 0.32561475409836066,
+      "grad_norm": 5.1572346687316895,
+      "learning_rate": 1.575854827297783e-05,
+      "loss": 1.2147,
+      "step": 1589
+    },
+    {
+      "epoch": 0.32581967213114754,
+      "grad_norm": 3.952069044113159,
+      "learning_rate": 1.5753120393259106e-05,
+      "loss": 1.0871,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3260245901639344,
+      "grad_norm": 3.5121958255767822,
+      "learning_rate": 1.5747689978819268e-05,
+      "loss": 0.8929,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3262295081967213,
+      "grad_norm": 4.054090976715088,
+      "learning_rate": 1.5742257032050855e-05,
+      "loss": 0.9341,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3264344262295082,
+      "grad_norm": 4.42818021774292,
+      "learning_rate": 1.5736821555347525e-05,
+      "loss": 1.0258,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3266393442622951,
+      "grad_norm": 6.482116222381592,
+      "learning_rate": 1.5731383551104054e-05,
+      "loss": 0.9906,
+      "step": 1594
+    },
+    {
+      "epoch": 0.32684426229508196,
+      "grad_norm": 4.349967002868652,
+      "learning_rate": 1.5725943021716324e-05,
+      "loss": 0.9788,
+      "step": 1595
+    },
+    {
+      "epoch": 0.32704918032786884,
+      "grad_norm": 2.5290989875793457,
+      "learning_rate": 1.5720499969581342e-05,
+      "loss": 1.0177,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3272540983606557,
+      "grad_norm": 2.5364205837249756,
+      "learning_rate": 1.571505439709721e-05,
+      "loss": 0.9134,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3274590163934426,
+      "grad_norm": 3.8951873779296875,
+      "learning_rate": 1.5709606306663155e-05,
+      "loss": 1.073,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3276639344262295,
+      "grad_norm": 2.6915369033813477,
+      "learning_rate": 1.5704155700679506e-05,
+      "loss": 0.9337,
+      "step": 1599
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 5.243301868438721,
+      "learning_rate": 1.56987025815477e-05,
+      "loss": 1.0919,
+      "step": 1600
+    },
+    {
+      "epoch": 0.32807377049180325,
+      "grad_norm": 7.807430267333984,
+      "learning_rate": 1.5693246951670285e-05,
+      "loss": 1.2742,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3282786885245902,
+      "grad_norm": 4.802052974700928,
+      "learning_rate": 1.568778881345091e-05,
+      "loss": 1.1545,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3284836065573771,
+      "grad_norm": 7.042609214782715,
+      "learning_rate": 1.5682328169294338e-05,
+      "loss": 1.3581,
+      "step": 1603
+    },
+    {
+      "epoch": 0.32868852459016396,
+      "grad_norm": 3.372318983078003,
+      "learning_rate": 1.567686502160642e-05,
+      "loss": 1.2029,
+      "step": 1604
+    },
+    {
+      "epoch": 0.32889344262295084,
+      "grad_norm": 2.7943155765533447,
+      "learning_rate": 1.5671399372794133e-05,
+      "loss": 0.7666,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3290983606557377,
+      "grad_norm": 3.658761978149414,
+      "learning_rate": 1.566593122526554e-05,
+      "loss": 0.8424,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3293032786885246,
+      "grad_norm": 4.309822082519531,
+      "learning_rate": 1.56604605814298e-05,
+      "loss": 0.9281,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3295081967213115,
+      "grad_norm": 4.732679843902588,
+      "learning_rate": 1.5654987443697186e-05,
+      "loss": 0.9498,
+      "step": 1608
+    },
+    {
+      "epoch": 0.32971311475409837,
+      "grad_norm": 3.180725574493408,
+      "learning_rate": 1.5649511814479065e-05,
+      "loss": 1.0048,
+      "step": 1609
+    },
+    {
+      "epoch": 0.32991803278688525,
+      "grad_norm": 4.041045188903809,
+      "learning_rate": 1.56440336961879e-05,
+      "loss": 1.1154,
+      "step": 1610
+    },
+    {
+      "epoch": 0.33012295081967213,
+      "grad_norm": 7.198296070098877,
+      "learning_rate": 1.563855309123725e-05,
+      "loss": 1.0119,
+      "step": 1611
+    },
+    {
+      "epoch": 0.330327868852459,
+      "grad_norm": 3.7422866821289062,
+      "learning_rate": 1.5633070002041772e-05,
+      "loss": 1.1168,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3305327868852459,
+      "grad_norm": 5.694290637969971,
+      "learning_rate": 1.5627584431017213e-05,
+      "loss": 1.1231,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3307377049180328,
+      "grad_norm": 4.079566478729248,
+      "learning_rate": 1.562209638058042e-05,
+      "loss": 0.9625,
+      "step": 1614
+    },
+    {
+      "epoch": 0.33094262295081966,
+      "grad_norm": 5.550624370574951,
+      "learning_rate": 1.561660585314933e-05,
+      "loss": 0.9169,
+      "step": 1615
+    },
+    {
+      "epoch": 0.33114754098360655,
+      "grad_norm": 6.9668965339660645,
+      "learning_rate": 1.561111285114296e-05,
+      "loss": 1.0023,
+      "step": 1616
+    },
+    {
+      "epoch": 0.33135245901639343,
+      "grad_norm": 3.927203416824341,
+      "learning_rate": 1.5605617376981447e-05,
+      "loss": 1.0184,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3315573770491803,
+      "grad_norm": 7.020673751831055,
+      "learning_rate": 1.560011943308598e-05,
+      "loss": 1.4712,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3317622950819672,
+      "grad_norm": 5.7177581787109375,
+      "learning_rate": 1.5594619021878864e-05,
+      "loss": 1.1607,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3319672131147541,
+      "grad_norm": 4.598093509674072,
+      "learning_rate": 1.5589116145783477e-05,
+      "loss": 1.0162,
+      "step": 1620
+    },
+    {
+      "epoch": 0.33217213114754096,
+      "grad_norm": 4.581098556518555,
+      "learning_rate": 1.5583610807224284e-05,
+      "loss": 1.2057,
+      "step": 1621
+    },
+    {
+      "epoch": 0.33237704918032784,
+      "grad_norm": 3.677827835083008,
+      "learning_rate": 1.5578103008626848e-05,
+      "loss": 1.1234,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3325819672131147,
+      "grad_norm": 4.11728572845459,
+      "learning_rate": 1.5572592752417797e-05,
+      "loss": 1.002,
+      "step": 1623
+    },
+    {
+      "epoch": 0.33278688524590166,
+      "grad_norm": 3.994401454925537,
+      "learning_rate": 1.5567080041024852e-05,
+      "loss": 0.9705,
+      "step": 1624
+    },
+    {
+      "epoch": 0.33299180327868855,
+      "grad_norm": 5.835907936096191,
+      "learning_rate": 1.5561564876876817e-05,
+      "loss": 0.9609,
+      "step": 1625
+    },
+    {
+      "epoch": 0.33319672131147543,
+      "grad_norm": 3.244267225265503,
+      "learning_rate": 1.555604726240357e-05,
+      "loss": 1.0457,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3334016393442623,
+      "grad_norm": 7.488487243652344,
+      "learning_rate": 1.5550527200036077e-05,
+      "loss": 0.8822,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3336065573770492,
+      "grad_norm": 4.826598644256592,
+      "learning_rate": 1.5545004692206372e-05,
+      "loss": 0.9069,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3338114754098361,
+      "grad_norm": 3.0855965614318848,
+      "learning_rate": 1.5539479741347578e-05,
+      "loss": 0.8744,
+      "step": 1629
+    },
+    {
+      "epoch": 0.33401639344262296,
+      "grad_norm": 5.414949417114258,
+      "learning_rate": 1.5533952349893887e-05,
+      "loss": 0.9079,
+      "step": 1630
+    },
+    {
+      "epoch": 0.33422131147540984,
+      "grad_norm": 8.861950874328613,
+      "learning_rate": 1.5528422520280568e-05,
+      "loss": 1.1132,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3344262295081967,
+      "grad_norm": 7.041595458984375,
+      "learning_rate": 1.552289025494396e-05,
+      "loss": 1.0636,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3346311475409836,
+      "grad_norm": 4.759159088134766,
+      "learning_rate": 1.5517355556321487e-05,
+      "loss": 1.0135,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3348360655737705,
+      "grad_norm": 4.959380149841309,
+      "learning_rate": 1.5511818426851634e-05,
+      "loss": 1.1825,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3350409836065574,
+      "grad_norm": 8.271422386169434,
+      "learning_rate": 1.550627886897396e-05,
+      "loss": 0.8507,
+      "step": 1635
+    },
+    {
+      "epoch": 0.33524590163934426,
+      "grad_norm": 2.148503303527832,
+      "learning_rate": 1.550073688512909e-05,
+      "loss": 0.9812,
+      "step": 1636
+    },
+    {
+      "epoch": 0.33545081967213114,
+      "grad_norm": 5.6088457107543945,
+      "learning_rate": 1.5495192477758734e-05,
+      "loss": 0.9063,
+      "step": 1637
+    },
+    {
+      "epoch": 0.335655737704918,
+      "grad_norm": 4.83995246887207,
+      "learning_rate": 1.5489645649305643e-05,
+      "loss": 0.9637,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3358606557377049,
+      "grad_norm": 4.5143537521362305,
+      "learning_rate": 1.548409640221366e-05,
+      "loss": 1.2042,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3360655737704918,
+      "grad_norm": 12.368703842163086,
+      "learning_rate": 1.5478544738927682e-05,
+      "loss": 1.2205,
+      "step": 1640
+    },
+    {
+      "epoch": 0.33627049180327867,
+      "grad_norm": 4.488269329071045,
+      "learning_rate": 1.547299066189367e-05,
+      "loss": 1.0421,
+      "step": 1641
+    },
+    {
+      "epoch": 0.33647540983606555,
+      "grad_norm": 5.847134590148926,
+      "learning_rate": 1.546743417355865e-05,
+      "loss": 1.2818,
+      "step": 1642
+    },
+    {
+      "epoch": 0.33668032786885244,
+      "grad_norm": 4.768396377563477,
+      "learning_rate": 1.546187527637071e-05,
+      "loss": 1.1176,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3368852459016393,
+      "grad_norm": 5.81901741027832,
+      "learning_rate": 1.5456313972779002e-05,
+      "loss": 1.1489,
+      "step": 1644
+    },
+    {
+      "epoch": 0.33709016393442626,
+      "grad_norm": 5.801312446594238,
+      "learning_rate": 1.5450750265233732e-05,
+      "loss": 1.2489,
+      "step": 1645
+    },
+    {
+      "epoch": 0.33729508196721314,
+      "grad_norm": 10.421825408935547,
+      "learning_rate": 1.5445184156186176e-05,
+      "loss": 1.5553,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 2.7016472816467285,
+      "learning_rate": 1.5439615648088653e-05,
+      "loss": 1.2488,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3377049180327869,
+      "grad_norm": 6.601657867431641,
+      "learning_rate": 1.5434044743394555e-05,
+      "loss": 0.91,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3379098360655738,
+      "grad_norm": 4.80570125579834,
+      "learning_rate": 1.5428471444558314e-05,
+      "loss": 0.8818,
+      "step": 1649
+    },
+    {
+      "epoch": 0.33811475409836067,
+      "grad_norm": 5.058017730712891,
+      "learning_rate": 1.5422895754035435e-05,
+      "loss": 1.1512,
+      "step": 1650
+    },
+    {
+      "epoch": 0.33831967213114755,
+      "grad_norm": 4.141129493713379,
+      "learning_rate": 1.541731767428246e-05,
+      "loss": 0.9459,
+      "step": 1651
+    },
+    {
+      "epoch": 0.33852459016393444,
+      "grad_norm": 3.152006149291992,
+      "learning_rate": 1.541173720775699e-05,
+      "loss": 0.6653,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3387295081967213,
+      "grad_norm": 5.508606433868408,
+      "learning_rate": 1.540615435691768e-05,
+      "loss": 1.2321,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3389344262295082,
+      "grad_norm": 5.269017219543457,
+      "learning_rate": 1.5400569124224237e-05,
+      "loss": 0.9167,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3391393442622951,
+      "grad_norm": 5.636336803436279,
+      "learning_rate": 1.5394981512137406e-05,
+      "loss": 1.0427,
+      "step": 1655
+    },
+    {
+      "epoch": 0.33934426229508197,
+      "grad_norm": 4.1122636795043945,
+      "learning_rate": 1.5389391523118994e-05,
+      "loss": 1.121,
+      "step": 1656
+    },
+    {
+      "epoch": 0.33954918032786885,
+      "grad_norm": 3.3682305812835693,
+      "learning_rate": 1.5383799159631843e-05,
+      "loss": 0.8537,
+      "step": 1657
+    },
+    {
+      "epoch": 0.33975409836065573,
+      "grad_norm": 9.52736759185791,
+      "learning_rate": 1.5378204424139858e-05,
+      "loss": 1.0313,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3399590163934426,
+      "grad_norm": 5.0220537185668945,
+      "learning_rate": 1.5372607319107972e-05,
+      "loss": 1.0343,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3401639344262295,
+      "grad_norm": 2.9400594234466553,
+      "learning_rate": 1.5367007847002166e-05,
+      "loss": 1.0002,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3403688524590164,
+      "grad_norm": 3.446505546569824,
+      "learning_rate": 1.5361406010289473e-05,
+      "loss": 1.1774,
+      "step": 1661
+    },
+    {
+      "epoch": 0.34057377049180326,
+      "grad_norm": 3.818256139755249,
+      "learning_rate": 1.5355801811437955e-05,
+      "loss": 0.9029,
+      "step": 1662
+    },
+    {
+      "epoch": 0.34077868852459015,
+      "grad_norm": 4.395934581756592,
+      "learning_rate": 1.5350195252916724e-05,
+      "loss": 0.9768,
+      "step": 1663
+    },
+    {
+      "epoch": 0.34098360655737703,
+      "grad_norm": 5.355147838592529,
+      "learning_rate": 1.5344586337195933e-05,
+      "loss": 1.0172,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3411885245901639,
+      "grad_norm": 4.839866638183594,
+      "learning_rate": 1.5338975066746757e-05,
+      "loss": 0.8432,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3413934426229508,
+      "grad_norm": 6.035134315490723,
+      "learning_rate": 1.533336144404143e-05,
+      "loss": 1.051,
+      "step": 1666
+    },
+    {
+      "epoch": 0.34159836065573773,
+      "grad_norm": 2.459001064300537,
+      "learning_rate": 1.5327745471553215e-05,
+      "loss": 0.8306,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3418032786885246,
+      "grad_norm": 5.235065460205078,
+      "learning_rate": 1.5322127151756406e-05,
+      "loss": 0.9515,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3420081967213115,
+      "grad_norm": 3.2183752059936523,
+      "learning_rate": 1.531650648712633e-05,
+      "loss": 0.9745,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3422131147540984,
+      "grad_norm": 5.042847156524658,
+      "learning_rate": 1.531088348013935e-05,
+      "loss": 1.1558,
+      "step": 1670
+    },
+    {
+      "epoch": 0.34241803278688526,
+      "grad_norm": 4.79826021194458,
+      "learning_rate": 1.530525813327286e-05,
+      "loss": 1.0881,
+      "step": 1671
+    },
+    {
+      "epoch": 0.34262295081967215,
+      "grad_norm": 9.2280912399292,
+      "learning_rate": 1.5299630449005295e-05,
+      "loss": 0.8319,
+      "step": 1672
+    },
+    {
+      "epoch": 0.34282786885245903,
+      "grad_norm": 5.522466659545898,
+      "learning_rate": 1.5294000429816108e-05,
+      "loss": 1.0348,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3430327868852459,
+      "grad_norm": 3.753108501434326,
+      "learning_rate": 1.528836807818578e-05,
+      "loss": 0.8663,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3432377049180328,
+      "grad_norm": 3.5305533409118652,
+      "learning_rate": 1.5282733396595824e-05,
+      "loss": 1.1317,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3434426229508197,
+      "grad_norm": 3.194035768508911,
+      "learning_rate": 1.5277096387528784e-05,
+      "loss": 1.0888,
+      "step": 1676
+    },
+    {
+      "epoch": 0.34364754098360656,
+      "grad_norm": 5.8319902420043945,
+      "learning_rate": 1.5271457053468223e-05,
+      "loss": 1.0642,
+      "step": 1677
+    },
+    {
+      "epoch": 0.34385245901639344,
+      "grad_norm": 2.9311530590057373,
+      "learning_rate": 1.5265815396898724e-05,
+      "loss": 0.9332,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3440573770491803,
+      "grad_norm": 3.390526056289673,
+      "learning_rate": 1.5260171420305904e-05,
+      "loss": 0.7603,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3442622950819672,
+      "grad_norm": 3.7307345867156982,
+      "learning_rate": 1.52545251261764e-05,
+      "loss": 0.7987,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3444672131147541,
+      "grad_norm": 5.712270736694336,
+      "learning_rate": 1.5248876516997864e-05,
+      "loss": 1.1034,
+      "step": 1681
+    },
+    {
+      "epoch": 0.34467213114754097,
+      "grad_norm": 3.958359718322754,
+      "learning_rate": 1.524322559525897e-05,
+      "loss": 1.043,
+      "step": 1682
+    },
+    {
+      "epoch": 0.34487704918032785,
+      "grad_norm": 6.307284355163574,
+      "learning_rate": 1.5237572363449418e-05,
+      "loss": 1.0801,
+      "step": 1683
+    },
+    {
+      "epoch": 0.34508196721311474,
+      "grad_norm": 4.124212741851807,
+      "learning_rate": 1.5231916824059912e-05,
+      "loss": 1.0321,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3452868852459016,
+      "grad_norm": 5.210524559020996,
+      "learning_rate": 1.5226258979582182e-05,
+      "loss": 1.1229,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3454918032786885,
+      "grad_norm": 3.2006444931030273,
+      "learning_rate": 1.5220598832508979e-05,
+      "loss": 1.1404,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3456967213114754,
+      "grad_norm": 5.683770656585693,
+      "learning_rate": 1.5214936385334059e-05,
+      "loss": 1.1884,
+      "step": 1687
+    },
+    {
+      "epoch": 0.34590163934426227,
+      "grad_norm": 3.0015242099761963,
+      "learning_rate": 1.5209271640552187e-05,
+      "loss": 1.3054,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3461065573770492,
+      "grad_norm": 9.5305814743042,
+      "learning_rate": 1.5203604600659161e-05,
+      "loss": 1.0417,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3463114754098361,
+      "grad_norm": 7.119845867156982,
+      "learning_rate": 1.5197935268151764e-05,
+      "loss": 1.1983,
+      "step": 1690
+    },
+    {
+      "epoch": 0.34651639344262297,
+      "grad_norm": 7.892723560333252,
+      "learning_rate": 1.5192263645527808e-05,
+      "loss": 1.3109,
+      "step": 1691
+    },
+    {
+      "epoch": 0.34672131147540985,
+      "grad_norm": 9.858922958374023,
+      "learning_rate": 1.5186589735286104e-05,
+      "loss": 1.0686,
+      "step": 1692
+    },
+    {
+      "epoch": 0.34692622950819674,
+      "grad_norm": 2.673138380050659,
+      "learning_rate": 1.5180913539926478e-05,
+      "loss": 0.7855,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3471311475409836,
+      "grad_norm": 3.808634042739868,
+      "learning_rate": 1.517523506194976e-05,
+      "loss": 1.0301,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3473360655737705,
+      "grad_norm": 8.06158447265625,
+      "learning_rate": 1.5169554303857778e-05,
+      "loss": 1.1626,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3475409836065574,
+      "grad_norm": 6.041512966156006,
+      "learning_rate": 1.5163871268153383e-05,
+      "loss": 0.9651,
+      "step": 1696
+    },
+    {
+      "epoch": 0.34774590163934427,
+      "grad_norm": 3.4610378742218018,
+      "learning_rate": 1.5158185957340407e-05,
+      "loss": 1.0012,
+      "step": 1697
+    },
+    {
+      "epoch": 0.34795081967213115,
+      "grad_norm": 5.7460551261901855,
+      "learning_rate": 1.5152498373923701e-05,
+      "loss": 1.0979,
+      "step": 1698
+    },
+    {
+      "epoch": 0.34815573770491803,
+      "grad_norm": 2.8422229290008545,
+      "learning_rate": 1.5146808520409115e-05,
+      "loss": 0.9517,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3483606557377049,
+      "grad_norm": 4.968901634216309,
+      "learning_rate": 1.5141116399303486e-05,
+      "loss": 0.9126,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3485655737704918,
+      "grad_norm": 4.709219932556152,
+      "learning_rate": 1.513542201311467e-05,
+      "loss": 0.9952,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3487704918032787,
+      "grad_norm": 3.778381586074829,
+      "learning_rate": 1.5129725364351504e-05,
+      "loss": 0.9276,
+      "step": 1702
+    },
+    {
+      "epoch": 0.34897540983606556,
+      "grad_norm": 4.007264614105225,
+      "learning_rate": 1.512402645552383e-05,
+      "loss": 1.1355,
+      "step": 1703
+    },
+    {
+      "epoch": 0.34918032786885245,
+      "grad_norm": 5.361089706420898,
+      "learning_rate": 1.511832528914249e-05,
+      "loss": 1.0928,
+      "step": 1704
+    },
+    {
+      "epoch": 0.34938524590163933,
+      "grad_norm": 5.3983378410339355,
+      "learning_rate": 1.5112621867719309e-05,
+      "loss": 0.9805,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3495901639344262,
+      "grad_norm": 4.910577297210693,
+      "learning_rate": 1.5106916193767114e-05,
+      "loss": 0.837,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3497950819672131,
+      "grad_norm": 3.1997921466827393,
+      "learning_rate": 1.5101208269799718e-05,
+      "loss": 0.6932,
+      "step": 1707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.373333930969238,
+      "learning_rate": 1.5095498098331937e-05,
+      "loss": 1.0926,
+      "step": 1708
+    },
+    {
+      "epoch": 0.35020491803278686,
+      "grad_norm": 4.771115779876709,
+      "learning_rate": 1.5089785681879562e-05,
+      "loss": 1.0445,
+      "step": 1709
+    },
+    {
+      "epoch": 0.35040983606557374,
+      "grad_norm": 3.0325698852539062,
+      "learning_rate": 1.5084071022959388e-05,
+      "loss": 0.9766,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3506147540983607,
+      "grad_norm": 5.2754411697387695,
+      "learning_rate": 1.5078354124089187e-05,
+      "loss": 0.8486,
+      "step": 1711
+    },
+    {
+      "epoch": 0.35081967213114756,
+      "grad_norm": 2.411959171295166,
+      "learning_rate": 1.5072634987787717e-05,
+      "loss": 0.8738,
+      "step": 1712
+    },
+    {
+      "epoch": 0.35102459016393445,
+      "grad_norm": 3.184126853942871,
+      "learning_rate": 1.5066913616574734e-05,
+      "loss": 1.0476,
+      "step": 1713
+    },
+    {
+      "epoch": 0.35122950819672133,
+      "grad_norm": 3.4831783771514893,
+      "learning_rate": 1.5061190012970968e-05,
+      "loss": 1.3668,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3514344262295082,
+      "grad_norm": 4.21259880065918,
+      "learning_rate": 1.5055464179498137e-05,
+      "loss": 1.1655,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3516393442622951,
+      "grad_norm": 4.044133186340332,
+      "learning_rate": 1.5049736118678934e-05,
+      "loss": 1.0329,
+      "step": 1716
+    },
+    {
+      "epoch": 0.351844262295082,
+      "grad_norm": 3.390869617462158,
+      "learning_rate": 1.5044005833037048e-05,
+      "loss": 1.1055,
+      "step": 1717
+    },
+    {
+      "epoch": 0.35204918032786886,
+      "grad_norm": 7.676429271697998,
+      "learning_rate": 1.5038273325097135e-05,
+      "loss": 1.1176,
+      "step": 1718
+    },
+    {
+      "epoch": 0.35225409836065574,
+      "grad_norm": 6.638105392456055,
+      "learning_rate": 1.5032538597384832e-05,
+      "loss": 1.0865,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3524590163934426,
+      "grad_norm": 6.087202548980713,
+      "learning_rate": 1.5026801652426761e-05,
+      "loss": 1.1695,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3526639344262295,
+      "grad_norm": 3.464566230773926,
+      "learning_rate": 1.5021062492750513e-05,
+      "loss": 0.9831,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3528688524590164,
+      "grad_norm": 4.970489978790283,
+      "learning_rate": 1.5015321120884662e-05,
+      "loss": 1.3101,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3530737704918033,
+      "grad_norm": 3.0767815113067627,
+      "learning_rate": 1.5009577539358753e-05,
+      "loss": 1.2025,
+      "step": 1723
+    },
+    {
+      "epoch": 0.35327868852459016,
+      "grad_norm": 6.74249267578125,
+      "learning_rate": 1.5003831750703302e-05,
+      "loss": 1.2026,
+      "step": 1724
+    },
+    {
+      "epoch": 0.35348360655737704,
+      "grad_norm": 8.652679443359375,
+      "learning_rate": 1.4998083757449799e-05,
+      "loss": 0.9747,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3536885245901639,
+      "grad_norm": 6.1603684425354,
+      "learning_rate": 1.4992333562130709e-05,
+      "loss": 1.2541,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3538934426229508,
+      "grad_norm": 4.209448337554932,
+      "learning_rate": 1.4986581167279464e-05,
+      "loss": 1.1016,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3540983606557377,
+      "grad_norm": 4.962892055511475,
+      "learning_rate": 1.4980826575430462e-05,
+      "loss": 1.0307,
+      "step": 1728
+    },
+    {
+      "epoch": 0.35430327868852457,
+      "grad_norm": 5.2751851081848145,
+      "learning_rate": 1.4975069789119079e-05,
+      "loss": 0.9535,
+      "step": 1729
+    },
+    {
+      "epoch": 0.35450819672131145,
+      "grad_norm": 17.011795043945312,
+      "learning_rate": 1.4969310810881644e-05,
+      "loss": 1.3323,
+      "step": 1730
+    },
+    {
+      "epoch": 0.35471311475409834,
+      "grad_norm": 6.225976943969727,
+      "learning_rate": 1.4963549643255464e-05,
+      "loss": 1.2265,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3549180327868853,
+      "grad_norm": 3.561882734298706,
+      "learning_rate": 1.4957786288778801e-05,
+      "loss": 1.2551,
+      "step": 1732
+    },
+    {
+      "epoch": 0.35512295081967216,
+      "grad_norm": 5.198137283325195,
+      "learning_rate": 1.495202074999089e-05,
+      "loss": 1.0752,
+      "step": 1733
+    },
+    {
+      "epoch": 0.35532786885245904,
+      "grad_norm": 5.8098297119140625,
+      "learning_rate": 1.4946253029431923e-05,
+      "loss": 1.1106,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3555327868852459,
+      "grad_norm": 4.293265342712402,
+      "learning_rate": 1.4940483129643048e-05,
+      "loss": 1.1884,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3557377049180328,
+      "grad_norm": 3.984328508377075,
+      "learning_rate": 1.4934711053166386e-05,
+      "loss": 1.0083,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3559426229508197,
+      "grad_norm": 3.804286003112793,
+      "learning_rate": 1.4928936802545004e-05,
+      "loss": 1.1736,
+      "step": 1737
+    },
+    {
+      "epoch": 0.35614754098360657,
+      "grad_norm": 3.8535032272338867,
+      "learning_rate": 1.492316038032294e-05,
+      "loss": 1.014,
+      "step": 1738
+    },
+    {
+      "epoch": 0.35635245901639345,
+      "grad_norm": 12.708379745483398,
+      "learning_rate": 1.4917381789045174e-05,
+      "loss": 1.2603,
+      "step": 1739
+    },
+    {
+      "epoch": 0.35655737704918034,
+      "grad_norm": 6.346560001373291,
+      "learning_rate": 1.4911601031257653e-05,
+      "loss": 1.1145,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3567622950819672,
+      "grad_norm": 3.612197160720825,
+      "learning_rate": 1.4905818109507274e-05,
+      "loss": 1.1868,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3569672131147541,
+      "grad_norm": 4.518301010131836,
+      "learning_rate": 1.4900033026341883e-05,
+      "loss": 0.6992,
+      "step": 1742
+    },
+    {
+      "epoch": 0.357172131147541,
+      "grad_norm": 3.592106342315674,
+      "learning_rate": 1.4894245784310293e-05,
+      "loss": 1.2389,
+      "step": 1743
+    },
+    {
+      "epoch": 0.35737704918032787,
+      "grad_norm": 3.390378475189209,
+      "learning_rate": 1.4888456385962253e-05,
+      "loss": 1.0585,
+      "step": 1744
+    },
+    {
+      "epoch": 0.35758196721311475,
+      "grad_norm": 5.413238525390625,
+      "learning_rate": 1.4882664833848467e-05,
+      "loss": 1.193,
+      "step": 1745
+    },
+    {
+      "epoch": 0.35778688524590163,
+      "grad_norm": 5.5519256591796875,
+      "learning_rate": 1.4876871130520591e-05,
+      "loss": 0.903,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3579918032786885,
+      "grad_norm": 4.7507524490356445,
+      "learning_rate": 1.4871075278531222e-05,
+      "loss": 1.0951,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3581967213114754,
+      "grad_norm": 4.978414058685303,
+      "learning_rate": 1.4865277280433912e-05,
+      "loss": 0.9198,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3584016393442623,
+      "grad_norm": 7.244028568267822,
+      "learning_rate": 1.4859477138783154e-05,
+      "loss": 1.2466,
+      "step": 1749
+    },
+    {
+      "epoch": 0.35860655737704916,
+      "grad_norm": 2.6156187057495117,
+      "learning_rate": 1.4853674856134384e-05,
+      "loss": 0.9189,
+      "step": 1750
+    },
+    {
+      "epoch": 0.35881147540983604,
+      "grad_norm": 6.485260486602783,
+      "learning_rate": 1.4847870435043982e-05,
+      "loss": 1.0881,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3590163934426229,
+      "grad_norm": 4.030226707458496,
+      "learning_rate": 1.4842063878069276e-05,
+      "loss": 1.0665,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3592213114754098,
+      "grad_norm": 5.907912731170654,
+      "learning_rate": 1.483625518776853e-05,
+      "loss": 0.8598,
+      "step": 1753
+    },
+    {
+      "epoch": 0.35942622950819675,
+      "grad_norm": 3.1827847957611084,
+      "learning_rate": 1.4830444366700942e-05,
+      "loss": 1.0731,
+      "step": 1754
+    },
+    {
+      "epoch": 0.35963114754098363,
+      "grad_norm": 4.036923885345459,
+      "learning_rate": 1.4824631417426661e-05,
+      "loss": 0.9333,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3598360655737705,
+      "grad_norm": 4.103682518005371,
+      "learning_rate": 1.4818816342506763e-05,
+      "loss": 1.1045,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3600409836065574,
+      "grad_norm": 3.251239776611328,
+      "learning_rate": 1.4812999144503268e-05,
+      "loss": 1.0188,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3602459016393443,
+      "grad_norm": 6.189571857452393,
+      "learning_rate": 1.4807179825979127e-05,
+      "loss": 1.2487,
+      "step": 1758
+    },
+    {
+      "epoch": 0.36045081967213116,
+      "grad_norm": 5.848077297210693,
+      "learning_rate": 1.4801358389498233e-05,
+      "loss": 1.2348,
+      "step": 1759
+    },
+    {
+      "epoch": 0.36065573770491804,
+      "grad_norm": 6.709429740905762,
+      "learning_rate": 1.4795534837625398e-05,
+      "loss": 1.2278,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3608606557377049,
+      "grad_norm": 4.3548455238342285,
+      "learning_rate": 1.4789709172926377e-05,
+      "loss": 0.8949,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3610655737704918,
+      "grad_norm": 3.3152990341186523,
+      "learning_rate": 1.4783881397967851e-05,
+      "loss": 0.751,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3612704918032787,
+      "grad_norm": 4.084930419921875,
+      "learning_rate": 1.4778051515317438e-05,
+      "loss": 1.0314,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3614754098360656,
+      "grad_norm": 2.4672443866729736,
+      "learning_rate": 1.4772219527543672e-05,
+      "loss": 0.9574,
+      "step": 1764
+    },
+    {
+      "epoch": 0.36168032786885246,
+      "grad_norm": 3.3610403537750244,
+      "learning_rate": 1.4766385437216026e-05,
+      "loss": 0.8777,
+      "step": 1765
+    },
+    {
+      "epoch": 0.36188524590163934,
+      "grad_norm": 4.798501968383789,
+      "learning_rate": 1.4760549246904896e-05,
+      "loss": 1.0574,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3620901639344262,
+      "grad_norm": 5.137192726135254,
+      "learning_rate": 1.4754710959181599e-05,
+      "loss": 0.9493,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3622950819672131,
+      "grad_norm": 3.187095880508423,
+      "learning_rate": 1.4748870576618382e-05,
+      "loss": 0.9466,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 4.6450934410095215,
+      "learning_rate": 1.4743028101788408e-05,
+      "loss": 1.0525,
+      "step": 1769
+    },
+    {
+      "epoch": 0.36270491803278687,
+      "grad_norm": 3.063927173614502,
+      "learning_rate": 1.4737183537265772e-05,
+      "loss": 1.0069,
+      "step": 1770
+    },
+    {
+      "epoch": 0.36290983606557375,
+      "grad_norm": 2.885754346847534,
+      "learning_rate": 1.473133688562548e-05,
+      "loss": 1.0514,
+      "step": 1771
+    },
+    {
+      "epoch": 0.36311475409836064,
+      "grad_norm": 3.067291021347046,
+      "learning_rate": 1.4725488149443462e-05,
+      "loss": 0.8044,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3633196721311475,
+      "grad_norm": 5.214171886444092,
+      "learning_rate": 1.4719637331296566e-05,
+      "loss": 1.0023,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3635245901639344,
+      "grad_norm": 2.8295176029205322,
+      "learning_rate": 1.4713784433762554e-05,
+      "loss": 0.692,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3637295081967213,
+      "grad_norm": 3.123028039932251,
+      "learning_rate": 1.4707929459420114e-05,
+      "loss": 1.2049,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3639344262295082,
+      "grad_norm": 5.3125386238098145,
+      "learning_rate": 1.4702072410848838e-05,
+      "loss": 0.9348,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3641393442622951,
+      "grad_norm": 5.872818470001221,
+      "learning_rate": 1.4696213290629231e-05,
+      "loss": 1.1547,
+      "step": 1777
+    },
+    {
+      "epoch": 0.364344262295082,
+      "grad_norm": 5.118954181671143,
+      "learning_rate": 1.4690352101342726e-05,
+      "loss": 1.1443,
+      "step": 1778
+    },
+    {
+      "epoch": 0.36454918032786887,
+      "grad_norm": 8.92236328125,
+      "learning_rate": 1.468448884557165e-05,
+      "loss": 1.031,
+      "step": 1779
+    },
+    {
+      "epoch": 0.36475409836065575,
+      "grad_norm": 5.891019344329834,
+      "learning_rate": 1.467862352589925e-05,
+      "loss": 0.9754,
+      "step": 1780
+    },
+    {
+      "epoch": 0.36495901639344264,
+      "grad_norm": 4.81038761138916,
+      "learning_rate": 1.4672756144909681e-05,
+      "loss": 0.8543,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3651639344262295,
+      "grad_norm": 5.8633503913879395,
+      "learning_rate": 1.4666886705188002e-05,
+      "loss": 1.1611,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3653688524590164,
+      "grad_norm": 5.731857776641846,
+      "learning_rate": 1.4661015209320186e-05,
+      "loss": 1.1669,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3655737704918033,
+      "grad_norm": 4.293800354003906,
+      "learning_rate": 1.4655141659893107e-05,
+      "loss": 1.0623,
+      "step": 1784
+    },
+    {
+      "epoch": 0.36577868852459017,
+      "grad_norm": 4.558681488037109,
+      "learning_rate": 1.4649266059494543e-05,
+      "loss": 1.0836,
+      "step": 1785
+    },
+    {
+      "epoch": 0.36598360655737705,
+      "grad_norm": 7.346732139587402,
+      "learning_rate": 1.464338841071318e-05,
+      "loss": 1.1563,
+      "step": 1786
+    },
+    {
+      "epoch": 0.36618852459016393,
+      "grad_norm": 6.903554916381836,
+      "learning_rate": 1.4637508716138605e-05,
+      "loss": 1.2301,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3663934426229508,
+      "grad_norm": 6.454011917114258,
+      "learning_rate": 1.4631626978361304e-05,
+      "loss": 1.2074,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3665983606557377,
+      "grad_norm": 5.600589752197266,
+      "learning_rate": 1.462574319997266e-05,
+      "loss": 1.1826,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3668032786885246,
+      "grad_norm": 2.308833360671997,
+      "learning_rate": 1.4619857383564968e-05,
+      "loss": 0.9123,
+      "step": 1790
+    },
+    {
+      "epoch": 0.36700819672131146,
+      "grad_norm": 3.6873695850372314,
+      "learning_rate": 1.4613969531731409e-05,
+      "loss": 1.0239,
+      "step": 1791
+    },
+    {
+      "epoch": 0.36721311475409835,
+      "grad_norm": 4.653580665588379,
+      "learning_rate": 1.4608079647066062e-05,
+      "loss": 1.372,
+      "step": 1792
+    },
+    {
+      "epoch": 0.36741803278688523,
+      "grad_norm": 4.347935676574707,
+      "learning_rate": 1.4602187732163909e-05,
+      "loss": 1.1604,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3676229508196721,
+      "grad_norm": 3.8710479736328125,
+      "learning_rate": 1.4596293789620814e-05,
+      "loss": 1.0831,
+      "step": 1794
+    },
+    {
+      "epoch": 0.367827868852459,
+      "grad_norm": 5.620660781860352,
+      "learning_rate": 1.4590397822033553e-05,
+      "loss": 1.1839,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3680327868852459,
+      "grad_norm": 4.623664855957031,
+      "learning_rate": 1.458449983199977e-05,
+      "loss": 1.1038,
+      "step": 1796
+    },
+    {
+      "epoch": 0.36823770491803276,
+      "grad_norm": 2.3031013011932373,
+      "learning_rate": 1.4578599822118022e-05,
+      "loss": 0.7616,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3684426229508197,
+      "grad_norm": 5.434330463409424,
+      "learning_rate": 1.4572697794987745e-05,
+      "loss": 1.0335,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3686475409836066,
+      "grad_norm": 3.0141501426696777,
+      "learning_rate": 1.4566793753209263e-05,
+      "loss": 1.1562,
+      "step": 1799
+    },
+    {
+      "epoch": 0.36885245901639346,
+      "grad_norm": 4.402254581451416,
+      "learning_rate": 1.4560887699383794e-05,
+      "loss": 0.8787,
+      "step": 1800
+    },
+    {
+      "epoch": 0.36905737704918035,
+      "grad_norm": 3.984282970428467,
+      "learning_rate": 1.455497963611344e-05,
+      "loss": 1.3157,
+      "step": 1801
+    },
+    {
+      "epoch": 0.36926229508196723,
+      "grad_norm": 4.581586837768555,
+      "learning_rate": 1.4549069566001181e-05,
+      "loss": 1.0576,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3694672131147541,
+      "grad_norm": 4.568691253662109,
+      "learning_rate": 1.454315749165089e-05,
+      "loss": 1.0788,
+      "step": 1803
+    },
+    {
+      "epoch": 0.369672131147541,
+      "grad_norm": 2.3496925830841064,
+      "learning_rate": 1.4537243415667324e-05,
+      "loss": 0.8794,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3698770491803279,
+      "grad_norm": 3.8501901626586914,
+      "learning_rate": 1.4531327340656114e-05,
+      "loss": 0.6662,
+      "step": 1805
+    },
+    {
+      "epoch": 0.37008196721311476,
+      "grad_norm": 7.713149070739746,
+      "learning_rate": 1.4525409269223779e-05,
+      "loss": 1.1998,
+      "step": 1806
+    },
+    {
+      "epoch": 0.37028688524590164,
+      "grad_norm": 3.537468433380127,
+      "learning_rate": 1.4519489203977708e-05,
+      "loss": 1.1821,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3704918032786885,
+      "grad_norm": 24.20016098022461,
+      "learning_rate": 1.4513567147526186e-05,
+      "loss": 1.1174,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3706967213114754,
+      "grad_norm": 7.602541923522949,
+      "learning_rate": 1.4507643102478355e-05,
+      "loss": 1.0234,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3709016393442623,
+      "grad_norm": 4.681624889373779,
+      "learning_rate": 1.4501717071444248e-05,
+      "loss": 0.8756,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3711065573770492,
+      "grad_norm": 3.1344172954559326,
+      "learning_rate": 1.4495789057034766e-05,
+      "loss": 1.0631,
+      "step": 1811
+    },
+    {
+      "epoch": 0.37131147540983606,
+      "grad_norm": 5.120546340942383,
+      "learning_rate": 1.448985906186168e-05,
+      "loss": 1.3174,
+      "step": 1812
+    },
+    {
+      "epoch": 0.37151639344262294,
+      "grad_norm": 4.491763591766357,
+      "learning_rate": 1.4483927088537646e-05,
+      "loss": 0.8278,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3717213114754098,
+      "grad_norm": 7.594021320343018,
+      "learning_rate": 1.447799313967618e-05,
+      "loss": 1.0128,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3719262295081967,
+      "grad_norm": 4.477758407592773,
+      "learning_rate": 1.4472057217891678e-05,
+      "loss": 1.146,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3721311475409836,
+      "grad_norm": 4.663478374481201,
+      "learning_rate": 1.4466119325799394e-05,
+      "loss": 0.8367,
+      "step": 1816
+    },
+    {
+      "epoch": 0.37233606557377047,
+      "grad_norm": 8.34043025970459,
+      "learning_rate": 1.446017946601546e-05,
+      "loss": 1.1599,
+      "step": 1817
+    },
+    {
+      "epoch": 0.37254098360655735,
+      "grad_norm": 4.3980207443237305,
+      "learning_rate": 1.4454237641156872e-05,
+      "loss": 1.0654,
+      "step": 1818
+    },
+    {
+      "epoch": 0.37274590163934423,
+      "grad_norm": 2.755558490753174,
+      "learning_rate": 1.4448293853841486e-05,
+      "loss": 1.1003,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3729508196721312,
+      "grad_norm": 4.234988689422607,
+      "learning_rate": 1.4442348106688033e-05,
+      "loss": 1.0801,
+      "step": 1820
+    },
+    {
+      "epoch": 0.37315573770491806,
+      "grad_norm": 3.3899893760681152,
+      "learning_rate": 1.4436400402316099e-05,
+      "loss": 1.1913,
+      "step": 1821
+    },
+    {
+      "epoch": 0.37336065573770494,
+      "grad_norm": 4.94830322265625,
+      "learning_rate": 1.4430450743346137e-05,
+      "loss": 1.0043,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3735655737704918,
+      "grad_norm": 6.9429826736450195,
+      "learning_rate": 1.4424499132399459e-05,
+      "loss": 1.1283,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3737704918032787,
+      "grad_norm": 9.119359970092773,
+      "learning_rate": 1.441854557209823e-05,
+      "loss": 1.0906,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3739754098360656,
+      "grad_norm": 3.9986066818237305,
+      "learning_rate": 1.4412590065065497e-05,
+      "loss": 1.2403,
+      "step": 1825
+    },
+    {
+      "epoch": 0.37418032786885247,
+      "grad_norm": 10.118354797363281,
+      "learning_rate": 1.4406632613925136e-05,
+      "loss": 1.121,
+      "step": 1826
+    },
+    {
+      "epoch": 0.37438524590163935,
+      "grad_norm": 8.326202392578125,
+      "learning_rate": 1.4400673221301898e-05,
+      "loss": 1.0717,
+      "step": 1827
+    },
+    {
+      "epoch": 0.37459016393442623,
+      "grad_norm": 5.7251667976379395,
+      "learning_rate": 1.439471188982138e-05,
+      "loss": 0.9343,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3747950819672131,
+      "grad_norm": 2.7658488750457764,
+      "learning_rate": 1.4388748622110043e-05,
+      "loss": 0.8168,
+      "step": 1829
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 5.866053581237793,
+      "learning_rate": 1.4382783420795193e-05,
+      "loss": 0.9311,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3752049180327869,
+      "grad_norm": 6.480401992797852,
+      "learning_rate": 1.4376816288504985e-05,
+      "loss": 1.1585,
+      "step": 1831
+    },
+    {
+      "epoch": 0.37540983606557377,
+      "grad_norm": 5.33212423324585,
+      "learning_rate": 1.4370847227868436e-05,
+      "loss": 0.9372,
+      "step": 1832
+    },
+    {
+      "epoch": 0.37561475409836065,
+      "grad_norm": 8.351614952087402,
+      "learning_rate": 1.4364876241515403e-05,
+      "loss": 1.267,
+      "step": 1833
+    },
+    {
+      "epoch": 0.37581967213114753,
+      "grad_norm": 8.580108642578125,
+      "learning_rate": 1.4358903332076601e-05,
+      "loss": 1.2281,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3760245901639344,
+      "grad_norm": 4.579860687255859,
+      "learning_rate": 1.4352928502183579e-05,
+      "loss": 1.015,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3762295081967213,
+      "grad_norm": 2.719040632247925,
+      "learning_rate": 1.4346951754468746e-05,
+      "loss": 0.936,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3764344262295082,
+      "grad_norm": 13.384531021118164,
+      "learning_rate": 1.434097309156534e-05,
+      "loss": 1.1772,
+      "step": 1837
+    },
+    {
+      "epoch": 0.37663934426229506,
+      "grad_norm": 4.610185146331787,
+      "learning_rate": 1.433499251610746e-05,
+      "loss": 0.8231,
+      "step": 1838
+    },
+    {
+      "epoch": 0.37684426229508194,
+      "grad_norm": 4.374967098236084,
+      "learning_rate": 1.4329010030730035e-05,
+      "loss": 0.9518,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3770491803278688,
+      "grad_norm": 4.039411544799805,
+      "learning_rate": 1.4323025638068843e-05,
+      "loss": 1.0004,
+      "step": 1840
+    },
+    {
+      "epoch": 0.37725409836065577,
+      "grad_norm": 5.983097553253174,
+      "learning_rate": 1.4317039340760501e-05,
+      "loss": 1.0062,
+      "step": 1841
+    },
+    {
+      "epoch": 0.37745901639344265,
+      "grad_norm": 6.8633527755737305,
+      "learning_rate": 1.4311051141442461e-05,
+      "loss": 0.893,
+      "step": 1842
+    },
+    {
+      "epoch": 0.37766393442622953,
+      "grad_norm": 4.870161056518555,
+      "learning_rate": 1.4305061042753017e-05,
+      "loss": 1.2116,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3778688524590164,
+      "grad_norm": 3.3281803131103516,
+      "learning_rate": 1.4299069047331295e-05,
+      "loss": 0.8836,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3780737704918033,
+      "grad_norm": 8.192225456237793,
+      "learning_rate": 1.4293075157817265e-05,
+      "loss": 1.2803,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3782786885245902,
+      "grad_norm": 3.750775098800659,
+      "learning_rate": 1.4287079376851724e-05,
+      "loss": 0.9843,
+      "step": 1846
+    },
+    {
+      "epoch": 0.37848360655737706,
+      "grad_norm": 3.7559735774993896,
+      "learning_rate": 1.4281081707076305e-05,
+      "loss": 1.1529,
+      "step": 1847
+    },
+    {
+      "epoch": 0.37868852459016394,
+      "grad_norm": 3.564563512802124,
+      "learning_rate": 1.4275082151133476e-05,
+      "loss": 1.0789,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3788934426229508,
+      "grad_norm": 6.837707042694092,
+      "learning_rate": 1.4269080711666525e-05,
+      "loss": 1.0042,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3790983606557377,
+      "grad_norm": 2.9677531719207764,
+      "learning_rate": 1.426307739131959e-05,
+      "loss": 0.992,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3793032786885246,
+      "grad_norm": 3.22823166847229,
+      "learning_rate": 1.4257072192737618e-05,
+      "loss": 1.1237,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3795081967213115,
+      "grad_norm": 4.8577799797058105,
+      "learning_rate": 1.4251065118566392e-05,
+      "loss": 1.0071,
+      "step": 1852
+    },
+    {
+      "epoch": 0.37971311475409836,
+      "grad_norm": 3.4197304248809814,
+      "learning_rate": 1.4245056171452523e-05,
+      "loss": 1.0,
+      "step": 1853
+    },
+    {
+      "epoch": 0.37991803278688524,
+      "grad_norm": 3.014448404312134,
+      "learning_rate": 1.423904535404344e-05,
+      "loss": 1.2337,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3801229508196721,
+      "grad_norm": 5.074424743652344,
+      "learning_rate": 1.4233032668987406e-05,
+      "loss": 1.0092,
+      "step": 1855
+    },
+    {
+      "epoch": 0.380327868852459,
+      "grad_norm": 17.643964767456055,
+      "learning_rate": 1.4227018118933497e-05,
+      "loss": 1.0567,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3805327868852459,
+      "grad_norm": 3.36885142326355,
+      "learning_rate": 1.422100170653162e-05,
+      "loss": 1.0402,
+      "step": 1857
+    },
+    {
+      "epoch": 0.38073770491803277,
+      "grad_norm": 30.74517059326172,
+      "learning_rate": 1.4214983434432496e-05,
+      "loss": 1.0064,
+      "step": 1858
+    },
+    {
+      "epoch": 0.38094262295081965,
+      "grad_norm": 2.5787880420684814,
+      "learning_rate": 1.4208963305287667e-05,
+      "loss": 0.75,
+      "step": 1859
+    },
+    {
+      "epoch": 0.38114754098360654,
+      "grad_norm": 5.604892730712891,
+      "learning_rate": 1.4202941321749493e-05,
+      "loss": 1.2579,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3813524590163934,
+      "grad_norm": 4.26726770401001,
+      "learning_rate": 1.4196917486471152e-05,
+      "loss": 0.9687,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3815573770491803,
+      "grad_norm": 5.292337417602539,
+      "learning_rate": 1.4190891802106635e-05,
+      "loss": 0.9522,
+      "step": 1862
+    },
+    {
+      "epoch": 0.38176229508196724,
+      "grad_norm": 2.54577898979187,
+      "learning_rate": 1.4184864271310752e-05,
+      "loss": 1.0324,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3819672131147541,
+      "grad_norm": 3.3163950443267822,
+      "learning_rate": 1.4178834896739124e-05,
+      "loss": 0.9714,
+      "step": 1864
+    },
+    {
+      "epoch": 0.382172131147541,
+      "grad_norm": 2.800747871398926,
+      "learning_rate": 1.4172803681048185e-05,
+      "loss": 0.7446,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3823770491803279,
+      "grad_norm": 7.46232271194458,
+      "learning_rate": 1.4166770626895179e-05,
+      "loss": 0.9757,
+      "step": 1866
+    },
+    {
+      "epoch": 0.38258196721311477,
+      "grad_norm": 8.699970245361328,
+      "learning_rate": 1.4160735736938165e-05,
+      "loss": 0.9478,
+      "step": 1867
+    },
+    {
+      "epoch": 0.38278688524590165,
+      "grad_norm": 2.7876229286193848,
+      "learning_rate": 1.4154699013836e-05,
+      "loss": 1.0281,
+      "step": 1868
+    },
+    {
+      "epoch": 0.38299180327868854,
+      "grad_norm": 4.007840633392334,
+      "learning_rate": 1.414866046024836e-05,
+      "loss": 1.026,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3831967213114754,
+      "grad_norm": 3.1090099811553955,
+      "learning_rate": 1.414262007883572e-05,
+      "loss": 0.9293,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3834016393442623,
+      "grad_norm": 7.137269973754883,
+      "learning_rate": 1.4136577872259365e-05,
+      "loss": 1.0302,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3836065573770492,
+      "grad_norm": 4.532566070556641,
+      "learning_rate": 1.4130533843181385e-05,
+      "loss": 0.9983,
+      "step": 1872
+    },
+    {
+      "epoch": 0.38381147540983607,
+      "grad_norm": 10.398268699645996,
+      "learning_rate": 1.4124487994264664e-05,
+      "loss": 1.0904,
+      "step": 1873
+    },
+    {
+      "epoch": 0.38401639344262295,
+      "grad_norm": 6.799644947052002,
+      "learning_rate": 1.4118440328172895e-05,
+      "loss": 0.9309,
+      "step": 1874
+    },
+    {
+      "epoch": 0.38422131147540983,
+      "grad_norm": 4.156732559204102,
+      "learning_rate": 1.4112390847570576e-05,
+      "loss": 1.065,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3844262295081967,
+      "grad_norm": 8.644786834716797,
+      "learning_rate": 1.4106339555122994e-05,
+      "loss": 1.0205,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3846311475409836,
+      "grad_norm": 2.8814947605133057,
+      "learning_rate": 1.4100286453496244e-05,
+      "loss": 0.9529,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3848360655737705,
+      "grad_norm": 4.568760395050049,
+      "learning_rate": 1.409423154535721e-05,
+      "loss": 0.7939,
+      "step": 1878
+    },
+    {
+      "epoch": 0.38504098360655736,
+      "grad_norm": 4.258334636688232,
+      "learning_rate": 1.4088174833373576e-05,
+      "loss": 0.8957,
+      "step": 1879
+    },
+    {
+      "epoch": 0.38524590163934425,
+      "grad_norm": 6.630258560180664,
+      "learning_rate": 1.4082116320213817e-05,
+      "loss": 1.0557,
+      "step": 1880
+    },
+    {
+      "epoch": 0.38545081967213113,
+      "grad_norm": 5.396305084228516,
+      "learning_rate": 1.407605600854721e-05,
+      "loss": 1.0702,
+      "step": 1881
+    },
+    {
+      "epoch": 0.385655737704918,
+      "grad_norm": 4.661121845245361,
+      "learning_rate": 1.4069993901043817e-05,
+      "loss": 0.9488,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3858606557377049,
+      "grad_norm": 9.813624382019043,
+      "learning_rate": 1.4063930000374495e-05,
+      "loss": 1.0806,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3860655737704918,
+      "grad_norm": 3.361328601837158,
+      "learning_rate": 1.4057864309210883e-05,
+      "loss": 1.1948,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3862704918032787,
+      "grad_norm": 3.8617069721221924,
+      "learning_rate": 1.4051796830225424e-05,
+      "loss": 1.0777,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3864754098360656,
+      "grad_norm": 3.0906295776367188,
+      "learning_rate": 1.4045727566091332e-05,
+      "loss": 0.893,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3866803278688525,
+      "grad_norm": 4.7717204093933105,
+      "learning_rate": 1.403965651948262e-05,
+      "loss": 1.0921,
+      "step": 1887
+    },
+    {
+      "epoch": 0.38688524590163936,
+      "grad_norm": 10.339519500732422,
+      "learning_rate": 1.4033583693074081e-05,
+      "loss": 1.1866,
+      "step": 1888
+    },
+    {
+      "epoch": 0.38709016393442625,
+      "grad_norm": 3.901482343673706,
+      "learning_rate": 1.4027509089541292e-05,
+      "loss": 1.0872,
+      "step": 1889
+    },
+    {
+      "epoch": 0.38729508196721313,
+      "grad_norm": 11.61197566986084,
+      "learning_rate": 1.4021432711560614e-05,
+      "loss": 1.3179,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 7.29616641998291,
+      "learning_rate": 1.4015354561809187e-05,
+      "loss": 1.1247,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3877049180327869,
+      "grad_norm": 3.2588589191436768,
+      "learning_rate": 1.400927464296494e-05,
+      "loss": 1.1049,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3879098360655738,
+      "grad_norm": 2.004760265350342,
+      "learning_rate": 1.4003192957706575e-05,
+      "loss": 0.9062,
+      "step": 1893
+    },
+    {
+      "epoch": 0.38811475409836066,
+      "grad_norm": 3.70276141166687,
+      "learning_rate": 1.3997109508713563e-05,
+      "loss": 1.1038,
+      "step": 1894
+    },
+    {
+      "epoch": 0.38831967213114754,
+      "grad_norm": 4.165141582489014,
+      "learning_rate": 1.3991024298666176e-05,
+      "loss": 0.8939,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3885245901639344,
+      "grad_norm": 5.068320274353027,
+      "learning_rate": 1.3984937330245436e-05,
+      "loss": 1.2123,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3887295081967213,
+      "grad_norm": 5.085746765136719,
+      "learning_rate": 1.3978848606133159e-05,
+      "loss": 1.0381,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3889344262295082,
+      "grad_norm": 2.434196710586548,
+      "learning_rate": 1.3972758129011925e-05,
+      "loss": 0.8631,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3891393442622951,
+      "grad_norm": 3.3028626441955566,
+      "learning_rate": 1.3966665901565087e-05,
+      "loss": 0.6566,
+      "step": 1899
+    },
+    {
+      "epoch": 0.38934426229508196,
+      "grad_norm": 3.714519500732422,
+      "learning_rate": 1.3960571926476774e-05,
+      "loss": 0.6583,
+      "step": 1900
+    },
+    {
+      "epoch": 0.38954918032786884,
+      "grad_norm": 4.620834827423096,
+      "learning_rate": 1.3954476206431874e-05,
+      "loss": 1.3068,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3897540983606557,
+      "grad_norm": 9.402534484863281,
+      "learning_rate": 1.3948378744116062e-05,
+      "loss": 0.9534,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3899590163934426,
+      "grad_norm": 10.08598804473877,
+      "learning_rate": 1.3942279542215764e-05,
+      "loss": 0.8733,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3901639344262295,
+      "grad_norm": 4.172461986541748,
+      "learning_rate": 1.3936178603418176e-05,
+      "loss": 1.3641,
+      "step": 1904
+    },
+    {
+      "epoch": 0.39036885245901637,
+      "grad_norm": 4.737673759460449,
+      "learning_rate": 1.3930075930411265e-05,
+      "loss": 1.0226,
+      "step": 1905
+    },
+    {
+      "epoch": 0.39057377049180325,
+      "grad_norm": 4.492729663848877,
+      "learning_rate": 1.392397152588376e-05,
+      "loss": 1.0959,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3907786885245902,
+      "grad_norm": 5.1359028816223145,
+      "learning_rate": 1.391786539252515e-05,
+      "loss": 1.028,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3909836065573771,
+      "grad_norm": 4.279147624969482,
+      "learning_rate": 1.3911757533025684e-05,
+      "loss": 1.1214,
+      "step": 1908
+    },
+    {
+      "epoch": 0.39118852459016396,
+      "grad_norm": 2.6443324089050293,
+      "learning_rate": 1.3905647950076382e-05,
+      "loss": 1.0703,
+      "step": 1909
+    },
+    {
+      "epoch": 0.39139344262295084,
+      "grad_norm": 9.472622871398926,
+      "learning_rate": 1.3899536646369009e-05,
+      "loss": 1.041,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3915983606557377,
+      "grad_norm": 5.148986339569092,
+      "learning_rate": 1.3893423624596098e-05,
+      "loss": 1.3338,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3918032786885246,
+      "grad_norm": 5.8152360916137695,
+      "learning_rate": 1.3887308887450937e-05,
+      "loss": 1.2292,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3920081967213115,
+      "grad_norm": 5.435270309448242,
+      "learning_rate": 1.3881192437627572e-05,
+      "loss": 0.898,
+      "step": 1913
+    },
+    {
+      "epoch": 0.39221311475409837,
+      "grad_norm": 4.239561557769775,
+      "learning_rate": 1.387507427782079e-05,
+      "loss": 1.1754,
+      "step": 1914
+    },
+    {
+      "epoch": 0.39241803278688525,
+      "grad_norm": 5.683990955352783,
+      "learning_rate": 1.3868954410726154e-05,
+      "loss": 0.8392,
+      "step": 1915
+    },
+    {
+      "epoch": 0.39262295081967213,
+      "grad_norm": 4.253211498260498,
+      "learning_rate": 1.386283283903996e-05,
+      "loss": 1.1851,
+      "step": 1916
+    },
+    {
+      "epoch": 0.392827868852459,
+      "grad_norm": 2.3173460960388184,
+      "learning_rate": 1.3856709565459263e-05,
+      "loss": 0.8024,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3930327868852459,
+      "grad_norm": 2.903177499771118,
+      "learning_rate": 1.3850584592681871e-05,
+      "loss": 0.9729,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3932377049180328,
+      "grad_norm": 5.254565238952637,
+      "learning_rate": 1.3844457923406327e-05,
+      "loss": 0.8462,
+      "step": 1919
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 3.1039884090423584,
+      "learning_rate": 1.3838329560331943e-05,
+      "loss": 0.9083,
+      "step": 1920
+    },
+    {
+      "epoch": 0.39364754098360655,
+      "grad_norm": 4.559187412261963,
+      "learning_rate": 1.383219950615876e-05,
+      "loss": 1.054,
+      "step": 1921
+    },
+    {
+      "epoch": 0.39385245901639343,
+      "grad_norm": 4.4355244636535645,
+      "learning_rate": 1.3826067763587565e-05,
+      "loss": 0.8853,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3940573770491803,
+      "grad_norm": 3.661328077316284,
+      "learning_rate": 1.3819934335319897e-05,
+      "loss": 0.9642,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3942622950819672,
+      "grad_norm": 4.7848005294799805,
+      "learning_rate": 1.3813799224058038e-05,
+      "loss": 1.0693,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3944672131147541,
+      "grad_norm": 4.3750481605529785,
+      "learning_rate": 1.3807662432505002e-05,
+      "loss": 1.1506,
+      "step": 1925
+    },
+    {
+      "epoch": 0.39467213114754096,
+      "grad_norm": 3.589367389678955,
+      "learning_rate": 1.380152396336455e-05,
+      "loss": 1.162,
+      "step": 1926
+    },
+    {
+      "epoch": 0.39487704918032784,
+      "grad_norm": 3.9026482105255127,
+      "learning_rate": 1.3795383819341184e-05,
+      "loss": 0.8915,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3950819672131147,
+      "grad_norm": 3.180377960205078,
+      "learning_rate": 1.3789242003140136e-05,
+      "loss": 0.7669,
+      "step": 1928
+    },
+    {
+      "epoch": 0.39528688524590166,
+      "grad_norm": 8.277277946472168,
+      "learning_rate": 1.3783098517467379e-05,
+      "loss": 0.9652,
+      "step": 1929
+    },
+    {
+      "epoch": 0.39549180327868855,
+      "grad_norm": 6.795473575592041,
+      "learning_rate": 1.3776953365029628e-05,
+      "loss": 1.2837,
+      "step": 1930
+    },
+    {
+      "epoch": 0.39569672131147543,
+      "grad_norm": 3.31337833404541,
+      "learning_rate": 1.3770806548534319e-05,
+      "loss": 0.8475,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3959016393442623,
+      "grad_norm": 2.5058460235595703,
+      "learning_rate": 1.3764658070689633e-05,
+      "loss": 0.9199,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3961065573770492,
+      "grad_norm": 1.9653961658477783,
+      "learning_rate": 1.3758507934204479e-05,
+      "loss": 0.8165,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3963114754098361,
+      "grad_norm": 6.84417200088501,
+      "learning_rate": 1.3752356141788496e-05,
+      "loss": 1.066,
+      "step": 1934
+    },
+    {
+      "epoch": 0.39651639344262296,
+      "grad_norm": 6.708113670349121,
+      "learning_rate": 1.3746202696152052e-05,
+      "loss": 1.4616,
+      "step": 1935
+    },
+    {
+      "epoch": 0.39672131147540984,
+      "grad_norm": 10.671759605407715,
+      "learning_rate": 1.3740047600006241e-05,
+      "loss": 0.9332,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3969262295081967,
+      "grad_norm": 8.480690956115723,
+      "learning_rate": 1.3733890856062893e-05,
+      "loss": 1.3333,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3971311475409836,
+      "grad_norm": 7.655148506164551,
+      "learning_rate": 1.3727732467034556e-05,
+      "loss": 0.8943,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3973360655737705,
+      "grad_norm": 8.431061744689941,
+      "learning_rate": 1.3721572435634502e-05,
+      "loss": 1.1987,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3975409836065574,
+      "grad_norm": 9.311687469482422,
+      "learning_rate": 1.3715410764576736e-05,
+      "loss": 1.2559,
+      "step": 1940
+    },
+    {
+      "epoch": 0.39774590163934426,
+      "grad_norm": 6.2136006355285645,
+      "learning_rate": 1.3709247456575974e-05,
+      "loss": 0.7985,
+      "step": 1941
+    },
+    {
+      "epoch": 0.39795081967213114,
+      "grad_norm": 5.250126361846924,
+      "learning_rate": 1.370308251434766e-05,
+      "loss": 0.8266,
+      "step": 1942
+    },
+    {
+      "epoch": 0.398155737704918,
+      "grad_norm": 3.5207834243774414,
+      "learning_rate": 1.3696915940607957e-05,
+      "loss": 1.272,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3983606557377049,
+      "grad_norm": 4.885398864746094,
+      "learning_rate": 1.3690747738073749e-05,
+      "loss": 1.3121,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3985655737704918,
+      "grad_norm": 3.30849552154541,
+      "learning_rate": 1.3684577909462626e-05,
+      "loss": 0.9227,
+      "step": 1945
+    },
+    {
+      "epoch": 0.39877049180327867,
+      "grad_norm": 2.9927494525909424,
+      "learning_rate": 1.3678406457492912e-05,
+      "loss": 1.1097,
+      "step": 1946
+    },
+    {
+      "epoch": 0.39897540983606555,
+      "grad_norm": 4.356511116027832,
+      "learning_rate": 1.3672233384883636e-05,
+      "loss": 0.8371,
+      "step": 1947
+    },
+    {
+      "epoch": 0.39918032786885244,
+      "grad_norm": 1.9214849472045898,
+      "learning_rate": 1.3666058694354538e-05,
+      "loss": 0.8495,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3993852459016393,
+      "grad_norm": 3.7468152046203613,
+      "learning_rate": 1.3659882388626077e-05,
+      "loss": 1.1746,
+      "step": 1949
+    },
+    {
+      "epoch": 0.39959016393442626,
+      "grad_norm": 5.114481449127197,
+      "learning_rate": 1.365370447041943e-05,
+      "loss": 1.2527,
+      "step": 1950
+    },
+    {
+      "epoch": 0.39979508196721314,
+      "grad_norm": 3.599224090576172,
+      "learning_rate": 1.3647524942456465e-05,
+      "loss": 0.9433,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.870095729827881,
+      "learning_rate": 1.3641343807459775e-05,
+      "loss": 1.1659,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4002049180327869,
+      "grad_norm": 4.1972479820251465,
+      "learning_rate": 1.3635161068152657e-05,
+      "loss": 0.9153,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4004098360655738,
+      "grad_norm": 6.317677021026611,
+      "learning_rate": 1.3628976727259113e-05,
+      "loss": 1.0138,
+      "step": 1954
+    },
+    {
+      "epoch": 0.40061475409836067,
+      "grad_norm": 4.843593120574951,
+      "learning_rate": 1.3622790787503855e-05,
+      "loss": 0.9696,
+      "step": 1955
+    },
+    {
+      "epoch": 0.40081967213114755,
+      "grad_norm": 3.497819423675537,
+      "learning_rate": 1.3616603251612293e-05,
+      "loss": 0.9924,
+      "step": 1956
+    },
+    {
+      "epoch": 0.40102459016393444,
+      "grad_norm": 4.357882976531982,
+      "learning_rate": 1.3610414122310549e-05,
+      "loss": 1.0093,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4012295081967213,
+      "grad_norm": 5.097222805023193,
+      "learning_rate": 1.3604223402325436e-05,
+      "loss": 1.0004,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4014344262295082,
+      "grad_norm": 9.275769233703613,
+      "learning_rate": 1.3598031094384475e-05,
+      "loss": 0.9037,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4016393442622951,
+      "grad_norm": 4.431663513183594,
+      "learning_rate": 1.3591837201215889e-05,
+      "loss": 1.1133,
+      "step": 1960
+    },
+    {
+      "epoch": 0.40184426229508197,
+      "grad_norm": 4.781447887420654,
+      "learning_rate": 1.358564172554859e-05,
+      "loss": 1.0012,
+      "step": 1961
+    },
+    {
+      "epoch": 0.40204918032786885,
+      "grad_norm": 4.395463466644287,
+      "learning_rate": 1.35794446701122e-05,
+      "loss": 1.005,
+      "step": 1962
+    },
+    {
+      "epoch": 0.40225409836065573,
+      "grad_norm": 4.230912685394287,
+      "learning_rate": 1.3573246037637023e-05,
+      "loss": 0.9627,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4024590163934426,
+      "grad_norm": 3.7451817989349365,
+      "learning_rate": 1.3567045830854066e-05,
+      "loss": 0.8276,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4026639344262295,
+      "grad_norm": 4.771474838256836,
+      "learning_rate": 1.3560844052495032e-05,
+      "loss": 0.9444,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4028688524590164,
+      "grad_norm": 3.20559024810791,
+      "learning_rate": 1.3554640705292307e-05,
+      "loss": 1.0078,
+      "step": 1966
+    },
+    {
+      "epoch": 0.40307377049180326,
+      "grad_norm": 4.007946968078613,
+      "learning_rate": 1.3548435791978982e-05,
+      "loss": 1.0077,
+      "step": 1967
+    },
+    {
+      "epoch": 0.40327868852459015,
+      "grad_norm": 5.040585994720459,
+      "learning_rate": 1.3542229315288822e-05,
+      "loss": 1.0493,
+      "step": 1968
+    },
+    {
+      "epoch": 0.40348360655737703,
+      "grad_norm": 5.03970193862915,
+      "learning_rate": 1.3536021277956292e-05,
+      "loss": 0.8373,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4036885245901639,
+      "grad_norm": 4.964225769042969,
+      "learning_rate": 1.3529811682716542e-05,
+      "loss": 1.2331,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4038934426229508,
+      "grad_norm": 5.9297637939453125,
+      "learning_rate": 1.3523600532305408e-05,
+      "loss": 0.8628,
+      "step": 1971
+    },
+    {
+      "epoch": 0.40409836065573773,
+      "grad_norm": 5.475839138031006,
+      "learning_rate": 1.3517387829459412e-05,
+      "loss": 0.8539,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4043032786885246,
+      "grad_norm": 3.3281736373901367,
+      "learning_rate": 1.351117357691575e-05,
+      "loss": 1.0536,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4045081967213115,
+      "grad_norm": 6.420851230621338,
+      "learning_rate": 1.3504957777412322e-05,
+      "loss": 1.2825,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4047131147540984,
+      "grad_norm": 3.006334066390991,
+      "learning_rate": 1.3498740433687689e-05,
+      "loss": 1.0144,
+      "step": 1975
+    },
+    {
+      "epoch": 0.40491803278688526,
+      "grad_norm": 2.086092948913574,
+      "learning_rate": 1.3492521548481104e-05,
+      "loss": 0.7581,
+      "step": 1976
+    },
+    {
+      "epoch": 0.40512295081967215,
+      "grad_norm": 5.731237411499023,
+      "learning_rate": 1.3486301124532494e-05,
+      "loss": 0.9279,
+      "step": 1977
+    },
+    {
+      "epoch": 0.40532786885245903,
+      "grad_norm": 4.29608678817749,
+      "learning_rate": 1.3480079164582467e-05,
+      "loss": 1.1236,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4055327868852459,
+      "grad_norm": 5.455349922180176,
+      "learning_rate": 1.3473855671372309e-05,
+      "loss": 1.1731,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4057377049180328,
+      "grad_norm": 3.5686802864074707,
+      "learning_rate": 1.3467630647643972e-05,
+      "loss": 0.9467,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4059426229508197,
+      "grad_norm": 3.4309475421905518,
+      "learning_rate": 1.3461404096140093e-05,
+      "loss": 0.7895,
+      "step": 1981
+    },
+    {
+      "epoch": 0.40614754098360656,
+      "grad_norm": 7.743010997772217,
+      "learning_rate": 1.345517601960398e-05,
+      "loss": 0.9685,
+      "step": 1982
+    },
+    {
+      "epoch": 0.40635245901639344,
+      "grad_norm": 5.765895843505859,
+      "learning_rate": 1.3448946420779607e-05,
+      "loss": 1.2825,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4065573770491803,
+      "grad_norm": 5.007030010223389,
+      "learning_rate": 1.3442715302411627e-05,
+      "loss": 0.9313,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4067622950819672,
+      "grad_norm": 3.9002809524536133,
+      "learning_rate": 1.3436482667245359e-05,
+      "loss": 1.1509,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4069672131147541,
+      "grad_norm": 3.839287281036377,
+      "learning_rate": 1.3430248518026786e-05,
+      "loss": 1.1562,
+      "step": 1986
+    },
+    {
+      "epoch": 0.40717213114754097,
+      "grad_norm": 5.150951385498047,
+      "learning_rate": 1.3424012857502559e-05,
+      "loss": 0.9514,
+      "step": 1987
+    },
+    {
+      "epoch": 0.40737704918032785,
+      "grad_norm": 7.934512615203857,
+      "learning_rate": 1.3417775688420008e-05,
+      "loss": 1.0779,
+      "step": 1988
+    },
+    {
+      "epoch": 0.40758196721311474,
+      "grad_norm": 4.939842700958252,
+      "learning_rate": 1.3411537013527106e-05,
+      "loss": 0.9952,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4077868852459016,
+      "grad_norm": 5.696869850158691,
+      "learning_rate": 1.340529683557251e-05,
+      "loss": 1.4329,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4079918032786885,
+      "grad_norm": 6.64450216293335,
+      "learning_rate": 1.339905515730552e-05,
+      "loss": 1.2588,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4081967213114754,
+      "grad_norm": 6.057649612426758,
+      "learning_rate": 1.3392811981476119e-05,
+      "loss": 1.2047,
+      "step": 1992
+    },
+    {
+      "epoch": 0.40840163934426227,
+      "grad_norm": 2.890815258026123,
+      "learning_rate": 1.3386567310834928e-05,
+      "loss": 0.9726,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4086065573770492,
+      "grad_norm": 4.374378681182861,
+      "learning_rate": 1.3380321148133242e-05,
+      "loss": 1.2448,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4088114754098361,
+      "grad_norm": 7.383669376373291,
+      "learning_rate": 1.3374073496123004e-05,
+      "loss": 1.1756,
+      "step": 1995
+    },
+    {
+      "epoch": 0.40901639344262297,
+      "grad_norm": 3.5708839893341064,
+      "learning_rate": 1.3367824357556815e-05,
+      "loss": 1.0367,
+      "step": 1996
+    },
+    {
+      "epoch": 0.40922131147540985,
+      "grad_norm": 3.8618876934051514,
+      "learning_rate": 1.336157373518794e-05,
+      "loss": 0.845,
+      "step": 1997
+    },
+    {
+      "epoch": 0.40942622950819674,
+      "grad_norm": 4.81962251663208,
+      "learning_rate": 1.335532163177028e-05,
+      "loss": 1.211,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4096311475409836,
+      "grad_norm": 5.195772647857666,
+      "learning_rate": 1.3349068050058414e-05,
+      "loss": 0.8164,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4098360655737705,
+      "grad_norm": 3.770683765411377,
+      "learning_rate": 1.3342812992807546e-05,
+      "loss": 1.1596,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4100409836065574,
+      "grad_norm": 5.612376689910889,
+      "learning_rate": 1.3336556462773542e-05,
+      "loss": 0.8298,
+      "step": 2001
+    },
+    {
+      "epoch": 0.41024590163934427,
+      "grad_norm": 3.234585762023926,
+      "learning_rate": 1.3330298462712921e-05,
+      "loss": 1.0364,
+      "step": 2002
+    },
+    {
+      "epoch": 0.41045081967213115,
+      "grad_norm": 3.163936138153076,
+      "learning_rate": 1.332403899538284e-05,
+      "loss": 0.9578,
+      "step": 2003
+    },
+    {
+      "epoch": 0.41065573770491803,
+      "grad_norm": 3.018179178237915,
+      "learning_rate": 1.3317778063541113e-05,
+      "loss": 1.2071,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4108606557377049,
+      "grad_norm": 9.120782852172852,
+      "learning_rate": 1.3311515669946187e-05,
+      "loss": 1.3224,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4110655737704918,
+      "grad_norm": 5.188650608062744,
+      "learning_rate": 1.3305251817357165e-05,
+      "loss": 1.1009,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4112704918032787,
+      "grad_norm": 2.7091925144195557,
+      "learning_rate": 1.3298986508533788e-05,
+      "loss": 0.9809,
+      "step": 2007
+    },
+    {
+      "epoch": 0.41147540983606556,
+      "grad_norm": 5.0567216873168945,
+      "learning_rate": 1.329271974623643e-05,
+      "loss": 1.0074,
+      "step": 2008
+    },
+    {
+      "epoch": 0.41168032786885245,
+      "grad_norm": 6.30210542678833,
+      "learning_rate": 1.3286451533226127e-05,
+      "loss": 1.0901,
+      "step": 2009
+    },
+    {
+      "epoch": 0.41188524590163933,
+      "grad_norm": 3.241075277328491,
+      "learning_rate": 1.328018187226453e-05,
+      "loss": 1.0661,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4120901639344262,
+      "grad_norm": 5.916114807128906,
+      "learning_rate": 1.3273910766113942e-05,
+      "loss": 0.9504,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4122950819672131,
+      "grad_norm": 5.018312931060791,
+      "learning_rate": 1.3267638217537297e-05,
+      "loss": 0.9278,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 3.8620402812957764,
+      "learning_rate": 1.326136422929817e-05,
+      "loss": 1.2448,
+      "step": 2013
+    },
+    {
+      "epoch": 0.41270491803278686,
+      "grad_norm": 4.305169105529785,
+      "learning_rate": 1.3255088804160768e-05,
+      "loss": 0.9862,
+      "step": 2014
+    },
+    {
+      "epoch": 0.41290983606557374,
+      "grad_norm": 3.6094796657562256,
+      "learning_rate": 1.3248811944889926e-05,
+      "loss": 1.239,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4131147540983607,
+      "grad_norm": 9.473259925842285,
+      "learning_rate": 1.3242533654251116e-05,
+      "loss": 0.9553,
+      "step": 2016
+    },
+    {
+      "epoch": 0.41331967213114756,
+      "grad_norm": 3.7792041301727295,
+      "learning_rate": 1.3236253935010441e-05,
+      "loss": 1.0648,
+      "step": 2017
+    },
+    {
+      "epoch": 0.41352459016393445,
+      "grad_norm": 3.0000205039978027,
+      "learning_rate": 1.3229972789934633e-05,
+      "loss": 0.9343,
+      "step": 2018
+    },
+    {
+      "epoch": 0.41372950819672133,
+      "grad_norm": 4.272130012512207,
+      "learning_rate": 1.322369022179105e-05,
+      "loss": 1.2943,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4139344262295082,
+      "grad_norm": 3.221705913543701,
+      "learning_rate": 1.3217406233347678e-05,
+      "loss": 0.7469,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4141393442622951,
+      "grad_norm": 3.8526062965393066,
+      "learning_rate": 1.3211120827373129e-05,
+      "loss": 0.9701,
+      "step": 2021
+    },
+    {
+      "epoch": 0.414344262295082,
+      "grad_norm": 2.694458246231079,
+      "learning_rate": 1.320483400663664e-05,
+      "loss": 0.846,
+      "step": 2022
+    },
+    {
+      "epoch": 0.41454918032786886,
+      "grad_norm": 2.5225422382354736,
+      "learning_rate": 1.3198545773908072e-05,
+      "loss": 1.066,
+      "step": 2023
+    },
+    {
+      "epoch": 0.41475409836065574,
+      "grad_norm": 3.0873212814331055,
+      "learning_rate": 1.3192256131957903e-05,
+      "loss": 0.9427,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4149590163934426,
+      "grad_norm": 5.970068454742432,
+      "learning_rate": 1.3185965083557242e-05,
+      "loss": 1.0781,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4151639344262295,
+      "grad_norm": 5.66378927230835,
+      "learning_rate": 1.3179672631477804e-05,
+      "loss": 0.7682,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4153688524590164,
+      "grad_norm": 7.894969940185547,
+      "learning_rate": 1.3173378778491939e-05,
+      "loss": 1.1128,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4155737704918033,
+      "grad_norm": 5.4107279777526855,
+      "learning_rate": 1.3167083527372598e-05,
+      "loss": 0.8886,
+      "step": 2028
+    },
+    {
+      "epoch": 0.41577868852459016,
+      "grad_norm": 5.569651126861572,
+      "learning_rate": 1.3160786880893355e-05,
+      "loss": 0.9928,
+      "step": 2029
+    },
+    {
+      "epoch": 0.41598360655737704,
+      "grad_norm": 3.90390682220459,
+      "learning_rate": 1.3154488841828404e-05,
+      "loss": 0.9751,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4161885245901639,
+      "grad_norm": 2.3807950019836426,
+      "learning_rate": 1.3148189412952542e-05,
+      "loss": 0.7745,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4163934426229508,
+      "grad_norm": 3.3378543853759766,
+      "learning_rate": 1.314188859704119e-05,
+      "loss": 0.9925,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4165983606557377,
+      "grad_norm": 5.574931621551514,
+      "learning_rate": 1.3135586396870366e-05,
+      "loss": 0.9267,
+      "step": 2033
+    },
+    {
+      "epoch": 0.41680327868852457,
+      "grad_norm": 6.108892440795898,
+      "learning_rate": 1.3129282815216715e-05,
+      "loss": 0.9481,
+      "step": 2034
+    },
+    {
+      "epoch": 0.41700819672131145,
+      "grad_norm": 3.6429927349090576,
+      "learning_rate": 1.3122977854857473e-05,
+      "loss": 1.2011,
+      "step": 2035
+    },
+    {
+      "epoch": 0.41721311475409834,
+      "grad_norm": 4.238472938537598,
+      "learning_rate": 1.3116671518570496e-05,
+      "loss": 1.1252,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4174180327868853,
+      "grad_norm": 2.6794450283050537,
+      "learning_rate": 1.3110363809134243e-05,
+      "loss": 0.9609,
+      "step": 2037
+    },
+    {
+      "epoch": 0.41762295081967216,
+      "grad_norm": 6.156083106994629,
+      "learning_rate": 1.3104054729327773e-05,
+      "loss": 0.9458,
+      "step": 2038
+    },
+    {
+      "epoch": 0.41782786885245904,
+      "grad_norm": 2.7086267471313477,
+      "learning_rate": 1.3097744281930758e-05,
+      "loss": 1.0754,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4180327868852459,
+      "grad_norm": 4.443950653076172,
+      "learning_rate": 1.3091432469723462e-05,
+      "loss": 0.9791,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4182377049180328,
+      "grad_norm": 5.6584930419921875,
+      "learning_rate": 1.3085119295486759e-05,
+      "loss": 1.315,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4184426229508197,
+      "grad_norm": 5.790865898132324,
+      "learning_rate": 1.3078804762002122e-05,
+      "loss": 0.9948,
+      "step": 2042
+    },
+    {
+      "epoch": 0.41864754098360657,
+      "grad_norm": 4.885839939117432,
+      "learning_rate": 1.3072488872051611e-05,
+      "loss": 1.0623,
+      "step": 2043
+    },
+    {
+      "epoch": 0.41885245901639345,
+      "grad_norm": 9.429014205932617,
+      "learning_rate": 1.3066171628417907e-05,
+      "loss": 1.2575,
+      "step": 2044
+    },
+    {
+      "epoch": 0.41905737704918034,
+      "grad_norm": 5.920489311218262,
+      "learning_rate": 1.3059853033884262e-05,
+      "loss": 0.9076,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4192622950819672,
+      "grad_norm": 2.844564914703369,
+      "learning_rate": 1.3053533091234542e-05,
+      "loss": 0.9083,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4194672131147541,
+      "grad_norm": 5.03860330581665,
+      "learning_rate": 1.3047211803253195e-05,
+      "loss": 1.0574,
+      "step": 2047
+    },
+    {
+      "epoch": 0.419672131147541,
+      "grad_norm": 7.5791215896606445,
+      "learning_rate": 1.3040889172725269e-05,
+      "loss": 1.1826,
+      "step": 2048
+    },
+    {
+      "epoch": 0.41987704918032787,
+      "grad_norm": 5.917967796325684,
+      "learning_rate": 1.3034565202436405e-05,
+      "loss": 1.2273,
+      "step": 2049
+    },
+    {
+      "epoch": 0.42008196721311475,
+      "grad_norm": 7.538298606872559,
+      "learning_rate": 1.3028239895172821e-05,
+      "loss": 0.8271,
+      "step": 2050
+    },
+    {
+      "epoch": 0.42028688524590163,
+      "grad_norm": 5.644692420959473,
+      "learning_rate": 1.3021913253721339e-05,
+      "loss": 1.1508,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4204918032786885,
+      "grad_norm": 8.369562149047852,
+      "learning_rate": 1.3015585280869361e-05,
+      "loss": 1.3095,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4206967213114754,
+      "grad_norm": 4.12898063659668,
+      "learning_rate": 1.3009255979404882e-05,
+      "loss": 0.7981,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4209016393442623,
+      "grad_norm": 5.5372538566589355,
+      "learning_rate": 1.3002925352116474e-05,
+      "loss": 1.1245,
+      "step": 2054
+    },
+    {
+      "epoch": 0.42110655737704916,
+      "grad_norm": 4.084104061126709,
+      "learning_rate": 1.2996593401793298e-05,
+      "loss": 0.8724,
+      "step": 2055
+    },
+    {
+      "epoch": 0.42131147540983604,
+      "grad_norm": 5.162708759307861,
+      "learning_rate": 1.29902601312251e-05,
+      "loss": 1.2049,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4215163934426229,
+      "grad_norm": 3.741560220718384,
+      "learning_rate": 1.29839255432022e-05,
+      "loss": 0.9752,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4217213114754098,
+      "grad_norm": 3.8502755165100098,
+      "learning_rate": 1.2977589640515507e-05,
+      "loss": 1.053,
+      "step": 2058
+    },
+    {
+      "epoch": 0.42192622950819675,
+      "grad_norm": 4.629134178161621,
+      "learning_rate": 1.2971252425956504e-05,
+      "loss": 0.9222,
+      "step": 2059
+    },
+    {
+      "epoch": 0.42213114754098363,
+      "grad_norm": 3.1456499099731445,
+      "learning_rate": 1.2964913902317254e-05,
+      "loss": 0.9551,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4223360655737705,
+      "grad_norm": 3.1817026138305664,
+      "learning_rate": 1.2958574072390396e-05,
+      "loss": 1.0724,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4225409836065574,
+      "grad_norm": 3.532932758331299,
+      "learning_rate": 1.2952232938969145e-05,
+      "loss": 1.3186,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4227459016393443,
+      "grad_norm": 2.8625805377960205,
+      "learning_rate": 1.294589050484729e-05,
+      "loss": 1.0436,
+      "step": 2063
+    },
+    {
+      "epoch": 0.42295081967213116,
+      "grad_norm": 9.031486511230469,
+      "learning_rate": 1.293954677281919e-05,
+      "loss": 1.1999,
+      "step": 2064
+    },
+    {
+      "epoch": 0.42315573770491804,
+      "grad_norm": 8.09088134765625,
+      "learning_rate": 1.2933201745679785e-05,
+      "loss": 1.0927,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4233606557377049,
+      "grad_norm": 3.760952949523926,
+      "learning_rate": 1.2926855426224574e-05,
+      "loss": 1.1726,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4235655737704918,
+      "grad_norm": 4.105858325958252,
+      "learning_rate": 1.2920507817249636e-05,
+      "loss": 1.0734,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4237704918032787,
+      "grad_norm": 3.535722255706787,
+      "learning_rate": 1.2914158921551607e-05,
+      "loss": 0.9952,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4239754098360656,
+      "grad_norm": 3.9806249141693115,
+      "learning_rate": 1.2907808741927702e-05,
+      "loss": 0.8194,
+      "step": 2069
+    },
+    {
+      "epoch": 0.42418032786885246,
+      "grad_norm": 4.939681053161621,
+      "learning_rate": 1.2901457281175693e-05,
+      "loss": 0.8737,
+      "step": 2070
+    },
+    {
+      "epoch": 0.42438524590163934,
+      "grad_norm": 3.928072214126587,
+      "learning_rate": 1.289510454209392e-05,
+      "loss": 1.0167,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4245901639344262,
+      "grad_norm": 5.263453483581543,
+      "learning_rate": 1.2888750527481285e-05,
+      "loss": 1.1674,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4247950819672131,
+      "grad_norm": 4.137607097625732,
+      "learning_rate": 1.2882395240137246e-05,
+      "loss": 0.9681,
+      "step": 2073
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 4.600144863128662,
+      "learning_rate": 1.287603868286184e-05,
+      "loss": 1.018,
+      "step": 2074
+    },
+    {
+      "epoch": 0.42520491803278687,
+      "grad_norm": 3.2041096687316895,
+      "learning_rate": 1.2869680858455642e-05,
+      "loss": 0.8042,
+      "step": 2075
+    },
+    {
+      "epoch": 0.42540983606557375,
+      "grad_norm": 3.2633378505706787,
+      "learning_rate": 1.2863321769719798e-05,
+      "loss": 1.0045,
+      "step": 2076
+    },
+    {
+      "epoch": 0.42561475409836064,
+      "grad_norm": 6.215106964111328,
+      "learning_rate": 1.2856961419456012e-05,
+      "loss": 1.0412,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4258196721311475,
+      "grad_norm": 2.7890939712524414,
+      "learning_rate": 1.2850599810466529e-05,
+      "loss": 0.6777,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4260245901639344,
+      "grad_norm": 4.888681888580322,
+      "learning_rate": 1.284423694555417e-05,
+      "loss": 1.1086,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4262295081967213,
+      "grad_norm": 9.252080917358398,
+      "learning_rate": 1.283787282752229e-05,
+      "loss": 0.9774,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4264344262295082,
+      "grad_norm": 4.89749813079834,
+      "learning_rate": 1.2831507459174814e-05,
+      "loss": 1.1859,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4266393442622951,
+      "grad_norm": 6.603305339813232,
+      "learning_rate": 1.2825140843316202e-05,
+      "loss": 1.0737,
+      "step": 2082
+    },
+    {
+      "epoch": 0.426844262295082,
+      "grad_norm": 3.1204285621643066,
+      "learning_rate": 1.2818772982751471e-05,
+      "loss": 1.1084,
+      "step": 2083
+    },
+    {
+      "epoch": 0.42704918032786887,
+      "grad_norm": 4.369685173034668,
+      "learning_rate": 1.2812403880286188e-05,
+      "loss": 1.1554,
+      "step": 2084
+    },
+    {
+      "epoch": 0.42725409836065575,
+      "grad_norm": 4.024925231933594,
+      "learning_rate": 1.2806033538726459e-05,
+      "loss": 0.9109,
+      "step": 2085
+    },
+    {
+      "epoch": 0.42745901639344264,
+      "grad_norm": 2.8431484699249268,
+      "learning_rate": 1.2799661960878948e-05,
+      "loss": 0.9621,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4276639344262295,
+      "grad_norm": 3.453000545501709,
+      "learning_rate": 1.2793289149550856e-05,
+      "loss": 1.0121,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4278688524590164,
+      "grad_norm": 2.6690876483917236,
+      "learning_rate": 1.2786915107549927e-05,
+      "loss": 0.8042,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4280737704918033,
+      "grad_norm": 3.2946808338165283,
+      "learning_rate": 1.2780539837684449e-05,
+      "loss": 1.0597,
+      "step": 2089
+    },
+    {
+      "epoch": 0.42827868852459017,
+      "grad_norm": 3.030409336090088,
+      "learning_rate": 1.277416334276325e-05,
+      "loss": 1.0798,
+      "step": 2090
+    },
+    {
+      "epoch": 0.42848360655737705,
+      "grad_norm": 2.6698787212371826,
+      "learning_rate": 1.2767785625595703e-05,
+      "loss": 0.7116,
+      "step": 2091
+    },
+    {
+      "epoch": 0.42868852459016393,
+      "grad_norm": 3.4560434818267822,
+      "learning_rate": 1.2761406688991713e-05,
+      "loss": 1.0079,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4288934426229508,
+      "grad_norm": 5.137183666229248,
+      "learning_rate": 1.275502653576172e-05,
+      "loss": 1.0514,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4290983606557377,
+      "grad_norm": 4.377679347991943,
+      "learning_rate": 1.2748645168716712e-05,
+      "loss": 0.9337,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4293032786885246,
+      "grad_norm": 3.397491455078125,
+      "learning_rate": 1.27422625906682e-05,
+      "loss": 1.0416,
+      "step": 2095
+    },
+    {
+      "epoch": 0.42950819672131146,
+      "grad_norm": 3.7292704582214355,
+      "learning_rate": 1.2735878804428233e-05,
+      "loss": 1.1664,
+      "step": 2096
+    },
+    {
+      "epoch": 0.42971311475409835,
+      "grad_norm": 2.786606788635254,
+      "learning_rate": 1.2729493812809392e-05,
+      "loss": 1.0455,
+      "step": 2097
+    },
+    {
+      "epoch": 0.42991803278688523,
+      "grad_norm": 3.7055413722991943,
+      "learning_rate": 1.272310761862479e-05,
+      "loss": 1.0345,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4301229508196721,
+      "grad_norm": 9.331771850585938,
+      "learning_rate": 1.271672022468807e-05,
+      "loss": 0.9897,
+      "step": 2099
+    },
+    {
+      "epoch": 0.430327868852459,
+      "grad_norm": 6.140835285186768,
+      "learning_rate": 1.27103316338134e-05,
+      "loss": 1.0688,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4305327868852459,
+      "grad_norm": 3.47841215133667,
+      "learning_rate": 1.270394184881548e-05,
+      "loss": 1.0867,
+      "step": 2101
+    },
+    {
+      "epoch": 0.43073770491803276,
+      "grad_norm": 3.9071993827819824,
+      "learning_rate": 1.2697550872509533e-05,
+      "loss": 1.1721,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4309426229508197,
+      "grad_norm": 4.834946155548096,
+      "learning_rate": 1.2691158707711306e-05,
+      "loss": 1.1536,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4311475409836066,
+      "grad_norm": 3.607616662979126,
+      "learning_rate": 1.2684765357237079e-05,
+      "loss": 1.0357,
+      "step": 2104
+    },
+    {
+      "epoch": 0.43135245901639346,
+      "grad_norm": 3.759399175643921,
+      "learning_rate": 1.267837082390364e-05,
+      "loss": 0.7958,
+      "step": 2105
+    },
+    {
+      "epoch": 0.43155737704918035,
+      "grad_norm": 6.660044193267822,
+      "learning_rate": 1.2671975110528304e-05,
+      "loss": 1.1164,
+      "step": 2106
+    },
+    {
+      "epoch": 0.43176229508196723,
+      "grad_norm": 3.075366973876953,
+      "learning_rate": 1.2665578219928914e-05,
+      "loss": 0.7565,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4319672131147541,
+      "grad_norm": 4.47213888168335,
+      "learning_rate": 1.2659180154923818e-05,
+      "loss": 1.1128,
+      "step": 2108
+    },
+    {
+      "epoch": 0.432172131147541,
+      "grad_norm": 3.275787591934204,
+      "learning_rate": 1.2652780918331894e-05,
+      "loss": 0.8418,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4323770491803279,
+      "grad_norm": 5.148368835449219,
+      "learning_rate": 1.2646380512972521e-05,
+      "loss": 0.914,
+      "step": 2110
+    },
+    {
+      "epoch": 0.43258196721311476,
+      "grad_norm": 3.945760488510132,
+      "learning_rate": 1.2639978941665614e-05,
+      "loss": 1.0924,
+      "step": 2111
+    },
+    {
+      "epoch": 0.43278688524590164,
+      "grad_norm": 3.3796780109405518,
+      "learning_rate": 1.263357620723158e-05,
+      "loss": 1.165,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4329918032786885,
+      "grad_norm": 3.9157185554504395,
+      "learning_rate": 1.2627172312491354e-05,
+      "loss": 0.9015,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4331967213114754,
+      "grad_norm": 6.128211975097656,
+      "learning_rate": 1.2620767260266378e-05,
+      "loss": 1.0697,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4334016393442623,
+      "grad_norm": 2.476499319076538,
+      "learning_rate": 1.2614361053378594e-05,
+      "loss": 1.0781,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4336065573770492,
+      "grad_norm": 5.691343784332275,
+      "learning_rate": 1.260795369465047e-05,
+      "loss": 1.3149,
+      "step": 2116
+    },
+    {
+      "epoch": 0.43381147540983606,
+      "grad_norm": 4.444729804992676,
+      "learning_rate": 1.2601545186904968e-05,
+      "loss": 1.2215,
+      "step": 2117
+    },
+    {
+      "epoch": 0.43401639344262294,
+      "grad_norm": 3.2853927612304688,
+      "learning_rate": 1.2595135532965563e-05,
+      "loss": 0.894,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4342213114754098,
+      "grad_norm": 5.254720687866211,
+      "learning_rate": 1.2588724735656235e-05,
+      "loss": 0.9351,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4344262295081967,
+      "grad_norm": 4.267833232879639,
+      "learning_rate": 1.2582312797801459e-05,
+      "loss": 1.0142,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4346311475409836,
+      "grad_norm": 7.486599922180176,
+      "learning_rate": 1.2575899722226229e-05,
+      "loss": 0.9172,
+      "step": 2121
+    },
+    {
+      "epoch": 0.43483606557377047,
+      "grad_norm": 5.801318168640137,
+      "learning_rate": 1.2569485511756019e-05,
+      "loss": 1.1572,
+      "step": 2122
+    },
+    {
+      "epoch": 0.43504098360655735,
+      "grad_norm": 2.5539071559906006,
+      "learning_rate": 1.256307016921682e-05,
+      "loss": 1.015,
+      "step": 2123
+    },
+    {
+      "epoch": 0.43524590163934423,
+      "grad_norm": 2.9481327533721924,
+      "learning_rate": 1.2556653697435119e-05,
+      "loss": 1.2075,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4354508196721312,
+      "grad_norm": 4.556431770324707,
+      "learning_rate": 1.2550236099237896e-05,
+      "loss": 1.1489,
+      "step": 2125
+    },
+    {
+      "epoch": 0.43565573770491806,
+      "grad_norm": 5.150708198547363,
+      "learning_rate": 1.2543817377452628e-05,
+      "loss": 0.8078,
+      "step": 2126
+    },
+    {
+      "epoch": 0.43586065573770494,
+      "grad_norm": 5.656988143920898,
+      "learning_rate": 1.2537397534907288e-05,
+      "loss": 1.01,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4360655737704918,
+      "grad_norm": 3.7042312622070312,
+      "learning_rate": 1.2530976574430342e-05,
+      "loss": 0.8784,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4362704918032787,
+      "grad_norm": 5.61859130859375,
+      "learning_rate": 1.2524554498850752e-05,
+      "loss": 0.9808,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4364754098360656,
+      "grad_norm": 3.552358388900757,
+      "learning_rate": 1.2518131310997968e-05,
+      "loss": 1.2708,
+      "step": 2130
+    },
+    {
+      "epoch": 0.43668032786885247,
+      "grad_norm": 3.745239019393921,
+      "learning_rate": 1.2511707013701928e-05,
+      "loss": 0.8995,
+      "step": 2131
+    },
+    {
+      "epoch": 0.43688524590163935,
+      "grad_norm": 4.721340179443359,
+      "learning_rate": 1.2505281609793066e-05,
+      "loss": 1.2726,
+      "step": 2132
+    },
+    {
+      "epoch": 0.43709016393442623,
+      "grad_norm": 3.0570333003997803,
+      "learning_rate": 1.2498855102102297e-05,
+      "loss": 1.1515,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4372950819672131,
+      "grad_norm": 3.701097011566162,
+      "learning_rate": 1.2492427493461024e-05,
+      "loss": 0.9439,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 16.614437103271484,
+      "learning_rate": 1.2485998786701135e-05,
+      "loss": 1.0592,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4377049180327869,
+      "grad_norm": 12.8952054977417,
+      "learning_rate": 1.2479568984655002e-05,
+      "loss": 0.9997,
+      "step": 2136
+    },
+    {
+      "epoch": 0.43790983606557377,
+      "grad_norm": 4.846699237823486,
+      "learning_rate": 1.2473138090155483e-05,
+      "loss": 1.1472,
+      "step": 2137
+    },
+    {
+      "epoch": 0.43811475409836065,
+      "grad_norm": 6.402317523956299,
+      "learning_rate": 1.2466706106035911e-05,
+      "loss": 1.1281,
+      "step": 2138
+    },
+    {
+      "epoch": 0.43831967213114753,
+      "grad_norm": 2.8718554973602295,
+      "learning_rate": 1.2460273035130106e-05,
+      "loss": 0.971,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4385245901639344,
+      "grad_norm": 4.165424346923828,
+      "learning_rate": 1.2453838880272358e-05,
+      "loss": 0.9212,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4387295081967213,
+      "grad_norm": 3.0852088928222656,
+      "learning_rate": 1.244740364429744e-05,
+      "loss": 1.1299,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4389344262295082,
+      "grad_norm": 5.05717658996582,
+      "learning_rate": 1.2440967330040606e-05,
+      "loss": 0.9412,
+      "step": 2142
+    },
+    {
+      "epoch": 0.43913934426229506,
+      "grad_norm": 3.1594045162200928,
+      "learning_rate": 1.2434529940337576e-05,
+      "loss": 1.0387,
+      "step": 2143
+    },
+    {
+      "epoch": 0.43934426229508194,
+      "grad_norm": 4.51142692565918,
+      "learning_rate": 1.242809147802455e-05,
+      "loss": 1.1285,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4395491803278688,
+      "grad_norm": 3.3049263954162598,
+      "learning_rate": 1.2421651945938194e-05,
+      "loss": 1.2325,
+      "step": 2145
+    },
+    {
+      "epoch": 0.43975409836065577,
+      "grad_norm": 4.288992881774902,
+      "learning_rate": 1.2415211346915657e-05,
+      "loss": 1.0566,
+      "step": 2146
+    },
+    {
+      "epoch": 0.43995901639344265,
+      "grad_norm": 3.373507261276245,
+      "learning_rate": 1.2408769683794544e-05,
+      "loss": 1.1644,
+      "step": 2147
+    },
+    {
+      "epoch": 0.44016393442622953,
+      "grad_norm": 3.349897861480713,
+      "learning_rate": 1.2402326959412938e-05,
+      "loss": 0.9433,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4403688524590164,
+      "grad_norm": 3.7010955810546875,
+      "learning_rate": 1.2395883176609391e-05,
+      "loss": 0.9023,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4405737704918033,
+      "grad_norm": 3.1350767612457275,
+      "learning_rate": 1.238943833822291e-05,
+      "loss": 0.9999,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4407786885245902,
+      "grad_norm": 5.908895015716553,
+      "learning_rate": 1.2382992447092983e-05,
+      "loss": 1.1022,
+      "step": 2151
+    },
+    {
+      "epoch": 0.44098360655737706,
+      "grad_norm": 7.280833721160889,
+      "learning_rate": 1.2376545506059545e-05,
+      "loss": 1.2506,
+      "step": 2152
+    },
+    {
+      "epoch": 0.44118852459016394,
+      "grad_norm": 3.043593645095825,
+      "learning_rate": 1.2370097517963006e-05,
+      "loss": 0.8414,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4413934426229508,
+      "grad_norm": 6.007325172424316,
+      "learning_rate": 1.2363648485644237e-05,
+      "loss": 1.1925,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4415983606557377,
+      "grad_norm": 5.704403877258301,
+      "learning_rate": 1.235719841194456e-05,
+      "loss": 0.9061,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4418032786885246,
+      "grad_norm": 6.445374965667725,
+      "learning_rate": 1.2350747299705763e-05,
+      "loss": 1.0315,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4420081967213115,
+      "grad_norm": 6.552412033081055,
+      "learning_rate": 1.2344295151770087e-05,
+      "loss": 1.1128,
+      "step": 2157
+    },
+    {
+      "epoch": 0.44221311475409836,
+      "grad_norm": 7.215743064880371,
+      "learning_rate": 1.233784197098024e-05,
+      "loss": 1.0393,
+      "step": 2158
+    },
+    {
+      "epoch": 0.44241803278688524,
+      "grad_norm": 4.774063587188721,
+      "learning_rate": 1.2331387760179368e-05,
+      "loss": 0.8902,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4426229508196721,
+      "grad_norm": 3.059889793395996,
+      "learning_rate": 1.2324932522211084e-05,
+      "loss": 0.9165,
+      "step": 2160
+    },
+    {
+      "epoch": 0.442827868852459,
+      "grad_norm": 4.437394142150879,
+      "learning_rate": 1.2318476259919451e-05,
+      "loss": 0.957,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4430327868852459,
+      "grad_norm": 2.400036096572876,
+      "learning_rate": 1.2312018976148976e-05,
+      "loss": 0.9905,
+      "step": 2162
+    },
+    {
+      "epoch": 0.44323770491803277,
+      "grad_norm": 3.7056362628936768,
+      "learning_rate": 1.2305560673744631e-05,
+      "loss": 0.8704,
+      "step": 2163
+    },
+    {
+      "epoch": 0.44344262295081965,
+      "grad_norm": 12.103838920593262,
+      "learning_rate": 1.2299101355551822e-05,
+      "loss": 1.2474,
+      "step": 2164
+    },
+    {
+      "epoch": 0.44364754098360654,
+      "grad_norm": 4.79158353805542,
+      "learning_rate": 1.2292641024416404e-05,
+      "loss": 0.9783,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4438524590163934,
+      "grad_norm": 2.9580142498016357,
+      "learning_rate": 1.228617968318469e-05,
+      "loss": 0.9058,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4440573770491803,
+      "grad_norm": 3.7836079597473145,
+      "learning_rate": 1.2279717334703424e-05,
+      "loss": 0.9278,
+      "step": 2167
+    },
+    {
+      "epoch": 0.44426229508196724,
+      "grad_norm": 2.361546277999878,
+      "learning_rate": 1.227325398181981e-05,
+      "loss": 0.8414,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4444672131147541,
+      "grad_norm": 3.096489906311035,
+      "learning_rate": 1.2266789627381475e-05,
+      "loss": 1.1027,
+      "step": 2169
+    },
+    {
+      "epoch": 0.444672131147541,
+      "grad_norm": 3.2285475730895996,
+      "learning_rate": 1.2260324274236499e-05,
+      "loss": 0.9128,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4448770491803279,
+      "grad_norm": 2.9011101722717285,
+      "learning_rate": 1.2253857925233401e-05,
+      "loss": 1.0439,
+      "step": 2171
+    },
+    {
+      "epoch": 0.44508196721311477,
+      "grad_norm": 7.82223653793335,
+      "learning_rate": 1.224739058322114e-05,
+      "loss": 1.2395,
+      "step": 2172
+    },
+    {
+      "epoch": 0.44528688524590165,
+      "grad_norm": 2.782752275466919,
+      "learning_rate": 1.2240922251049105e-05,
+      "loss": 1.0172,
+      "step": 2173
+    },
+    {
+      "epoch": 0.44549180327868854,
+      "grad_norm": 7.223694324493408,
+      "learning_rate": 1.2234452931567135e-05,
+      "loss": 1.0785,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4456967213114754,
+      "grad_norm": 5.377253532409668,
+      "learning_rate": 1.222798262762549e-05,
+      "loss": 0.8013,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4459016393442623,
+      "grad_norm": 3.2472565174102783,
+      "learning_rate": 1.2221511342074867e-05,
+      "loss": 1.0451,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4461065573770492,
+      "grad_norm": 4.081872463226318,
+      "learning_rate": 1.2215039077766402e-05,
+      "loss": 0.8517,
+      "step": 2177
+    },
+    {
+      "epoch": 0.44631147540983607,
+      "grad_norm": 3.378549575805664,
+      "learning_rate": 1.2208565837551653e-05,
+      "loss": 1.1263,
+      "step": 2178
+    },
+    {
+      "epoch": 0.44651639344262295,
+      "grad_norm": 6.841521263122559,
+      "learning_rate": 1.2202091624282622e-05,
+      "loss": 1.0325,
+      "step": 2179
+    },
+    {
+      "epoch": 0.44672131147540983,
+      "grad_norm": 3.775071620941162,
+      "learning_rate": 1.219561644081172e-05,
+      "loss": 0.7168,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4469262295081967,
+      "grad_norm": 4.82589864730835,
+      "learning_rate": 1.2189140289991805e-05,
+      "loss": 0.9461,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4471311475409836,
+      "grad_norm": 3.729494333267212,
+      "learning_rate": 1.2182663174676147e-05,
+      "loss": 1.0077,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4473360655737705,
+      "grad_norm": 5.081676483154297,
+      "learning_rate": 1.2176185097718447e-05,
+      "loss": 1.0188,
+      "step": 2183
+    },
+    {
+      "epoch": 0.44754098360655736,
+      "grad_norm": 7.323000431060791,
+      "learning_rate": 1.216970606197283e-05,
+      "loss": 0.9469,
+      "step": 2184
+    },
+    {
+      "epoch": 0.44774590163934425,
+      "grad_norm": 5.472822666168213,
+      "learning_rate": 1.2163226070293841e-05,
+      "loss": 0.8897,
+      "step": 2185
+    },
+    {
+      "epoch": 0.44795081967213113,
+      "grad_norm": 3.8541598320007324,
+      "learning_rate": 1.215674512553645e-05,
+      "loss": 1.065,
+      "step": 2186
+    },
+    {
+      "epoch": 0.448155737704918,
+      "grad_norm": 2.8564958572387695,
+      "learning_rate": 1.2150263230556041e-05,
+      "loss": 0.9552,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4483606557377049,
+      "grad_norm": 4.54953670501709,
+      "learning_rate": 1.2143780388208424e-05,
+      "loss": 0.9898,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4485655737704918,
+      "grad_norm": 5.728074550628662,
+      "learning_rate": 1.2137296601349817e-05,
+      "loss": 0.952,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4487704918032787,
+      "grad_norm": 3.315885543823242,
+      "learning_rate": 1.2130811872836865e-05,
+      "loss": 0.8861,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4489754098360656,
+      "grad_norm": 4.513580322265625,
+      "learning_rate": 1.2124326205526622e-05,
+      "loss": 1.0801,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4491803278688525,
+      "grad_norm": 10.095090866088867,
+      "learning_rate": 1.211783960227655e-05,
+      "loss": 1.0935,
+      "step": 2192
+    },
+    {
+      "epoch": 0.44938524590163936,
+      "grad_norm": 3.776561737060547,
+      "learning_rate": 1.2111352065944536e-05,
+      "loss": 0.9007,
+      "step": 2193
+    },
+    {
+      "epoch": 0.44959016393442625,
+      "grad_norm": 6.468218803405762,
+      "learning_rate": 1.2104863599388873e-05,
+      "loss": 0.8815,
+      "step": 2194
+    },
+    {
+      "epoch": 0.44979508196721313,
+      "grad_norm": 7.2963995933532715,
+      "learning_rate": 1.2098374205468253e-05,
+      "loss": 1.1407,
+      "step": 2195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.470483779907227,
+      "learning_rate": 1.2091883887041795e-05,
+      "loss": 1.2084,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4502049180327869,
+      "grad_norm": 4.2841410636901855,
+      "learning_rate": 1.208539264696901e-05,
+      "loss": 0.9077,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4504098360655738,
+      "grad_norm": 5.1359734535217285,
+      "learning_rate": 1.2078900488109828e-05,
+      "loss": 1.0116,
+      "step": 2198
+    },
+    {
+      "epoch": 0.45061475409836066,
+      "grad_norm": 3.9523978233337402,
+      "learning_rate": 1.2072407413324572e-05,
+      "loss": 1.0341,
+      "step": 2199
+    },
+    {
+      "epoch": 0.45081967213114754,
+      "grad_norm": 3.6387507915496826,
+      "learning_rate": 1.2065913425473973e-05,
+      "loss": 1.1019,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4510245901639344,
+      "grad_norm": 5.512958526611328,
+      "learning_rate": 1.205941852741917e-05,
+      "loss": 1.1319,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4512295081967213,
+      "grad_norm": 2.518993854522705,
+      "learning_rate": 1.2052922722021694e-05,
+      "loss": 0.9448,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4514344262295082,
+      "grad_norm": 7.6417012214660645,
+      "learning_rate": 1.2046426012143484e-05,
+      "loss": 1.2606,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4516393442622951,
+      "grad_norm": 4.063118934631348,
+      "learning_rate": 1.2039928400646865e-05,
+      "loss": 1.1774,
+      "step": 2204
+    },
+    {
+      "epoch": 0.45184426229508196,
+      "grad_norm": 6.36499547958374,
+      "learning_rate": 1.2033429890394575e-05,
+      "loss": 1.0382,
+      "step": 2205
+    },
+    {
+      "epoch": 0.45204918032786884,
+      "grad_norm": 9.054524421691895,
+      "learning_rate": 1.2026930484249738e-05,
+      "loss": 0.9437,
+      "step": 2206
+    },
+    {
+      "epoch": 0.4522540983606557,
+      "grad_norm": 5.376067161560059,
+      "learning_rate": 1.2020430185075874e-05,
+      "loss": 1.0811,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4524590163934426,
+      "grad_norm": 4.087770938873291,
+      "learning_rate": 1.2013928995736898e-05,
+      "loss": 0.8194,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4526639344262295,
+      "grad_norm": 13.804874420166016,
+      "learning_rate": 1.200742691909712e-05,
+      "loss": 0.939,
+      "step": 2209
+    },
+    {
+      "epoch": 0.45286885245901637,
+      "grad_norm": 3.374622344970703,
+      "learning_rate": 1.2000923958021235e-05,
+      "loss": 0.8777,
+      "step": 2210
+    },
+    {
+      "epoch": 0.45307377049180325,
+      "grad_norm": 3.3910882472991943,
+      "learning_rate": 1.199442011537433e-05,
+      "loss": 1.0049,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4532786885245902,
+      "grad_norm": 3.6709303855895996,
+      "learning_rate": 1.198791539402188e-05,
+      "loss": 1.0492,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4534836065573771,
+      "grad_norm": 4.9755167961120605,
+      "learning_rate": 1.1981409796829753e-05,
+      "loss": 1.0113,
+      "step": 2213
+    },
+    {
+      "epoch": 0.45368852459016396,
+      "grad_norm": 2.907677412033081,
+      "learning_rate": 1.1974903326664191e-05,
+      "loss": 0.7893,
+      "step": 2214
+    },
+    {
+      "epoch": 0.45389344262295084,
+      "grad_norm": 4.66806697845459,
+      "learning_rate": 1.1968395986391832e-05,
+      "loss": 1.0443,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4540983606557377,
+      "grad_norm": 4.390023708343506,
+      "learning_rate": 1.1961887778879693e-05,
+      "loss": 1.0252,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4543032786885246,
+      "grad_norm": 6.241100788116455,
+      "learning_rate": 1.195537870699517e-05,
+      "loss": 1.2693,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4545081967213115,
+      "grad_norm": 4.752009868621826,
+      "learning_rate": 1.1948868773606044e-05,
+      "loss": 0.9618,
+      "step": 2218
+    },
+    {
+      "epoch": 0.45471311475409837,
+      "grad_norm": 5.363521099090576,
+      "learning_rate": 1.1942357981580472e-05,
+      "loss": 0.9092,
+      "step": 2219
+    },
+    {
+      "epoch": 0.45491803278688525,
+      "grad_norm": 3.2249221801757812,
+      "learning_rate": 1.1935846333786998e-05,
+      "loss": 1.0715,
+      "step": 2220
+    },
+    {
+      "epoch": 0.45512295081967213,
+      "grad_norm": 5.101828575134277,
+      "learning_rate": 1.1929333833094532e-05,
+      "loss": 1.0128,
+      "step": 2221
+    },
+    {
+      "epoch": 0.455327868852459,
+      "grad_norm": 4.146220684051514,
+      "learning_rate": 1.1922820482372362e-05,
+      "loss": 0.9411,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4555327868852459,
+      "grad_norm": 6.718583583831787,
+      "learning_rate": 1.191630628449016e-05,
+      "loss": 0.7002,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4557377049180328,
+      "grad_norm": 5.399123668670654,
+      "learning_rate": 1.1909791242317957e-05,
+      "loss": 1.317,
+      "step": 2224
+    },
+    {
+      "epoch": 0.45594262295081966,
+      "grad_norm": 5.5568013191223145,
+      "learning_rate": 1.1903275358726168e-05,
+      "loss": 0.9962,
+      "step": 2225
+    },
+    {
+      "epoch": 0.45614754098360655,
+      "grad_norm": 6.204396724700928,
+      "learning_rate": 1.189675863658557e-05,
+      "loss": 1.178,
+      "step": 2226
+    },
+    {
+      "epoch": 0.45635245901639343,
+      "grad_norm": 5.663088321685791,
+      "learning_rate": 1.1890241078767316e-05,
+      "loss": 1.0359,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4565573770491803,
+      "grad_norm": 4.537768363952637,
+      "learning_rate": 1.1883722688142925e-05,
+      "loss": 1.1882,
+      "step": 2228
+    },
+    {
+      "epoch": 0.4567622950819672,
+      "grad_norm": 7.045298099517822,
+      "learning_rate": 1.187720346758428e-05,
+      "loss": 0.8725,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4569672131147541,
+      "grad_norm": 8.433988571166992,
+      "learning_rate": 1.1870683419963631e-05,
+      "loss": 1.0577,
+      "step": 2230
+    },
+    {
+      "epoch": 0.45717213114754096,
+      "grad_norm": 3.7153515815734863,
+      "learning_rate": 1.18641625481536e-05,
+      "loss": 0.9696,
+      "step": 2231
+    },
+    {
+      "epoch": 0.45737704918032784,
+      "grad_norm": 3.2031331062316895,
+      "learning_rate": 1.1857640855027157e-05,
+      "loss": 0.7982,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4575819672131147,
+      "grad_norm": 5.166006565093994,
+      "learning_rate": 1.185111834345765e-05,
+      "loss": 1.028,
+      "step": 2233
+    },
+    {
+      "epoch": 0.45778688524590166,
+      "grad_norm": 3.5086190700531006,
+      "learning_rate": 1.1844595016318774e-05,
+      "loss": 0.8105,
+      "step": 2234
+    },
+    {
+      "epoch": 0.45799180327868855,
+      "grad_norm": 3.4865915775299072,
+      "learning_rate": 1.1838070876484593e-05,
+      "loss": 1.1736,
+      "step": 2235
+    },
+    {
+      "epoch": 0.45819672131147543,
+      "grad_norm": 6.831033229827881,
+      "learning_rate": 1.1831545926829522e-05,
+      "loss": 1.4684,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4584016393442623,
+      "grad_norm": 3.929307460784912,
+      "learning_rate": 1.1825020170228341e-05,
+      "loss": 0.9376,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4586065573770492,
+      "grad_norm": 4.63611364364624,
+      "learning_rate": 1.181849360955618e-05,
+      "loss": 0.9013,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4588114754098361,
+      "grad_norm": 3.664745807647705,
+      "learning_rate": 1.1811966247688516e-05,
+      "loss": 1.1066,
+      "step": 2239
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 3.9980247020721436,
+      "learning_rate": 1.1805438087501198e-05,
+      "loss": 1.0084,
+      "step": 2240
+    },
+    {
+      "epoch": 0.45922131147540984,
+      "grad_norm": 4.673801898956299,
+      "learning_rate": 1.1798909131870407e-05,
+      "loss": 1.0355,
+      "step": 2241
+    },
+    {
+      "epoch": 0.4594262295081967,
+      "grad_norm": 3.0889580249786377,
+      "learning_rate": 1.179237938367269e-05,
+      "loss": 0.9626,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4596311475409836,
+      "grad_norm": 3.758612632751465,
+      "learning_rate": 1.178584884578493e-05,
+      "loss": 0.8674,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4598360655737705,
+      "grad_norm": 3.518897533416748,
+      "learning_rate": 1.1779317521084368e-05,
+      "loss": 0.8344,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4600409836065574,
+      "grad_norm": 3.687730073928833,
+      "learning_rate": 1.1772785412448591e-05,
+      "loss": 1.0063,
+      "step": 2245
+    },
+    {
+      "epoch": 0.46024590163934426,
+      "grad_norm": 12.569208145141602,
+      "learning_rate": 1.1766252522755521e-05,
+      "loss": 1.1131,
+      "step": 2246
+    },
+    {
+      "epoch": 0.46045081967213114,
+      "grad_norm": 7.914841651916504,
+      "learning_rate": 1.1759718854883434e-05,
+      "loss": 0.868,
+      "step": 2247
+    },
+    {
+      "epoch": 0.460655737704918,
+      "grad_norm": 3.916210889816284,
+      "learning_rate": 1.175318441171095e-05,
+      "loss": 0.9297,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4608606557377049,
+      "grad_norm": 2.7618601322174072,
+      "learning_rate": 1.1746649196117024e-05,
+      "loss": 0.8389,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4610655737704918,
+      "grad_norm": 3.413602352142334,
+      "learning_rate": 1.1740113210980952e-05,
+      "loss": 0.6869,
+      "step": 2250
+    },
+    {
+      "epoch": 0.46127049180327867,
+      "grad_norm": 4.273172378540039,
+      "learning_rate": 1.1733576459182376e-05,
+      "loss": 1.2029,
+      "step": 2251
+    },
+    {
+      "epoch": 0.46147540983606555,
+      "grad_norm": 5.626925468444824,
+      "learning_rate": 1.1727038943601268e-05,
+      "loss": 1.0944,
+      "step": 2252
+    },
+    {
+      "epoch": 0.46168032786885244,
+      "grad_norm": 4.22699499130249,
+      "learning_rate": 1.1720500667117937e-05,
+      "loss": 1.0263,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4618852459016393,
+      "grad_norm": 4.072995662689209,
+      "learning_rate": 1.1713961632613036e-05,
+      "loss": 1.0358,
+      "step": 2254
+    },
+    {
+      "epoch": 0.46209016393442626,
+      "grad_norm": 2.8077518939971924,
+      "learning_rate": 1.1707421842967541e-05,
+      "loss": 1.0769,
+      "step": 2255
+    },
+    {
+      "epoch": 0.46229508196721314,
+      "grad_norm": 4.799870491027832,
+      "learning_rate": 1.1700881301062767e-05,
+      "loss": 1.3439,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 4.8334856033325195,
+      "learning_rate": 1.1694340009780356e-05,
+      "loss": 1.0639,
+      "step": 2257
+    },
+    {
+      "epoch": 0.4627049180327869,
+      "grad_norm": 4.4328694343566895,
+      "learning_rate": 1.1687797972002291e-05,
+      "loss": 1.024,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4629098360655738,
+      "grad_norm": 5.5,
+      "learning_rate": 1.1681255190610868e-05,
+      "loss": 0.9335,
+      "step": 2259
+    },
+    {
+      "epoch": 0.46311475409836067,
+      "grad_norm": 5.060070514678955,
+      "learning_rate": 1.1674711668488719e-05,
+      "loss": 0.9159,
+      "step": 2260
+    },
+    {
+      "epoch": 0.46331967213114755,
+      "grad_norm": 2.671255111694336,
+      "learning_rate": 1.1668167408518808e-05,
+      "loss": 0.9079,
+      "step": 2261
+    },
+    {
+      "epoch": 0.46352459016393444,
+      "grad_norm": 4.269006729125977,
+      "learning_rate": 1.1661622413584408e-05,
+      "loss": 0.6925,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4637295081967213,
+      "grad_norm": 4.751982688903809,
+      "learning_rate": 1.1655076686569137e-05,
+      "loss": 1.0628,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4639344262295082,
+      "grad_norm": 8.520039558410645,
+      "learning_rate": 1.1648530230356914e-05,
+      "loss": 1.191,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4641393442622951,
+      "grad_norm": 3.7908153533935547,
+      "learning_rate": 1.1641983047832e-05,
+      "loss": 0.9266,
+      "step": 2265
+    },
+    {
+      "epoch": 0.46434426229508197,
+      "grad_norm": 3.7034263610839844,
+      "learning_rate": 1.1635435141878957e-05,
+      "loss": 1.0836,
+      "step": 2266
+    },
+    {
+      "epoch": 0.46454918032786885,
+      "grad_norm": 5.963608264923096,
+      "learning_rate": 1.1628886515382675e-05,
+      "loss": 1.2506,
+      "step": 2267
+    },
+    {
+      "epoch": 0.46475409836065573,
+      "grad_norm": 7.731484889984131,
+      "learning_rate": 1.1622337171228368e-05,
+      "loss": 1.0275,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4649590163934426,
+      "grad_norm": 6.0005035400390625,
+      "learning_rate": 1.161578711230155e-05,
+      "loss": 1.3609,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4651639344262295,
+      "grad_norm": 6.219449043273926,
+      "learning_rate": 1.1609236341488066e-05,
+      "loss": 1.1183,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4653688524590164,
+      "grad_norm": 5.036273002624512,
+      "learning_rate": 1.1602684861674064e-05,
+      "loss": 0.9673,
+      "step": 2271
+    },
+    {
+      "epoch": 0.46557377049180326,
+      "grad_norm": 3.9358813762664795,
+      "learning_rate": 1.1596132675746006e-05,
+      "loss": 0.7,
+      "step": 2272
+    },
+    {
+      "epoch": 0.46577868852459015,
+      "grad_norm": 5.283418655395508,
+      "learning_rate": 1.1589579786590673e-05,
+      "loss": 1.1319,
+      "step": 2273
+    },
+    {
+      "epoch": 0.46598360655737703,
+      "grad_norm": 5.680464267730713,
+      "learning_rate": 1.1583026197095141e-05,
+      "loss": 1.2031,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4661885245901639,
+      "grad_norm": 3.147228956222534,
+      "learning_rate": 1.1576471910146815e-05,
+      "loss": 1.0085,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4663934426229508,
+      "grad_norm": 4.565055847167969,
+      "learning_rate": 1.1569916928633385e-05,
+      "loss": 1.024,
+      "step": 2276
+    },
+    {
+      "epoch": 0.46659836065573773,
+      "grad_norm": 2.8703980445861816,
+      "learning_rate": 1.156336125544286e-05,
+      "loss": 1.1299,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4668032786885246,
+      "grad_norm": 2.8556652069091797,
+      "learning_rate": 1.1556804893463556e-05,
+      "loss": 1.0912,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4670081967213115,
+      "grad_norm": 4.4446892738342285,
+      "learning_rate": 1.1550247845584084e-05,
+      "loss": 1.0958,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4672131147540984,
+      "grad_norm": 3.6720409393310547,
+      "learning_rate": 1.1543690114693364e-05,
+      "loss": 0.9187,
+      "step": 2280
+    },
+    {
+      "epoch": 0.46741803278688526,
+      "grad_norm": 3.79304575920105,
+      "learning_rate": 1.153713170368061e-05,
+      "loss": 1.2282,
+      "step": 2281
+    },
+    {
+      "epoch": 0.46762295081967215,
+      "grad_norm": 3.704146146774292,
+      "learning_rate": 1.1530572615435343e-05,
+      "loss": 1.0376,
+      "step": 2282
+    },
+    {
+      "epoch": 0.46782786885245903,
+      "grad_norm": 4.554152011871338,
+      "learning_rate": 1.1524012852847377e-05,
+      "loss": 0.9538,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4680327868852459,
+      "grad_norm": 8.657910346984863,
+      "learning_rate": 1.1517452418806826e-05,
+      "loss": 1.1442,
+      "step": 2284
+    },
+    {
+      "epoch": 0.4682377049180328,
+      "grad_norm": 3.4082303047180176,
+      "learning_rate": 1.15108913162041e-05,
+      "loss": 1.0586,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4684426229508197,
+      "grad_norm": 2.8735926151275635,
+      "learning_rate": 1.1504329547929901e-05,
+      "loss": 1.1133,
+      "step": 2286
+    },
+    {
+      "epoch": 0.46864754098360656,
+      "grad_norm": 2.8333446979522705,
+      "learning_rate": 1.1497767116875227e-05,
+      "loss": 1.026,
+      "step": 2287
+    },
+    {
+      "epoch": 0.46885245901639344,
+      "grad_norm": 5.56522798538208,
+      "learning_rate": 1.1491204025931363e-05,
+      "loss": 0.9788,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4690573770491803,
+      "grad_norm": 6.175342559814453,
+      "learning_rate": 1.1484640277989894e-05,
+      "loss": 1.0578,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4692622950819672,
+      "grad_norm": 2.984683036804199,
+      "learning_rate": 1.1478075875942688e-05,
+      "loss": 0.8885,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4694672131147541,
+      "grad_norm": 2.7350122928619385,
+      "learning_rate": 1.1471510822681898e-05,
+      "loss": 0.9938,
+      "step": 2291
+    },
+    {
+      "epoch": 0.46967213114754097,
+      "grad_norm": 4.135010719299316,
+      "learning_rate": 1.1464945121099971e-05,
+      "loss": 0.9954,
+      "step": 2292
+    },
+    {
+      "epoch": 0.46987704918032785,
+      "grad_norm": 4.406982898712158,
+      "learning_rate": 1.1458378774089638e-05,
+      "loss": 1.2162,
+      "step": 2293
+    },
+    {
+      "epoch": 0.47008196721311474,
+      "grad_norm": 2.4742393493652344,
+      "learning_rate": 1.145181178454391e-05,
+      "loss": 0.7823,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4702868852459016,
+      "grad_norm": 5.053757667541504,
+      "learning_rate": 1.1445244155356087e-05,
+      "loss": 0.8801,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4704918032786885,
+      "grad_norm": 5.7018303871154785,
+      "learning_rate": 1.1438675889419746e-05,
+      "loss": 0.9133,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4706967213114754,
+      "grad_norm": 3.510981321334839,
+      "learning_rate": 1.1432106989628748e-05,
+      "loss": 0.9215,
+      "step": 2297
+    },
+    {
+      "epoch": 0.47090163934426227,
+      "grad_norm": 2.8304378986358643,
+      "learning_rate": 1.1425537458877232e-05,
+      "loss": 0.8446,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4711065573770492,
+      "grad_norm": 3.9545788764953613,
+      "learning_rate": 1.141896730005961e-05,
+      "loss": 0.8089,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4713114754098361,
+      "grad_norm": 5.172813415527344,
+      "learning_rate": 1.1412396516070585e-05,
+      "loss": 0.9764,
+      "step": 2300
+    },
+    {
+      "epoch": 0.47151639344262297,
+      "grad_norm": 2.8520333766937256,
+      "learning_rate": 1.140582510980512e-05,
+      "loss": 0.7991,
+      "step": 2301
+    },
+    {
+      "epoch": 0.47172131147540985,
+      "grad_norm": 3.207030773162842,
+      "learning_rate": 1.1399253084158455e-05,
+      "loss": 1.0941,
+      "step": 2302
+    },
+    {
+      "epoch": 0.47192622950819674,
+      "grad_norm": 5.182816505432129,
+      "learning_rate": 1.1392680442026114e-05,
+      "loss": 0.9944,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4721311475409836,
+      "grad_norm": 4.613626480102539,
+      "learning_rate": 1.1386107186303877e-05,
+      "loss": 1.0266,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4723360655737705,
+      "grad_norm": 4.133020877838135,
+      "learning_rate": 1.1379533319887805e-05,
+      "loss": 0.939,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4725409836065574,
+      "grad_norm": 5.417807102203369,
+      "learning_rate": 1.1372958845674226e-05,
+      "loss": 1.0196,
+      "step": 2306
+    },
+    {
+      "epoch": 0.47274590163934427,
+      "grad_norm": 4.906064510345459,
+      "learning_rate": 1.1366383766559734e-05,
+      "loss": 1.2204,
+      "step": 2307
+    },
+    {
+      "epoch": 0.47295081967213115,
+      "grad_norm": 6.330846309661865,
+      "learning_rate": 1.1359808085441194e-05,
+      "loss": 1.0774,
+      "step": 2308
+    },
+    {
+      "epoch": 0.47315573770491803,
+      "grad_norm": 4.9008564949035645,
+      "learning_rate": 1.1353231805215722e-05,
+      "loss": 1.1131,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4733606557377049,
+      "grad_norm": 4.528264045715332,
+      "learning_rate": 1.134665492878072e-05,
+      "loss": 1.0091,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4735655737704918,
+      "grad_norm": 6.418875217437744,
+      "learning_rate": 1.1340077459033836e-05,
+      "loss": 1.0141,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4737704918032787,
+      "grad_norm": 3.4506020545959473,
+      "learning_rate": 1.1333499398872983e-05,
+      "loss": 1.0817,
+      "step": 2312
+    },
+    {
+      "epoch": 0.47397540983606556,
+      "grad_norm": 4.767189025878906,
+      "learning_rate": 1.132692075119634e-05,
+      "loss": 1.0609,
+      "step": 2313
+    },
+    {
+      "epoch": 0.47418032786885245,
+      "grad_norm": 2.7522125244140625,
+      "learning_rate": 1.1320341518902336e-05,
+      "loss": 0.7748,
+      "step": 2314
+    },
+    {
+      "epoch": 0.47438524590163933,
+      "grad_norm": 5.039668083190918,
+      "learning_rate": 1.1313761704889668e-05,
+      "loss": 1.131,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4745901639344262,
+      "grad_norm": 5.564464569091797,
+      "learning_rate": 1.1307181312057275e-05,
+      "loss": 0.8508,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4747950819672131,
+      "grad_norm": 4.54846715927124,
+      "learning_rate": 1.130060034330437e-05,
+      "loss": 0.9142,
+      "step": 2317
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 4.100391864776611,
+      "learning_rate": 1.12940188015304e-05,
+      "loss": 1.0753,
+      "step": 2318
+    },
+    {
+      "epoch": 0.47520491803278686,
+      "grad_norm": 4.676609039306641,
+      "learning_rate": 1.128743668963508e-05,
+      "loss": 1.2459,
+      "step": 2319
+    },
+    {
+      "epoch": 0.47540983606557374,
+      "grad_norm": 2.823354482650757,
+      "learning_rate": 1.1280854010518364e-05,
+      "loss": 1.1359,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4756147540983607,
+      "grad_norm": 3.3732922077178955,
+      "learning_rate": 1.1274270767080465e-05,
+      "loss": 0.8237,
+      "step": 2321
+    },
+    {
+      "epoch": 0.47581967213114756,
+      "grad_norm": 5.350398540496826,
+      "learning_rate": 1.1267686962221843e-05,
+      "loss": 1.0487,
+      "step": 2322
+    },
+    {
+      "epoch": 0.47602459016393445,
+      "grad_norm": 2.6470279693603516,
+      "learning_rate": 1.1261102598843198e-05,
+      "loss": 0.7326,
+      "step": 2323
+    },
+    {
+      "epoch": 0.47622950819672133,
+      "grad_norm": 6.9887847900390625,
+      "learning_rate": 1.125451767984549e-05,
+      "loss": 1.1099,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4764344262295082,
+      "grad_norm": 3.7903661727905273,
+      "learning_rate": 1.1247932208129909e-05,
+      "loss": 0.9804,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4766393442622951,
+      "grad_norm": 5.6975626945495605,
+      "learning_rate": 1.1241346186597897e-05,
+      "loss": 1.086,
+      "step": 2326
+    },
+    {
+      "epoch": 0.476844262295082,
+      "grad_norm": 6.521388530731201,
+      "learning_rate": 1.1234759618151137e-05,
+      "loss": 1.1192,
+      "step": 2327
+    },
+    {
+      "epoch": 0.47704918032786886,
+      "grad_norm": 4.776442527770996,
+      "learning_rate": 1.1228172505691555e-05,
+      "loss": 0.7688,
+      "step": 2328
+    },
+    {
+      "epoch": 0.47725409836065574,
+      "grad_norm": 3.8123791217803955,
+      "learning_rate": 1.1221584852121307e-05,
+      "loss": 1.0818,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4774590163934426,
+      "grad_norm": 6.605536937713623,
+      "learning_rate": 1.1214996660342799e-05,
+      "loss": 1.1747,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4776639344262295,
+      "grad_norm": 4.678340911865234,
+      "learning_rate": 1.1208407933258669e-05,
+      "loss": 1.2595,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4778688524590164,
+      "grad_norm": 5.203702926635742,
+      "learning_rate": 1.1201818673771791e-05,
+      "loss": 1.1677,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4780737704918033,
+      "grad_norm": 4.20292329788208,
+      "learning_rate": 1.1195228884785276e-05,
+      "loss": 1.2453,
+      "step": 2333
+    },
+    {
+      "epoch": 0.47827868852459016,
+      "grad_norm": 4.4127092361450195,
+      "learning_rate": 1.1188638569202462e-05,
+      "loss": 1.1468,
+      "step": 2334
+    },
+    {
+      "epoch": 0.47848360655737704,
+      "grad_norm": 3.0011723041534424,
+      "learning_rate": 1.1182047729926926e-05,
+      "loss": 0.9007,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4786885245901639,
+      "grad_norm": 5.331995964050293,
+      "learning_rate": 1.1175456369862474e-05,
+      "loss": 1.0983,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4788934426229508,
+      "grad_norm": 2.270374059677124,
+      "learning_rate": 1.1168864491913136e-05,
+      "loss": 0.9827,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4790983606557377,
+      "grad_norm": 2.856156587600708,
+      "learning_rate": 1.1162272098983183e-05,
+      "loss": 0.9004,
+      "step": 2338
+    },
+    {
+      "epoch": 0.47930327868852457,
+      "grad_norm": 6.994837284088135,
+      "learning_rate": 1.115567919397709e-05,
+      "loss": 1.0769,
+      "step": 2339
+    },
+    {
+      "epoch": 0.47950819672131145,
+      "grad_norm": 4.382102966308594,
+      "learning_rate": 1.1149085779799589e-05,
+      "loss": 1.2367,
+      "step": 2340
+    },
+    {
+      "epoch": 0.47971311475409834,
+      "grad_norm": 5.024289608001709,
+      "learning_rate": 1.1142491859355607e-05,
+      "loss": 1.3308,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4799180327868853,
+      "grad_norm": 3.8781700134277344,
+      "learning_rate": 1.1135897435550313e-05,
+      "loss": 0.9014,
+      "step": 2342
+    },
+    {
+      "epoch": 0.48012295081967216,
+      "grad_norm": 6.580192565917969,
+      "learning_rate": 1.1129302511289091e-05,
+      "loss": 1.1041,
+      "step": 2343
+    },
+    {
+      "epoch": 0.48032786885245904,
+      "grad_norm": 4.785983085632324,
+      "learning_rate": 1.1122707089477539e-05,
+      "loss": 1.0238,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4805327868852459,
+      "grad_norm": 3.0041873455047607,
+      "learning_rate": 1.111611117302149e-05,
+      "loss": 0.9421,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4807377049180328,
+      "grad_norm": 3.4305331707000732,
+      "learning_rate": 1.1109514764826976e-05,
+      "loss": 0.6829,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4809426229508197,
+      "grad_norm": 3.4143149852752686,
+      "learning_rate": 1.1102917867800267e-05,
+      "loss": 1.1524,
+      "step": 2347
+    },
+    {
+      "epoch": 0.48114754098360657,
+      "grad_norm": 3.4402682781219482,
+      "learning_rate": 1.109632048484783e-05,
+      "loss": 0.8992,
+      "step": 2348
+    },
+    {
+      "epoch": 0.48135245901639345,
+      "grad_norm": 2.9864485263824463,
+      "learning_rate": 1.1089722618876351e-05,
+      "loss": 0.9759,
+      "step": 2349
+    },
+    {
+      "epoch": 0.48155737704918034,
+      "grad_norm": 5.168068885803223,
+      "learning_rate": 1.1083124272792739e-05,
+      "loss": 0.6819,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4817622950819672,
+      "grad_norm": 4.26582670211792,
+      "learning_rate": 1.1076525449504095e-05,
+      "loss": 1.1196,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4819672131147541,
+      "grad_norm": 5.729946613311768,
+      "learning_rate": 1.1069926151917751e-05,
+      "loss": 1.1011,
+      "step": 2352
+    },
+    {
+      "epoch": 0.482172131147541,
+      "grad_norm": 5.019975662231445,
+      "learning_rate": 1.1063326382941235e-05,
+      "loss": 1.2283,
+      "step": 2353
+    },
+    {
+      "epoch": 0.48237704918032787,
+      "grad_norm": 8.698945999145508,
+      "learning_rate": 1.1056726145482286e-05,
+      "loss": 1.1935,
+      "step": 2354
+    },
+    {
+      "epoch": 0.48258196721311475,
+      "grad_norm": 4.521785259246826,
+      "learning_rate": 1.1050125442448849e-05,
+      "loss": 0.9627,
+      "step": 2355
+    },
+    {
+      "epoch": 0.48278688524590163,
+      "grad_norm": 4.523580551147461,
+      "learning_rate": 1.1043524276749078e-05,
+      "loss": 1.1396,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4829918032786885,
+      "grad_norm": 3.8483264446258545,
+      "learning_rate": 1.1036922651291327e-05,
+      "loss": 1.0745,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4831967213114754,
+      "grad_norm": 6.339890480041504,
+      "learning_rate": 1.1030320568984148e-05,
+      "loss": 1.1111,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4834016393442623,
+      "grad_norm": 4.482868671417236,
+      "learning_rate": 1.10237180327363e-05,
+      "loss": 1.2819,
+      "step": 2359
+    },
+    {
+      "epoch": 0.48360655737704916,
+      "grad_norm": 5.522195816040039,
+      "learning_rate": 1.101711504545675e-05,
+      "loss": 0.9917,
+      "step": 2360
+    },
+    {
+      "epoch": 0.48381147540983604,
+      "grad_norm": 2.901610851287842,
+      "learning_rate": 1.1010511610054649e-05,
+      "loss": 0.8886,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4840163934426229,
+      "grad_norm": 6.219361782073975,
+      "learning_rate": 1.1003907729439349e-05,
+      "loss": 1.1041,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4842213114754098,
+      "grad_norm": 4.471076488494873,
+      "learning_rate": 1.0997303406520407e-05,
+      "loss": 1.0524,
+      "step": 2363
+    },
+    {
+      "epoch": 0.48442622950819675,
+      "grad_norm": 5.008403778076172,
+      "learning_rate": 1.0990698644207565e-05,
+      "loss": 1.2627,
+      "step": 2364
+    },
+    {
+      "epoch": 0.48463114754098363,
+      "grad_norm": 7.256374835968018,
+      "learning_rate": 1.0984093445410759e-05,
+      "loss": 0.8888,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4848360655737705,
+      "grad_norm": 4.349891662597656,
+      "learning_rate": 1.0977487813040128e-05,
+      "loss": 1.1349,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4850409836065574,
+      "grad_norm": 3.4002795219421387,
+      "learning_rate": 1.0970881750005989e-05,
+      "loss": 1.2222,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4852459016393443,
+      "grad_norm": 3.3179574012756348,
+      "learning_rate": 1.096427525921886e-05,
+      "loss": 1.06,
+      "step": 2368
+    },
+    {
+      "epoch": 0.48545081967213116,
+      "grad_norm": 5.733511924743652,
+      "learning_rate": 1.095766834358943e-05,
+      "loss": 0.7141,
+      "step": 2369
+    },
+    {
+      "epoch": 0.48565573770491804,
+      "grad_norm": 2.6865181922912598,
+      "learning_rate": 1.09510610060286e-05,
+      "loss": 1.1443,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4858606557377049,
+      "grad_norm": 7.74459981918335,
+      "learning_rate": 1.0944453249447437e-05,
+      "loss": 0.7953,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4860655737704918,
+      "grad_norm": 2.411824941635132,
+      "learning_rate": 1.0937845076757201e-05,
+      "loss": 1.1153,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4862704918032787,
+      "grad_norm": 2.4925332069396973,
+      "learning_rate": 1.0931236490869333e-05,
+      "loss": 0.9997,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4864754098360656,
+      "grad_norm": 6.0027055740356445,
+      "learning_rate": 1.0924627494695457e-05,
+      "loss": 1.3898,
+      "step": 2374
+    },
+    {
+      "epoch": 0.48668032786885246,
+      "grad_norm": 3.7718961238861084,
+      "learning_rate": 1.091801809114738e-05,
+      "loss": 1.3434,
+      "step": 2375
+    },
+    {
+      "epoch": 0.48688524590163934,
+      "grad_norm": 3.515932083129883,
+      "learning_rate": 1.0911408283137079e-05,
+      "loss": 1.0102,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4870901639344262,
+      "grad_norm": 4.427034378051758,
+      "learning_rate": 1.0904798073576728e-05,
+      "loss": 0.787,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4872950819672131,
+      "grad_norm": 9.80677604675293,
+      "learning_rate": 1.0898187465378654e-05,
+      "loss": 1.3969,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 3.123809814453125,
+      "learning_rate": 1.0891576461455378e-05,
+      "loss": 0.6517,
+      "step": 2379
+    },
+    {
+      "epoch": 0.48770491803278687,
+      "grad_norm": 3.0228757858276367,
+      "learning_rate": 1.0884965064719593e-05,
+      "loss": 0.9452,
+      "step": 2380
+    },
+    {
+      "epoch": 0.48790983606557375,
+      "grad_norm": 5.6180033683776855,
+      "learning_rate": 1.0878353278084148e-05,
+      "loss": 0.8572,
+      "step": 2381
+    },
+    {
+      "epoch": 0.48811475409836064,
+      "grad_norm": 2.8955650329589844,
+      "learning_rate": 1.0871741104462092e-05,
+      "loss": 1.1519,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4883196721311475,
+      "grad_norm": 3.440260171890259,
+      "learning_rate": 1.0865128546766619e-05,
+      "loss": 0.8338,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4885245901639344,
+      "grad_norm": 3.9037415981292725,
+      "learning_rate": 1.0858515607911108e-05,
+      "loss": 1.1004,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4887295081967213,
+      "grad_norm": 6.166625499725342,
+      "learning_rate": 1.0851902290809102e-05,
+      "loss": 1.1542,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4889344262295082,
+      "grad_norm": 2.8911218643188477,
+      "learning_rate": 1.0845288598374301e-05,
+      "loss": 0.9657,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4891393442622951,
+      "grad_norm": 6.561974048614502,
+      "learning_rate": 1.0838674533520587e-05,
+      "loss": 0.8828,
+      "step": 2387
+    },
+    {
+      "epoch": 0.489344262295082,
+      "grad_norm": 2.3017570972442627,
+      "learning_rate": 1.0832060099161996e-05,
+      "loss": 0.9242,
+      "step": 2388
+    },
+    {
+      "epoch": 0.48954918032786887,
+      "grad_norm": 2.697690010070801,
+      "learning_rate": 1.0825445298212724e-05,
+      "loss": 1.0064,
+      "step": 2389
+    },
+    {
+      "epoch": 0.48975409836065575,
+      "grad_norm": 3.788198709487915,
+      "learning_rate": 1.0818830133587139e-05,
+      "loss": 0.9873,
+      "step": 2390
+    },
+    {
+      "epoch": 0.48995901639344264,
+      "grad_norm": 4.118875980377197,
+      "learning_rate": 1.0812214608199761e-05,
+      "loss": 1.2066,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4901639344262295,
+      "grad_norm": 3.4462180137634277,
+      "learning_rate": 1.0805598724965269e-05,
+      "loss": 0.8756,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4903688524590164,
+      "grad_norm": 6.2484893798828125,
+      "learning_rate": 1.0798982486798505e-05,
+      "loss": 0.8132,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4905737704918033,
+      "grad_norm": 3.4035940170288086,
+      "learning_rate": 1.0792365896614462e-05,
+      "loss": 1.0806,
+      "step": 2394
+    },
+    {
+      "epoch": 0.49077868852459017,
+      "grad_norm": 4.293105602264404,
+      "learning_rate": 1.0785748957328293e-05,
+      "loss": 0.7655,
+      "step": 2395
+    },
+    {
+      "epoch": 0.49098360655737705,
+      "grad_norm": 12.465897560119629,
+      "learning_rate": 1.0779131671855297e-05,
+      "loss": 0.8361,
+      "step": 2396
+    },
+    {
+      "epoch": 0.49118852459016393,
+      "grad_norm": 7.889960765838623,
+      "learning_rate": 1.0772514043110935e-05,
+      "loss": 1.1865,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4913934426229508,
+      "grad_norm": 3.1491503715515137,
+      "learning_rate": 1.0765896074010815e-05,
+      "loss": 1.0211,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4915983606557377,
+      "grad_norm": 2.2576634883880615,
+      "learning_rate": 1.0759277767470698e-05,
+      "loss": 1.0068,
+      "step": 2399
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 5.158161163330078,
+      "learning_rate": 1.0752659126406482e-05,
+      "loss": 1.1495,
+      "step": 2400
+    },
+    {
+      "epoch": 0.49200819672131146,
+      "grad_norm": 5.836902141571045,
+      "learning_rate": 1.0746040153734225e-05,
+      "loss": 1.2325,
+      "step": 2401
+    },
+    {
+      "epoch": 0.49221311475409835,
+      "grad_norm": 4.041435241699219,
+      "learning_rate": 1.0739420852370128e-05,
+      "loss": 0.9929,
+      "step": 2402
+    },
+    {
+      "epoch": 0.49241803278688523,
+      "grad_norm": 2.1019387245178223,
+      "learning_rate": 1.0732801225230534e-05,
+      "loss": 0.8788,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4926229508196721,
+      "grad_norm": 6.333626747131348,
+      "learning_rate": 1.0726181275231935e-05,
+      "loss": 1.1977,
+      "step": 2404
+    },
+    {
+      "epoch": 0.492827868852459,
+      "grad_norm": 5.542392730712891,
+      "learning_rate": 1.0719561005290959e-05,
+      "loss": 0.904,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4930327868852459,
+      "grad_norm": 4.672683238983154,
+      "learning_rate": 1.0712940418324375e-05,
+      "loss": 1.1501,
+      "step": 2406
+    },
+    {
+      "epoch": 0.49323770491803276,
+      "grad_norm": 3.9679369926452637,
+      "learning_rate": 1.07063195172491e-05,
+      "loss": 0.7809,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4934426229508197,
+      "grad_norm": 3.4090864658355713,
+      "learning_rate": 1.0699698304982178e-05,
+      "loss": 1.0075,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4936475409836066,
+      "grad_norm": 4.4592084884643555,
+      "learning_rate": 1.06930767844408e-05,
+      "loss": 1.2361,
+      "step": 2409
+    },
+    {
+      "epoch": 0.49385245901639346,
+      "grad_norm": 3.7311770915985107,
+      "learning_rate": 1.068645495854229e-05,
+      "loss": 1.3177,
+      "step": 2410
+    },
+    {
+      "epoch": 0.49405737704918035,
+      "grad_norm": 3.5681397914886475,
+      "learning_rate": 1.0679832830204095e-05,
+      "loss": 1.0006,
+      "step": 2411
+    },
+    {
+      "epoch": 0.49426229508196723,
+      "grad_norm": 3.3702051639556885,
+      "learning_rate": 1.0673210402343818e-05,
+      "loss": 1.0942,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4944672131147541,
+      "grad_norm": 3.7243518829345703,
+      "learning_rate": 1.0666587677879175e-05,
+      "loss": 0.8957,
+      "step": 2413
+    },
+    {
+      "epoch": 0.494672131147541,
+      "grad_norm": 4.6438117027282715,
+      "learning_rate": 1.0659964659728017e-05,
+      "loss": 0.9752,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4948770491803279,
+      "grad_norm": 3.5301878452301025,
+      "learning_rate": 1.0653341350808333e-05,
+      "loss": 1.2681,
+      "step": 2415
+    },
+    {
+      "epoch": 0.49508196721311476,
+      "grad_norm": 7.681845664978027,
+      "learning_rate": 1.0646717754038226e-05,
+      "loss": 1.2466,
+      "step": 2416
+    },
+    {
+      "epoch": 0.49528688524590164,
+      "grad_norm": 3.539548635482788,
+      "learning_rate": 1.0640093872335936e-05,
+      "loss": 0.9116,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4954918032786885,
+      "grad_norm": 4.802485942840576,
+      "learning_rate": 1.063346970861983e-05,
+      "loss": 0.8781,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4956967213114754,
+      "grad_norm": 2.8960678577423096,
+      "learning_rate": 1.062684526580839e-05,
+      "loss": 0.9539,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4959016393442623,
+      "grad_norm": 4.327597141265869,
+      "learning_rate": 1.062022054682023e-05,
+      "loss": 1.1996,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4961065573770492,
+      "grad_norm": 5.7696332931518555,
+      "learning_rate": 1.0613595554574076e-05,
+      "loss": 0.964,
+      "step": 2421
+    },
+    {
+      "epoch": 0.49631147540983606,
+      "grad_norm": 5.889580249786377,
+      "learning_rate": 1.0606970291988787e-05,
+      "loss": 0.9185,
+      "step": 2422
+    },
+    {
+      "epoch": 0.49651639344262294,
+      "grad_norm": 4.208503723144531,
+      "learning_rate": 1.0600344761983327e-05,
+      "loss": 1.0664,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4967213114754098,
+      "grad_norm": 3.5910866260528564,
+      "learning_rate": 1.0593718967476796e-05,
+      "loss": 1.018,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4969262295081967,
+      "grad_norm": 2.5141851902008057,
+      "learning_rate": 1.0587092911388388e-05,
+      "loss": 0.8483,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4971311475409836,
+      "grad_norm": 5.177306652069092,
+      "learning_rate": 1.0580466596637428e-05,
+      "loss": 0.8697,
+      "step": 2426
+    },
+    {
+      "epoch": 0.49733606557377047,
+      "grad_norm": 3.3987159729003906,
+      "learning_rate": 1.0573840026143353e-05,
+      "loss": 0.9197,
+      "step": 2427
+    },
+    {
+      "epoch": 0.49754098360655735,
+      "grad_norm": 4.8787841796875,
+      "learning_rate": 1.0567213202825709e-05,
+      "loss": 1.0261,
+      "step": 2428
+    },
+    {
+      "epoch": 0.49774590163934423,
+      "grad_norm": 2.764641284942627,
+      "learning_rate": 1.0560586129604154e-05,
+      "loss": 0.8275,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4979508196721312,
+      "grad_norm": 6.355806827545166,
+      "learning_rate": 1.0553958809398456e-05,
+      "loss": 1.1474,
+      "step": 2430
+    },
+    {
+      "epoch": 0.49815573770491806,
+      "grad_norm": 3.9715490341186523,
+      "learning_rate": 1.0547331245128495e-05,
+      "loss": 1.063,
+      "step": 2431
+    },
+    {
+      "epoch": 0.49836065573770494,
+      "grad_norm": 4.181251049041748,
+      "learning_rate": 1.054070343971426e-05,
+      "loss": 1.1862,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4985655737704918,
+      "grad_norm": 2.862741470336914,
+      "learning_rate": 1.0534075396075833e-05,
+      "loss": 0.77,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4987704918032787,
+      "grad_norm": 2.677647113800049,
+      "learning_rate": 1.052744711713342e-05,
+      "loss": 1.0361,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4989754098360656,
+      "grad_norm": 5.854428768157959,
+      "learning_rate": 1.0520818605807315e-05,
+      "loss": 1.0014,
+      "step": 2435
+    },
+    {
+      "epoch": 0.49918032786885247,
+      "grad_norm": 3.8384740352630615,
+      "learning_rate": 1.0514189865017922e-05,
+      "loss": 1.002,
+      "step": 2436
+    },
+    {
+      "epoch": 0.49938524590163935,
+      "grad_norm": 4.62101411819458,
+      "learning_rate": 1.0507560897685744e-05,
+      "loss": 0.9048,
+      "step": 2437
+    },
+    {
+      "epoch": 0.49959016393442623,
+      "grad_norm": 3.224613666534424,
+      "learning_rate": 1.050093170673139e-05,
+      "loss": 1.0623,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4997950819672131,
+      "grad_norm": 7.602325439453125,
+      "learning_rate": 1.0494302295075553e-05,
+      "loss": 1.1732,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.496480464935303,
+      "learning_rate": 1.0487672665639042e-05,
+      "loss": 1.0007,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5002049180327869,
+      "grad_norm": 3.7961368560791016,
+      "learning_rate": 1.0481042821342745e-05,
+      "loss": 0.9644,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5004098360655738,
+      "grad_norm": 3.182710886001587,
+      "learning_rate": 1.0474412765107653e-05,
+      "loss": 0.9374,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5006147540983606,
+      "grad_norm": 4.228931427001953,
+      "learning_rate": 1.0467782499854851e-05,
+      "loss": 0.746,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5008196721311475,
+      "grad_norm": 3.2326526641845703,
+      "learning_rate": 1.0461152028505514e-05,
+      "loss": 0.6949,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5010245901639344,
+      "grad_norm": 4.789170742034912,
+      "learning_rate": 1.0454521353980908e-05,
+      "loss": 1.0035,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5012295081967213,
+      "grad_norm": 2.8456459045410156,
+      "learning_rate": 1.0447890479202388e-05,
+      "loss": 0.7245,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5014344262295082,
+      "grad_norm": 4.047671794891357,
+      "learning_rate": 1.04412594070914e-05,
+      "loss": 1.3114,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5016393442622951,
+      "grad_norm": 2.9569056034088135,
+      "learning_rate": 1.0434628140569471e-05,
+      "loss": 0.804,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5018442622950819,
+      "grad_norm": 2.9426167011260986,
+      "learning_rate": 1.0427996682558219e-05,
+      "loss": 0.9218,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5020491803278688,
+      "grad_norm": 3.6683387756347656,
+      "learning_rate": 1.0421365035979346e-05,
+      "loss": 1.0705,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5022540983606557,
+      "grad_norm": 7.150694370269775,
+      "learning_rate": 1.0414733203754637e-05,
+      "loss": 0.9638,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5024590163934426,
+      "grad_norm": 7.3196516036987305,
+      "learning_rate": 1.0408101188805957e-05,
+      "loss": 1.271,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5026639344262295,
+      "grad_norm": 3.1651365756988525,
+      "learning_rate": 1.0401468994055245e-05,
+      "loss": 0.882,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5028688524590164,
+      "grad_norm": 4.654633522033691,
+      "learning_rate": 1.039483662242454e-05,
+      "loss": 1.2377,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5030737704918032,
+      "grad_norm": 6.739443778991699,
+      "learning_rate": 1.0388204076835935e-05,
+      "loss": 1.0244,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5032786885245901,
+      "grad_norm": 3.005746364593506,
+      "learning_rate": 1.0381571360211613e-05,
+      "loss": 0.9105,
+      "step": 2456
+    },
+    {
+      "epoch": 0.503483606557377,
+      "grad_norm": 2.8768234252929688,
+      "learning_rate": 1.0374938475473832e-05,
+      "loss": 0.9826,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5036885245901639,
+      "grad_norm": 4.732824325561523,
+      "learning_rate": 1.0368305425544913e-05,
+      "loss": 1.0635,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5038934426229508,
+      "grad_norm": 8.31392765045166,
+      "learning_rate": 1.0361672213347268e-05,
+      "loss": 0.9908,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5040983606557377,
+      "grad_norm": 7.147120475769043,
+      "learning_rate": 1.0355038841803366e-05,
+      "loss": 1.3845,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5043032786885245,
+      "grad_norm": 3.0534536838531494,
+      "learning_rate": 1.0348405313835752e-05,
+      "loss": 1.1484,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5045081967213115,
+      "grad_norm": 2.181628465652466,
+      "learning_rate": 1.0341771632367042e-05,
+      "loss": 1.0392,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5047131147540984,
+      "grad_norm": 3.6983518600463867,
+      "learning_rate": 1.0335137800319905e-05,
+      "loss": 1.2573,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5049180327868853,
+      "grad_norm": 4.796201229095459,
+      "learning_rate": 1.0328503820617101e-05,
+      "loss": 0.947,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5051229508196722,
+      "grad_norm": 2.7933263778686523,
+      "learning_rate": 1.0321869696181437e-05,
+      "loss": 0.8864,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5053278688524591,
+      "grad_norm": 2.9933557510375977,
+      "learning_rate": 1.0315235429935786e-05,
+      "loss": 0.9238,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5055327868852459,
+      "grad_norm": 11.998370170593262,
+      "learning_rate": 1.0308601024803089e-05,
+      "loss": 1.0497,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5057377049180328,
+      "grad_norm": 7.05670166015625,
+      "learning_rate": 1.0301966483706343e-05,
+      "loss": 1.2548,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5059426229508197,
+      "grad_norm": 2.989391326904297,
+      "learning_rate": 1.0295331809568612e-05,
+      "loss": 0.7959,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5061475409836066,
+      "grad_norm": 5.948931694030762,
+      "learning_rate": 1.028869700531301e-05,
+      "loss": 0.8807,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5063524590163935,
+      "grad_norm": 11.44007396697998,
+      "learning_rate": 1.028206207386271e-05,
+      "loss": 1.2476,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5065573770491804,
+      "grad_norm": 3.285367250442505,
+      "learning_rate": 1.0275427018140946e-05,
+      "loss": 0.7795,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5067622950819672,
+      "grad_norm": 6.1609368324279785,
+      "learning_rate": 1.0268791841071004e-05,
+      "loss": 0.8442,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5069672131147541,
+      "grad_norm": 9.504800796508789,
+      "learning_rate": 1.0262156545576225e-05,
+      "loss": 1.1427,
+      "step": 2474
+    },
+    {
+      "epoch": 0.507172131147541,
+      "grad_norm": 4.080223560333252,
+      "learning_rate": 1.0255521134579997e-05,
+      "loss": 1.177,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5073770491803279,
+      "grad_norm": 5.539357662200928,
+      "learning_rate": 1.0248885611005768e-05,
+      "loss": 1.1837,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5075819672131148,
+      "grad_norm": 4.1226325035095215,
+      "learning_rate": 1.0242249977777022e-05,
+      "loss": 1.0269,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5077868852459017,
+      "grad_norm": 4.166797161102295,
+      "learning_rate": 1.0235614237817304e-05,
+      "loss": 0.9095,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5079918032786885,
+      "grad_norm": 5.738528251647949,
+      "learning_rate": 1.0228978394050201e-05,
+      "loss": 1.1002,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5081967213114754,
+      "grad_norm": 3.56195068359375,
+      "learning_rate": 1.0222342449399348e-05,
+      "loss": 0.9097,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5084016393442623,
+      "grad_norm": 3.2952075004577637,
+      "learning_rate": 1.0215706406788417e-05,
+      "loss": 1.1076,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5086065573770492,
+      "grad_norm": 5.990411281585693,
+      "learning_rate": 1.0209070269141138e-05,
+      "loss": 1.1013,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5088114754098361,
+      "grad_norm": 3.0738320350646973,
+      "learning_rate": 1.0202434039381264e-05,
+      "loss": 1.0183,
+      "step": 2483
+    },
+    {
+      "epoch": 0.509016393442623,
+      "grad_norm": 4.458913803100586,
+      "learning_rate": 1.0195797720432601e-05,
+      "loss": 1.0097,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5092213114754098,
+      "grad_norm": 3.030350923538208,
+      "learning_rate": 1.0189161315218994e-05,
+      "loss": 0.8472,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5094262295081967,
+      "grad_norm": 4.270596504211426,
+      "learning_rate": 1.0182524826664322e-05,
+      "loss": 1.0027,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5096311475409836,
+      "grad_norm": 3.7733209133148193,
+      "learning_rate": 1.0175888257692506e-05,
+      "loss": 0.9917,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5098360655737705,
+      "grad_norm": 4.7685980796813965,
+      "learning_rate": 1.016925161122749e-05,
+      "loss": 1.2032,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5100409836065574,
+      "grad_norm": 5.626981258392334,
+      "learning_rate": 1.0162614890193267e-05,
+      "loss": 0.8712,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5102459016393442,
+      "grad_norm": 3.9570000171661377,
+      "learning_rate": 1.0155978097513854e-05,
+      "loss": 0.9241,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5104508196721311,
+      "grad_norm": 4.74062442779541,
+      "learning_rate": 1.0149341236113306e-05,
+      "loss": 1.1416,
+      "step": 2491
+    },
+    {
+      "epoch": 0.510655737704918,
+      "grad_norm": 4.7518134117126465,
+      "learning_rate": 1.0142704308915702e-05,
+      "loss": 1.0356,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5108606557377049,
+      "grad_norm": 3.5203065872192383,
+      "learning_rate": 1.013606731884515e-05,
+      "loss": 1.0271,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5110655737704918,
+      "grad_norm": 4.401505947113037,
+      "learning_rate": 1.0129430268825795e-05,
+      "loss": 0.8811,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5112704918032787,
+      "grad_norm": 4.243074893951416,
+      "learning_rate": 1.0122793161781792e-05,
+      "loss": 0.9964,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5114754098360655,
+      "grad_norm": 3.3323190212249756,
+      "learning_rate": 1.0116156000637338e-05,
+      "loss": 1.0832,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5116803278688524,
+      "grad_norm": 7.5317702293396,
+      "learning_rate": 1.0109518788316648e-05,
+      "loss": 1.0025,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5118852459016393,
+      "grad_norm": 5.229051113128662,
+      "learning_rate": 1.0102881527743948e-05,
+      "loss": 1.2698,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5120901639344262,
+      "grad_norm": 5.160327911376953,
+      "learning_rate": 1.0096244221843508e-05,
+      "loss": 0.8824,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5122950819672131,
+      "grad_norm": 5.313085079193115,
+      "learning_rate": 1.0089606873539592e-05,
+      "loss": 1.2214,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 3.4995884895324707,
+      "learning_rate": 1.0082969485756509e-05,
+      "loss": 0.9132,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5127049180327868,
+      "grad_norm": 5.700722694396973,
+      "learning_rate": 1.0076332061418565e-05,
+      "loss": 1.1802,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5129098360655737,
+      "grad_norm": 3.528224468231201,
+      "learning_rate": 1.0069694603450092e-05,
+      "loss": 1.3116,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5131147540983606,
+      "grad_norm": 3.2894198894500732,
+      "learning_rate": 1.0063057114775432e-05,
+      "loss": 0.7747,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5133196721311475,
+      "grad_norm": 4.4766764640808105,
+      "learning_rate": 1.005641959831894e-05,
+      "loss": 0.95,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5135245901639345,
+      "grad_norm": 5.803834915161133,
+      "learning_rate": 1.0049782057004998e-05,
+      "loss": 1.2217,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5137295081967214,
+      "grad_norm": 2.8112356662750244,
+      "learning_rate": 1.0043144493757974e-05,
+      "loss": 0.8897,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5139344262295082,
+      "grad_norm": 6.333378791809082,
+      "learning_rate": 1.0036506911502266e-05,
+      "loss": 1.0309,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5141393442622951,
+      "grad_norm": 3.1242051124572754,
+      "learning_rate": 1.002986931316227e-05,
+      "loss": 0.7774,
+      "step": 2509
+    },
+    {
+      "epoch": 0.514344262295082,
+      "grad_norm": 2.8687520027160645,
+      "learning_rate": 1.0023231701662396e-05,
+      "loss": 1.0331,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5145491803278689,
+      "grad_norm": 4.471157073974609,
+      "learning_rate": 1.001659407992705e-05,
+      "loss": 1.098,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5147540983606558,
+      "grad_norm": 5.07621955871582,
+      "learning_rate": 1.0009956450880654e-05,
+      "loss": 1.1249,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5149590163934427,
+      "grad_norm": 4.230279445648193,
+      "learning_rate": 1.0003318817447622e-05,
+      "loss": 1.0192,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5151639344262295,
+      "grad_norm": 5.937757968902588,
+      "learning_rate": 9.99668118255238e-06,
+      "loss": 1.0874,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5153688524590164,
+      "grad_norm": 5.508362770080566,
+      "learning_rate": 9.99004354911935e-06,
+      "loss": 0.983,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5155737704918033,
+      "grad_norm": 2.7983500957489014,
+      "learning_rate": 9.983405920072954e-06,
+      "loss": 1.1725,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5157786885245902,
+      "grad_norm": 4.259829521179199,
+      "learning_rate": 9.97676829833761e-06,
+      "loss": 0.939,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5159836065573771,
+      "grad_norm": 3.0207040309906006,
+      "learning_rate": 9.97013068683773e-06,
+      "loss": 0.9543,
+      "step": 2518
+    },
+    {
+      "epoch": 0.516188524590164,
+      "grad_norm": 5.187236309051514,
+      "learning_rate": 9.96349308849774e-06,
+      "loss": 1.2305,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5163934426229508,
+      "grad_norm": 4.47042989730835,
+      "learning_rate": 9.956855506242027e-06,
+      "loss": 1.2339,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5165983606557377,
+      "grad_norm": 8.764968872070312,
+      "learning_rate": 9.950217942995005e-06,
+      "loss": 1.0629,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5168032786885246,
+      "grad_norm": 6.083300590515137,
+      "learning_rate": 9.943580401681058e-06,
+      "loss": 1.1874,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5170081967213115,
+      "grad_norm": 2.7794864177703857,
+      "learning_rate": 9.936942885224573e-06,
+      "loss": 0.8903,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5172131147540984,
+      "grad_norm": 3.4040274620056152,
+      "learning_rate": 9.930305396549913e-06,
+      "loss": 0.8636,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5174180327868853,
+      "grad_norm": 7.108668804168701,
+      "learning_rate": 9.923667938581436e-06,
+      "loss": 1.3315,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5176229508196721,
+      "grad_norm": 3.560728073120117,
+      "learning_rate": 9.917030514243494e-06,
+      "loss": 0.8382,
+      "step": 2526
+    },
+    {
+      "epoch": 0.517827868852459,
+      "grad_norm": 4.401824474334717,
+      "learning_rate": 9.91039312646041e-06,
+      "loss": 1.0086,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5180327868852459,
+      "grad_norm": 5.1824951171875,
+      "learning_rate": 9.903755778156496e-06,
+      "loss": 0.8715,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5182377049180328,
+      "grad_norm": 3.232264518737793,
+      "learning_rate": 9.897118472256054e-06,
+      "loss": 0.9704,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5184426229508197,
+      "grad_norm": 4.460423946380615,
+      "learning_rate": 9.890481211683358e-06,
+      "loss": 0.9671,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5186475409836065,
+      "grad_norm": 4.8398027420043945,
+      "learning_rate": 9.883843999362664e-06,
+      "loss": 1.1672,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5188524590163934,
+      "grad_norm": 2.915750503540039,
+      "learning_rate": 9.87720683821821e-06,
+      "loss": 0.9902,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5190573770491803,
+      "grad_norm": 4.7178754806518555,
+      "learning_rate": 9.87056973117421e-06,
+      "loss": 1.1148,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5192622950819672,
+      "grad_norm": 3.6348676681518555,
+      "learning_rate": 9.863932681154853e-06,
+      "loss": 0.8186,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5194672131147541,
+      "grad_norm": 8.097248077392578,
+      "learning_rate": 9.8572956910843e-06,
+      "loss": 1.0651,
+      "step": 2535
+    },
+    {
+      "epoch": 0.519672131147541,
+      "grad_norm": 1.9907177686691284,
+      "learning_rate": 9.850658763886694e-06,
+      "loss": 0.6591,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5198770491803278,
+      "grad_norm": 8.26021957397461,
+      "learning_rate": 9.844021902486148e-06,
+      "loss": 1.1723,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5200819672131147,
+      "grad_norm": 4.648985862731934,
+      "learning_rate": 9.837385109806736e-06,
+      "loss": 1.1288,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5202868852459016,
+      "grad_norm": 2.5668814182281494,
+      "learning_rate": 9.830748388772512e-06,
+      "loss": 0.9256,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5204918032786885,
+      "grad_norm": 4.8465189933776855,
+      "learning_rate": 9.824111742307501e-06,
+      "loss": 0.992,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5206967213114754,
+      "grad_norm": 4.210499286651611,
+      "learning_rate": 9.817475173335681e-06,
+      "loss": 1.2377,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5209016393442623,
+      "grad_norm": 2.9954652786254883,
+      "learning_rate": 9.81083868478101e-06,
+      "loss": 1.0993,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5211065573770491,
+      "grad_norm": 4.889657974243164,
+      "learning_rate": 9.804202279567399e-06,
+      "loss": 0.872,
+      "step": 2543
+    },
+    {
+      "epoch": 0.521311475409836,
+      "grad_norm": 3.310940980911255,
+      "learning_rate": 9.797565960618743e-06,
+      "loss": 1.1348,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5215163934426229,
+      "grad_norm": 5.866358757019043,
+      "learning_rate": 9.790929730858866e-06,
+      "loss": 0.9576,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5217213114754098,
+      "grad_norm": 4.514755725860596,
+      "learning_rate": 9.784293593211584e-06,
+      "loss": 1.2222,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5219262295081967,
+      "grad_norm": 4.675336837768555,
+      "learning_rate": 9.777657550600657e-06,
+      "loss": 0.698,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5221311475409836,
+      "grad_norm": 5.438223838806152,
+      "learning_rate": 9.771021605949802e-06,
+      "loss": 1.1693,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5223360655737705,
+      "grad_norm": 2.9111216068267822,
+      "learning_rate": 9.764385762182699e-06,
+      "loss": 0.946,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5225409836065574,
+      "grad_norm": 5.630061149597168,
+      "learning_rate": 9.75775002222298e-06,
+      "loss": 0.9822,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5227459016393443,
+      "grad_norm": 3.4655210971832275,
+      "learning_rate": 9.751114388994239e-06,
+      "loss": 0.7771,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5229508196721312,
+      "grad_norm": 3.1668860912323,
+      "learning_rate": 9.744478865420006e-06,
+      "loss": 1.1336,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5231557377049181,
+      "grad_norm": 6.685107707977295,
+      "learning_rate": 9.737843454423775e-06,
+      "loss": 0.8615,
+      "step": 2553
+    },
+    {
+      "epoch": 0.523360655737705,
+      "grad_norm": 5.305113792419434,
+      "learning_rate": 9.731208158929e-06,
+      "loss": 0.9194,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5235655737704918,
+      "grad_norm": 4.817692279815674,
+      "learning_rate": 9.724572981859056e-06,
+      "loss": 0.7259,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5237704918032787,
+      "grad_norm": 6.927541732788086,
+      "learning_rate": 9.717937926137293e-06,
+      "loss": 1.0333,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5239754098360656,
+      "grad_norm": 5.360337257385254,
+      "learning_rate": 9.711302994686994e-06,
+      "loss": 1.0036,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5241803278688525,
+      "grad_norm": 2.8377466201782227,
+      "learning_rate": 9.704668190431392e-06,
+      "loss": 0.9749,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5243852459016394,
+      "grad_norm": 3.6724905967712402,
+      "learning_rate": 9.698033516293659e-06,
+      "loss": 0.7425,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 3.9380078315734863,
+      "learning_rate": 9.691398975196913e-06,
+      "loss": 1.0131,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5247950819672131,
+      "grad_norm": 9.484538078308105,
+      "learning_rate": 9.68476457006422e-06,
+      "loss": 1.2069,
+      "step": 2561
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 5.371415615081787,
+      "learning_rate": 9.678130303818565e-06,
+      "loss": 0.9656,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5252049180327869,
+      "grad_norm": 10.304100036621094,
+      "learning_rate": 9.6714961793829e-06,
+      "loss": 0.7947,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5254098360655738,
+      "grad_norm": 4.354328155517578,
+      "learning_rate": 9.664862199680095e-06,
+      "loss": 0.9775,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5256147540983607,
+      "grad_norm": 5.065672874450684,
+      "learning_rate": 9.658228367632963e-06,
+      "loss": 1.0054,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5258196721311476,
+      "grad_norm": 5.126008987426758,
+      "learning_rate": 9.65159468616425e-06,
+      "loss": 1.2335,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5260245901639344,
+      "grad_norm": 3.9491159915924072,
+      "learning_rate": 9.644961158196635e-06,
+      "loss": 0.9864,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5262295081967213,
+      "grad_norm": 3.626469612121582,
+      "learning_rate": 9.638327786652737e-06,
+      "loss": 1.2007,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5264344262295082,
+      "grad_norm": 4.0126953125,
+      "learning_rate": 9.63169457445509e-06,
+      "loss": 0.9709,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5266393442622951,
+      "grad_norm": 4.082720756530762,
+      "learning_rate": 9.625061524526171e-06,
+      "loss": 1.0379,
+      "step": 2570
+    },
+    {
+      "epoch": 0.526844262295082,
+      "grad_norm": 9.4889554977417,
+      "learning_rate": 9.618428639788387e-06,
+      "loss": 1.1266,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5270491803278688,
+      "grad_norm": 3.9822165966033936,
+      "learning_rate": 9.611795923164068e-06,
+      "loss": 0.8413,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5272540983606557,
+      "grad_norm": 3.2945773601531982,
+      "learning_rate": 9.605163377575463e-06,
+      "loss": 0.8493,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5274590163934426,
+      "grad_norm": 3.950878858566284,
+      "learning_rate": 9.598531005944755e-06,
+      "loss": 0.9443,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5276639344262295,
+      "grad_norm": 4.681489944458008,
+      "learning_rate": 9.591898811194048e-06,
+      "loss": 1.1064,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5278688524590164,
+      "grad_norm": 2.6617696285247803,
+      "learning_rate": 9.585266796245367e-06,
+      "loss": 0.9347,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5280737704918033,
+      "grad_norm": 7.010955810546875,
+      "learning_rate": 9.578634964020655e-06,
+      "loss": 1.0514,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5282786885245901,
+      "grad_norm": 3.335714340209961,
+      "learning_rate": 9.572003317441781e-06,
+      "loss": 0.8793,
+      "step": 2578
+    },
+    {
+      "epoch": 0.528483606557377,
+      "grad_norm": 4.527212619781494,
+      "learning_rate": 9.565371859430534e-06,
+      "loss": 1.3816,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5286885245901639,
+      "grad_norm": 3.636521577835083,
+      "learning_rate": 9.558740592908602e-06,
+      "loss": 0.8013,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5288934426229508,
+      "grad_norm": 3.8914687633514404,
+      "learning_rate": 9.552109520797614e-06,
+      "loss": 0.9273,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5290983606557377,
+      "grad_norm": 4.246948719024658,
+      "learning_rate": 9.545478646019095e-06,
+      "loss": 1.1728,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5293032786885246,
+      "grad_norm": 6.18946647644043,
+      "learning_rate": 9.538847971494489e-06,
+      "loss": 0.8655,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5295081967213114,
+      "grad_norm": 4.522332668304443,
+      "learning_rate": 9.53221750014515e-06,
+      "loss": 1.0718,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5297131147540983,
+      "grad_norm": 4.346995830535889,
+      "learning_rate": 9.525587234892347e-06,
+      "loss": 0.9999,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5299180327868852,
+      "grad_norm": 2.3372607231140137,
+      "learning_rate": 9.51895717865726e-06,
+      "loss": 0.7138,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5301229508196721,
+      "grad_norm": 4.073917388916016,
+      "learning_rate": 9.51232733436096e-06,
+      "loss": 1.2228,
+      "step": 2587
+    },
+    {
+      "epoch": 0.530327868852459,
+      "grad_norm": 7.180388927459717,
+      "learning_rate": 9.505697704924447e-06,
+      "loss": 1.1111,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5305327868852459,
+      "grad_norm": 5.556789875030518,
+      "learning_rate": 9.499068293268615e-06,
+      "loss": 0.8585,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5307377049180327,
+      "grad_norm": 4.067225933074951,
+      "learning_rate": 9.492439102314257e-06,
+      "loss": 1.1493,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5309426229508196,
+      "grad_norm": 5.0722479820251465,
+      "learning_rate": 9.48581013498208e-06,
+      "loss": 1.2278,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5311475409836065,
+      "grad_norm": 3.665172576904297,
+      "learning_rate": 9.479181394192688e-06,
+      "loss": 0.8788,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5313524590163935,
+      "grad_norm": 3.188403367996216,
+      "learning_rate": 9.472552882866585e-06,
+      "loss": 0.8867,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5315573770491804,
+      "grad_norm": 6.73194694519043,
+      "learning_rate": 9.46592460392417e-06,
+      "loss": 1.0489,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5317622950819673,
+      "grad_norm": 5.215070724487305,
+      "learning_rate": 9.459296560285742e-06,
+      "loss": 1.1661,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5319672131147541,
+      "grad_norm": 5.975355625152588,
+      "learning_rate": 9.452668754871508e-06,
+      "loss": 0.7404,
+      "step": 2596
+    },
+    {
+      "epoch": 0.532172131147541,
+      "grad_norm": 5.054249286651611,
+      "learning_rate": 9.446041190601545e-06,
+      "loss": 0.8131,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5323770491803279,
+      "grad_norm": 4.05131196975708,
+      "learning_rate": 9.43941387039585e-06,
+      "loss": 0.7198,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5325819672131148,
+      "grad_norm": 7.3958611488342285,
+      "learning_rate": 9.432786797174293e-06,
+      "loss": 1.219,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5327868852459017,
+      "grad_norm": 2.743455648422241,
+      "learning_rate": 9.42615997385665e-06,
+      "loss": 0.9315,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5329918032786886,
+      "grad_norm": 3.1403615474700928,
+      "learning_rate": 9.419533403362575e-06,
+      "loss": 0.8464,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5331967213114754,
+      "grad_norm": 5.53405237197876,
+      "learning_rate": 9.412907088611617e-06,
+      "loss": 1.0983,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5334016393442623,
+      "grad_norm": 4.879096508026123,
+      "learning_rate": 9.40628103252321e-06,
+      "loss": 0.8451,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5336065573770492,
+      "grad_norm": 3.330064058303833,
+      "learning_rate": 9.399655238016674e-06,
+      "loss": 1.1912,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5338114754098361,
+      "grad_norm": 3.1048531532287598,
+      "learning_rate": 9.393029708011214e-06,
+      "loss": 1.187,
+      "step": 2605
+    },
+    {
+      "epoch": 0.534016393442623,
+      "grad_norm": 6.3796610832214355,
+      "learning_rate": 9.386404445425924e-06,
+      "loss": 1.1214,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5342213114754099,
+      "grad_norm": 3.790882110595703,
+      "learning_rate": 9.379779453179773e-06,
+      "loss": 0.8966,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5344262295081967,
+      "grad_norm": 2.388540744781494,
+      "learning_rate": 9.373154734191612e-06,
+      "loss": 0.9482,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5346311475409836,
+      "grad_norm": 5.926338195800781,
+      "learning_rate": 9.366530291380172e-06,
+      "loss": 1.2589,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5348360655737705,
+      "grad_norm": 3.762467861175537,
+      "learning_rate": 9.359906127664067e-06,
+      "loss": 0.7122,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5350409836065574,
+      "grad_norm": 2.3983023166656494,
+      "learning_rate": 9.35328224596178e-06,
+      "loss": 0.6423,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5352459016393443,
+      "grad_norm": 2.3740310668945312,
+      "learning_rate": 9.34665864919167e-06,
+      "loss": 0.7703,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5354508196721312,
+      "grad_norm": 2.614448070526123,
+      "learning_rate": 9.340035340271984e-06,
+      "loss": 0.9527,
+      "step": 2613
+    },
+    {
+      "epoch": 0.535655737704918,
+      "grad_norm": 5.487371921539307,
+      "learning_rate": 9.33341232212083e-06,
+      "loss": 0.8633,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5358606557377049,
+      "grad_norm": 7.00946569442749,
+      "learning_rate": 9.326789597656185e-06,
+      "loss": 1.1205,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5360655737704918,
+      "grad_norm": 3.8489956855773926,
+      "learning_rate": 9.320167169795905e-06,
+      "loss": 0.92,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5362704918032787,
+      "grad_norm": 2.386122226715088,
+      "learning_rate": 9.313545041457716e-06,
+      "loss": 0.737,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5364754098360656,
+      "grad_norm": 3.844996213912964,
+      "learning_rate": 9.306923215559202e-06,
+      "loss": 0.97,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5366803278688524,
+      "grad_norm": 3.618054151535034,
+      "learning_rate": 9.300301695017824e-06,
+      "loss": 1.0787,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5368852459016393,
+      "grad_norm": 4.082974433898926,
+      "learning_rate": 9.293680482750901e-06,
+      "loss": 1.0745,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5370901639344262,
+      "grad_norm": 5.80845308303833,
+      "learning_rate": 9.287059581675628e-06,
+      "loss": 0.8469,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5372950819672131,
+      "grad_norm": 5.592921257019043,
+      "learning_rate": 9.280438994709043e-06,
+      "loss": 1.1888,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 5.363214015960693,
+      "learning_rate": 9.273818724768066e-06,
+      "loss": 1.1473,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5377049180327869,
+      "grad_norm": 3.3239946365356445,
+      "learning_rate": 9.267198774769468e-06,
+      "loss": 1.0932,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5379098360655737,
+      "grad_norm": 3.7138726711273193,
+      "learning_rate": 9.260579147629875e-06,
+      "loss": 0.8995,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5381147540983606,
+      "grad_norm": 3.2749595642089844,
+      "learning_rate": 9.253959846265779e-06,
+      "loss": 0.8095,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5383196721311475,
+      "grad_norm": 2.999748945236206,
+      "learning_rate": 9.247340873593522e-06,
+      "loss": 1.183,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5385245901639344,
+      "grad_norm": 4.2548604011535645,
+      "learning_rate": 9.240722232529307e-06,
+      "loss": 0.9873,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5387295081967213,
+      "grad_norm": 6.084436893463135,
+      "learning_rate": 9.234103925989186e-06,
+      "loss": 0.9536,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5389344262295082,
+      "grad_norm": 3.3554980754852295,
+      "learning_rate": 9.227485956889063e-06,
+      "loss": 0.8684,
+      "step": 2630
+    },
+    {
+      "epoch": 0.539139344262295,
+      "grad_norm": 2.695868968963623,
+      "learning_rate": 9.220868328144708e-06,
+      "loss": 0.8159,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5393442622950819,
+      "grad_norm": 7.277331352233887,
+      "learning_rate": 9.21425104267171e-06,
+      "loss": 1.0678,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5395491803278688,
+      "grad_norm": 3.2276790142059326,
+      "learning_rate": 9.207634103385541e-06,
+      "loss": 0.99,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5397540983606557,
+      "grad_norm": 3.8584516048431396,
+      "learning_rate": 9.201017513201498e-06,
+      "loss": 0.814,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5399590163934426,
+      "grad_norm": 3.781993865966797,
+      "learning_rate": 9.194401275034734e-06,
+      "loss": 0.9401,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5401639344262295,
+      "grad_norm": 4.335963726043701,
+      "learning_rate": 9.187785391800244e-06,
+      "loss": 0.9809,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5403688524590164,
+      "grad_norm": 2.947711706161499,
+      "learning_rate": 9.18116986641286e-06,
+      "loss": 0.9705,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5405737704918033,
+      "grad_norm": 3.962214231491089,
+      "learning_rate": 9.174554701787279e-06,
+      "loss": 1.0652,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5407786885245902,
+      "grad_norm": 6.967640399932861,
+      "learning_rate": 9.167939900838007e-06,
+      "loss": 1.3788,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5409836065573771,
+      "grad_norm": 3.08829665184021,
+      "learning_rate": 9.161325466479414e-06,
+      "loss": 0.786,
+      "step": 2640
+    },
+    {
+      "epoch": 0.541188524590164,
+      "grad_norm": 3.39744234085083,
+      "learning_rate": 9.1547114016257e-06,
+      "loss": 1.1692,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5413934426229509,
+      "grad_norm": 6.462059497833252,
+      "learning_rate": 9.148097709190903e-06,
+      "loss": 1.0132,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5415983606557377,
+      "grad_norm": 5.8720245361328125,
+      "learning_rate": 9.141484392088893e-06,
+      "loss": 1.0808,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5418032786885246,
+      "grad_norm": 2.421905517578125,
+      "learning_rate": 9.134871453233383e-06,
+      "loss": 0.953,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5420081967213115,
+      "grad_norm": 5.694443225860596,
+      "learning_rate": 9.128258895537913e-06,
+      "loss": 1.0325,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5422131147540984,
+      "grad_norm": 5.198636531829834,
+      "learning_rate": 9.121646721915854e-06,
+      "loss": 1.131,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5424180327868853,
+      "grad_norm": 4.6321024894714355,
+      "learning_rate": 9.11503493528041e-06,
+      "loss": 1.0429,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5426229508196722,
+      "grad_norm": 3.637981414794922,
+      "learning_rate": 9.108423538544622e-06,
+      "loss": 0.7907,
+      "step": 2648
+    },
+    {
+      "epoch": 0.542827868852459,
+      "grad_norm": 3.483645439147949,
+      "learning_rate": 9.101812534621348e-06,
+      "loss": 1.0119,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5430327868852459,
+      "grad_norm": 3.242119789123535,
+      "learning_rate": 9.095201926423275e-06,
+      "loss": 0.6574,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5432377049180328,
+      "grad_norm": 6.392824172973633,
+      "learning_rate": 9.088591716862921e-06,
+      "loss": 0.8724,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5434426229508197,
+      "grad_norm": 4.431067943572998,
+      "learning_rate": 9.081981908852626e-06,
+      "loss": 1.3567,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5436475409836066,
+      "grad_norm": 3.7763450145721436,
+      "learning_rate": 9.075372505304547e-06,
+      "loss": 0.9635,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5438524590163935,
+      "grad_norm": 5.97493839263916,
+      "learning_rate": 9.06876350913067e-06,
+      "loss": 0.8682,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5440573770491803,
+      "grad_norm": 3.647275447845459,
+      "learning_rate": 9.062154923242799e-06,
+      "loss": 1.0932,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5442622950819672,
+      "grad_norm": 4.437153339385986,
+      "learning_rate": 9.055546750552568e-06,
+      "loss": 1.1842,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5444672131147541,
+      "grad_norm": 5.961724281311035,
+      "learning_rate": 9.048938993971403e-06,
+      "loss": 1.0451,
+      "step": 2657
+    },
+    {
+      "epoch": 0.544672131147541,
+      "grad_norm": 3.900040626525879,
+      "learning_rate": 9.04233165641057e-06,
+      "loss": 1.1282,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5448770491803279,
+      "grad_norm": 6.093821048736572,
+      "learning_rate": 9.035724740781146e-06,
+      "loss": 1.4322,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5450819672131147,
+      "grad_norm": 4.618945121765137,
+      "learning_rate": 9.029118249994013e-06,
+      "loss": 0.8658,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5452868852459016,
+      "grad_norm": 5.765988349914551,
+      "learning_rate": 9.022512186959875e-06,
+      "loss": 1.1843,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5454918032786885,
+      "grad_norm": 4.653562545776367,
+      "learning_rate": 9.01590655458924e-06,
+      "loss": 0.935,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5456967213114754,
+      "grad_norm": 2.3071110248565674,
+      "learning_rate": 9.009301355792442e-06,
+      "loss": 0.9103,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5459016393442623,
+      "grad_norm": 4.943557262420654,
+      "learning_rate": 9.002696593479594e-06,
+      "loss": 0.9063,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5461065573770492,
+      "grad_norm": 9.073668479919434,
+      "learning_rate": 8.996092270560653e-06,
+      "loss": 1.0118,
+      "step": 2665
+    },
+    {
+      "epoch": 0.546311475409836,
+      "grad_norm": 3.4731860160827637,
+      "learning_rate": 8.989488389945356e-06,
+      "loss": 1.0547,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5465163934426229,
+      "grad_norm": 7.783008575439453,
+      "learning_rate": 8.982884954543252e-06,
+      "loss": 0.9875,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5467213114754098,
+      "grad_norm": 4.154879093170166,
+      "learning_rate": 8.976281967263701e-06,
+      "loss": 0.6651,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5469262295081967,
+      "grad_norm": 4.409552097320557,
+      "learning_rate": 8.969679431015857e-06,
+      "loss": 0.9856,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5471311475409836,
+      "grad_norm": 4.55764102935791,
+      "learning_rate": 8.96307734870868e-06,
+      "loss": 1.0901,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5473360655737705,
+      "grad_norm": 3.241694927215576,
+      "learning_rate": 8.956475723250927e-06,
+      "loss": 0.8658,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5475409836065573,
+      "grad_norm": 6.328608512878418,
+      "learning_rate": 8.949874557551151e-06,
+      "loss": 0.9418,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5477459016393442,
+      "grad_norm": 4.988103866577148,
+      "learning_rate": 8.94327385451772e-06,
+      "loss": 1.2845,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5479508196721311,
+      "grad_norm": 6.3113603591918945,
+      "learning_rate": 8.936673617058767e-06,
+      "loss": 1.0238,
+      "step": 2674
+    },
+    {
+      "epoch": 0.548155737704918,
+      "grad_norm": 4.686042308807373,
+      "learning_rate": 8.93007384808225e-06,
+      "loss": 0.9815,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5483606557377049,
+      "grad_norm": 3.223557949066162,
+      "learning_rate": 8.923474550495907e-06,
+      "loss": 0.9308,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5485655737704918,
+      "grad_norm": 4.654681205749512,
+      "learning_rate": 8.916875727207266e-06,
+      "loss": 1.0716,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5487704918032786,
+      "grad_norm": 5.186700820922852,
+      "learning_rate": 8.91027738112365e-06,
+      "loss": 0.9799,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5489754098360655,
+      "grad_norm": 3.2020132541656494,
+      "learning_rate": 8.903679515152174e-06,
+      "loss": 1.0742,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5491803278688525,
+      "grad_norm": 3.7320501804351807,
+      "learning_rate": 8.897082132199738e-06,
+      "loss": 1.0372,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5493852459016394,
+      "grad_norm": 3.299164056777954,
+      "learning_rate": 8.890485235173026e-06,
+      "loss": 0.9553,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5495901639344263,
+      "grad_norm": 7.440738677978516,
+      "learning_rate": 8.883888826978512e-06,
+      "loss": 1.0931,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5497950819672132,
+      "grad_norm": 4.967722415924072,
+      "learning_rate": 8.877292910522461e-06,
+      "loss": 1.0437,
+      "step": 2683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.34787917137146,
+      "learning_rate": 8.870697488710914e-06,
+      "loss": 0.9205,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5502049180327869,
+      "grad_norm": 2.5406711101531982,
+      "learning_rate": 8.864102564449689e-06,
+      "loss": 1.0214,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5504098360655738,
+      "grad_norm": 5.517638206481934,
+      "learning_rate": 8.857508140644394e-06,
+      "loss": 0.8849,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5506147540983607,
+      "grad_norm": 3.4555909633636475,
+      "learning_rate": 8.850914220200416e-06,
+      "loss": 0.9394,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5508196721311476,
+      "grad_norm": 6.3233842849731445,
+      "learning_rate": 8.844320806022911e-06,
+      "loss": 0.9219,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5510245901639345,
+      "grad_norm": 3.32389760017395,
+      "learning_rate": 8.83772790101682e-06,
+      "loss": 0.7599,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5512295081967213,
+      "grad_norm": 3.682178020477295,
+      "learning_rate": 8.831135508086864e-06,
+      "loss": 0.904,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5514344262295082,
+      "grad_norm": 5.424732208251953,
+      "learning_rate": 8.82454363013753e-06,
+      "loss": 1.0806,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5516393442622951,
+      "grad_norm": 3.360597848892212,
+      "learning_rate": 8.817952270073075e-06,
+      "loss": 0.8792,
+      "step": 2692
+    },
+    {
+      "epoch": 0.551844262295082,
+      "grad_norm": 4.039639472961426,
+      "learning_rate": 8.81136143079754e-06,
+      "loss": 0.9026,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5520491803278689,
+      "grad_norm": 3.522721767425537,
+      "learning_rate": 8.804771115214729e-06,
+      "loss": 0.8099,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5522540983606558,
+      "grad_norm": 5.991953372955322,
+      "learning_rate": 8.798181326228212e-06,
+      "loss": 0.8699,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5524590163934426,
+      "grad_norm": 5.157790660858154,
+      "learning_rate": 8.791592066741335e-06,
+      "loss": 0.9268,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5526639344262295,
+      "grad_norm": 3.2096126079559326,
+      "learning_rate": 8.785003339657203e-06,
+      "loss": 1.0465,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5528688524590164,
+      "grad_norm": 3.201239824295044,
+      "learning_rate": 8.778415147878698e-06,
+      "loss": 1.0195,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5530737704918033,
+      "grad_norm": 3.357103109359741,
+      "learning_rate": 8.771827494308448e-06,
+      "loss": 1.0248,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5532786885245902,
+      "grad_norm": 5.170840740203857,
+      "learning_rate": 8.765240381848864e-06,
+      "loss": 0.9546,
+      "step": 2700
+    },
+    {
+      "epoch": 0.553483606557377,
+      "grad_norm": 3.780447483062744,
+      "learning_rate": 8.758653813402107e-06,
+      "loss": 0.9611,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5536885245901639,
+      "grad_norm": 4.652418613433838,
+      "learning_rate": 8.752067791870094e-06,
+      "loss": 1.2451,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5538934426229508,
+      "grad_norm": 3.9020092487335205,
+      "learning_rate": 8.745482320154514e-06,
+      "loss": 0.9443,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5540983606557377,
+      "grad_norm": 4.7623395919799805,
+      "learning_rate": 8.738897401156803e-06,
+      "loss": 0.982,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5543032786885246,
+      "grad_norm": 6.2035322189331055,
+      "learning_rate": 8.732313037778162e-06,
+      "loss": 0.7261,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5545081967213115,
+      "grad_norm": 3.893033742904663,
+      "learning_rate": 8.72572923291954e-06,
+      "loss": 1.0218,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5547131147540983,
+      "grad_norm": 3.333167791366577,
+      "learning_rate": 8.719145989481638e-06,
+      "loss": 0.9302,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5549180327868852,
+      "grad_norm": 3.155221939086914,
+      "learning_rate": 8.712563310364928e-06,
+      "loss": 0.7543,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5551229508196721,
+      "grad_norm": 4.056007385253906,
+      "learning_rate": 8.705981198469603e-06,
+      "loss": 1.1529,
+      "step": 2709
+    },
+    {
+      "epoch": 0.555327868852459,
+      "grad_norm": 4.454907417297363,
+      "learning_rate": 8.699399656695632e-06,
+      "loss": 0.9198,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5555327868852459,
+      "grad_norm": 2.710010528564453,
+      "learning_rate": 8.692818687942726e-06,
+      "loss": 0.9383,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5557377049180328,
+      "grad_norm": 4.430636882781982,
+      "learning_rate": 8.686238295110337e-06,
+      "loss": 0.9068,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5559426229508196,
+      "grad_norm": 3.8734583854675293,
+      "learning_rate": 8.679658481097668e-06,
+      "loss": 0.8977,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5561475409836065,
+      "grad_norm": 3.910374879837036,
+      "learning_rate": 8.673079248803662e-06,
+      "loss": 0.7578,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5563524590163934,
+      "grad_norm": 2.7227578163146973,
+      "learning_rate": 8.666500601127022e-06,
+      "loss": 0.9753,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5565573770491803,
+      "grad_norm": 3.310981035232544,
+      "learning_rate": 8.659922540966167e-06,
+      "loss": 0.974,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5567622950819672,
+      "grad_norm": 2.800330877304077,
+      "learning_rate": 8.653345071219283e-06,
+      "loss": 0.8107,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5569672131147541,
+      "grad_norm": 3.930035352706909,
+      "learning_rate": 8.646768194784278e-06,
+      "loss": 1.0896,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5571721311475409,
+      "grad_norm": 8.80814266204834,
+      "learning_rate": 8.640191914558811e-06,
+      "loss": 1.3373,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5573770491803278,
+      "grad_norm": 3.6826343536376953,
+      "learning_rate": 8.633616233440268e-06,
+      "loss": 1.061,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5575819672131147,
+      "grad_norm": 7.692549228668213,
+      "learning_rate": 8.627041154325775e-06,
+      "loss": 0.8963,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5577868852459016,
+      "grad_norm": 6.1796369552612305,
+      "learning_rate": 8.6204666801122e-06,
+      "loss": 1.1873,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5579918032786885,
+      "grad_norm": 4.979324817657471,
+      "learning_rate": 8.613892813696128e-06,
+      "loss": 0.8016,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5581967213114755,
+      "grad_norm": 4.033623218536377,
+      "learning_rate": 8.607319557973889e-06,
+      "loss": 1.0282,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5584016393442623,
+      "grad_norm": 4.140589237213135,
+      "learning_rate": 8.600746915841545e-06,
+      "loss": 0.8424,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5586065573770492,
+      "grad_norm": 5.365126609802246,
+      "learning_rate": 8.594174890194884e-06,
+      "loss": 1.3026,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5588114754098361,
+      "grad_norm": 2.9215915203094482,
+      "learning_rate": 8.587603483929417e-06,
+      "loss": 0.8575,
+      "step": 2727
+    },
+    {
+      "epoch": 0.559016393442623,
+      "grad_norm": 4.30432653427124,
+      "learning_rate": 8.58103269994039e-06,
+      "loss": 0.7097,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5592213114754099,
+      "grad_norm": 3.9046590328216553,
+      "learning_rate": 8.574462541122772e-06,
+      "loss": 0.8808,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5594262295081968,
+      "grad_norm": 3.015761613845825,
+      "learning_rate": 8.567893010371255e-06,
+      "loss": 1.0562,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5596311475409836,
+      "grad_norm": 5.069972038269043,
+      "learning_rate": 8.561324110580257e-06,
+      "loss": 1.0176,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5598360655737705,
+      "grad_norm": 3.851755142211914,
+      "learning_rate": 8.554755844643915e-06,
+      "loss": 0.815,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5600409836065574,
+      "grad_norm": 6.002426624298096,
+      "learning_rate": 8.548188215456094e-06,
+      "loss": 1.0082,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5602459016393443,
+      "grad_norm": 4.038698196411133,
+      "learning_rate": 8.541621225910366e-06,
+      "loss": 0.9453,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5604508196721312,
+      "grad_norm": 6.060431003570557,
+      "learning_rate": 8.53505487890003e-06,
+      "loss": 1.0168,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5606557377049181,
+      "grad_norm": 2.481205940246582,
+      "learning_rate": 8.528489177318107e-06,
+      "loss": 0.9062,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5608606557377049,
+      "grad_norm": 2.8082468509674072,
+      "learning_rate": 8.521924124057317e-06,
+      "loss": 1.0066,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5610655737704918,
+      "grad_norm": 4.123013973236084,
+      "learning_rate": 8.515359722010108e-06,
+      "loss": 0.9455,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5612704918032787,
+      "grad_norm": 3.994553565979004,
+      "learning_rate": 8.508795974068637e-06,
+      "loss": 0.9681,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5614754098360656,
+      "grad_norm": 3.364335536956787,
+      "learning_rate": 8.50223288312478e-06,
+      "loss": 0.7839,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5616803278688525,
+      "grad_norm": 2.934497117996216,
+      "learning_rate": 8.4956704520701e-06,
+      "loss": 0.9965,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5618852459016394,
+      "grad_norm": 3.6623849868774414,
+      "learning_rate": 8.489108683795902e-06,
+      "loss": 0.8277,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5620901639344262,
+      "grad_norm": 9.834036827087402,
+      "learning_rate": 8.482547581193177e-06,
+      "loss": 0.9429,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5622950819672131,
+      "grad_norm": 2.982361316680908,
+      "learning_rate": 8.475987147152626e-06,
+      "loss": 0.8881,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 3.102435350418091,
+      "learning_rate": 8.46942738456466e-06,
+      "loss": 0.9567,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5627049180327869,
+      "grad_norm": 7.003721714019775,
+      "learning_rate": 8.462868296319392e-06,
+      "loss": 1.0801,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5629098360655738,
+      "grad_norm": 2.9333434104919434,
+      "learning_rate": 8.456309885306641e-06,
+      "loss": 1.0337,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5631147540983606,
+      "grad_norm": 3.794985055923462,
+      "learning_rate": 8.44975215441592e-06,
+      "loss": 0.8101,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5633196721311475,
+      "grad_norm": 3.614070177078247,
+      "learning_rate": 8.443195106536445e-06,
+      "loss": 0.8299,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5635245901639344,
+      "grad_norm": 6.21343994140625,
+      "learning_rate": 8.436638744557143e-06,
+      "loss": 1.1121,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5637295081967213,
+      "grad_norm": 9.450660705566406,
+      "learning_rate": 8.43008307136662e-06,
+      "loss": 1.0386,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5639344262295082,
+      "grad_norm": 10.612711906433105,
+      "learning_rate": 8.42352808985319e-06,
+      "loss": 0.906,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5641393442622951,
+      "grad_norm": 2.831038475036621,
+      "learning_rate": 8.41697380290486e-06,
+      "loss": 0.9042,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5643442622950819,
+      "grad_norm": 3.3264667987823486,
+      "learning_rate": 8.410420213409332e-06,
+      "loss": 0.9895,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5645491803278688,
+      "grad_norm": 2.535308599472046,
+      "learning_rate": 8.403867324253996e-06,
+      "loss": 1.074,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5647540983606557,
+      "grad_norm": 3.0060184001922607,
+      "learning_rate": 8.39731513832594e-06,
+      "loss": 0.8718,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5649590163934426,
+      "grad_norm": 2.3842828273773193,
+      "learning_rate": 8.390763658511938e-06,
+      "loss": 0.8258,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5651639344262295,
+      "grad_norm": 2.9812686443328857,
+      "learning_rate": 8.384212887698453e-06,
+      "loss": 1.1015,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5653688524590164,
+      "grad_norm": 2.880849838256836,
+      "learning_rate": 8.377662828771634e-06,
+      "loss": 0.9204,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5655737704918032,
+      "grad_norm": 3.3016886711120605,
+      "learning_rate": 8.371113484617323e-06,
+      "loss": 0.6566,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5657786885245901,
+      "grad_norm": 5.4532976150512695,
+      "learning_rate": 8.364564858121047e-06,
+      "loss": 1.0049,
+      "step": 2761
+    },
+    {
+      "epoch": 0.565983606557377,
+      "grad_norm": 4.73813009262085,
+      "learning_rate": 8.358016952168004e-06,
+      "loss": 0.8523,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5661885245901639,
+      "grad_norm": 3.548567056655884,
+      "learning_rate": 8.351469769643087e-06,
+      "loss": 1.0073,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5663934426229508,
+      "grad_norm": 5.142325401306152,
+      "learning_rate": 8.344923313430868e-06,
+      "loss": 1.1294,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5665983606557377,
+      "grad_norm": 5.806443691253662,
+      "learning_rate": 8.338377586415595e-06,
+      "loss": 0.9803,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5668032786885245,
+      "grad_norm": 6.300876140594482,
+      "learning_rate": 8.331832591481194e-06,
+      "loss": 1.1142,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5670081967213115,
+      "grad_norm": 4.571526050567627,
+      "learning_rate": 8.32528833151128e-06,
+      "loss": 1.0426,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5672131147540984,
+      "grad_norm": 2.892002820968628,
+      "learning_rate": 8.318744809389135e-06,
+      "loss": 1.0331,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5674180327868853,
+      "grad_norm": 5.960763931274414,
+      "learning_rate": 8.312202027997712e-06,
+      "loss": 1.1332,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5676229508196722,
+      "grad_norm": 4.332674503326416,
+      "learning_rate": 8.305659990219644e-06,
+      "loss": 1.0592,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5678278688524591,
+      "grad_norm": 5.814267158508301,
+      "learning_rate": 8.299118698937237e-06,
+      "loss": 1.0653,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5680327868852459,
+      "grad_norm": 4.577042102813721,
+      "learning_rate": 8.292578157032462e-06,
+      "loss": 1.0821,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5682377049180328,
+      "grad_norm": 3.9962167739868164,
+      "learning_rate": 8.286038367386967e-06,
+      "loss": 1.066,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5684426229508197,
+      "grad_norm": 4.843022346496582,
+      "learning_rate": 8.279499332882063e-06,
+      "loss": 0.9041,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5686475409836066,
+      "grad_norm": 4.127387523651123,
+      "learning_rate": 8.272961056398739e-06,
+      "loss": 0.9047,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5688524590163935,
+      "grad_norm": 3.5057525634765625,
+      "learning_rate": 8.266423540817627e-06,
+      "loss": 0.8189,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5690573770491804,
+      "grad_norm": 6.430992603302002,
+      "learning_rate": 8.25988678901905e-06,
+      "loss": 1.2788,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5692622950819672,
+      "grad_norm": 4.142039775848389,
+      "learning_rate": 8.25335080388298e-06,
+      "loss": 0.9891,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5694672131147541,
+      "grad_norm": 4.975154399871826,
+      "learning_rate": 8.246815588289052e-06,
+      "loss": 1.0343,
+      "step": 2779
+    },
+    {
+      "epoch": 0.569672131147541,
+      "grad_norm": 4.084733009338379,
+      "learning_rate": 8.240281145116567e-06,
+      "loss": 1.0077,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5698770491803279,
+      "grad_norm": 3.3650546073913574,
+      "learning_rate": 8.23374747724448e-06,
+      "loss": 0.83,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5700819672131148,
+      "grad_norm": 4.697071552276611,
+      "learning_rate": 8.227214587551414e-06,
+      "loss": 1.129,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5702868852459017,
+      "grad_norm": 5.2644171714782715,
+      "learning_rate": 8.220682478915635e-06,
+      "loss": 0.8028,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5704918032786885,
+      "grad_norm": 2.533003568649292,
+      "learning_rate": 8.214151154215072e-06,
+      "loss": 0.9348,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5706967213114754,
+      "grad_norm": 4.872348308563232,
+      "learning_rate": 8.207620616327316e-06,
+      "loss": 0.8801,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5709016393442623,
+      "grad_norm": 4.922986030578613,
+      "learning_rate": 8.201090868129595e-06,
+      "loss": 0.9077,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5711065573770492,
+      "grad_norm": 6.5681657791137695,
+      "learning_rate": 8.194561912498807e-06,
+      "loss": 1.0875,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5713114754098361,
+      "grad_norm": 2.9934144020080566,
+      "learning_rate": 8.188033752311486e-06,
+      "loss": 0.9709,
+      "step": 2788
+    },
+    {
+      "epoch": 0.571516393442623,
+      "grad_norm": 4.259118556976318,
+      "learning_rate": 8.181506390443828e-06,
+      "loss": 0.9913,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5717213114754098,
+      "grad_norm": 3.56583571434021,
+      "learning_rate": 8.174979829771662e-06,
+      "loss": 0.9498,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5719262295081967,
+      "grad_norm": 4.312347412109375,
+      "learning_rate": 8.168454073170478e-06,
+      "loss": 1.0764,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5721311475409836,
+      "grad_norm": 6.218612194061279,
+      "learning_rate": 8.161929123515413e-06,
+      "loss": 1.0324,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5723360655737705,
+      "grad_norm": 3.3720006942749023,
+      "learning_rate": 8.15540498368123e-06,
+      "loss": 1.001,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5725409836065574,
+      "grad_norm": 3.338216543197632,
+      "learning_rate": 8.148881656542353e-06,
+      "loss": 0.8889,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5727459016393442,
+      "grad_norm": 3.497420310974121,
+      "learning_rate": 8.142359144972845e-06,
+      "loss": 1.0363,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5729508196721311,
+      "grad_norm": 3.7384514808654785,
+      "learning_rate": 8.135837451846404e-06,
+      "loss": 1.1788,
+      "step": 2796
+    },
+    {
+      "epoch": 0.573155737704918,
+      "grad_norm": 7.093274116516113,
+      "learning_rate": 8.12931658003637e-06,
+      "loss": 1.0658,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5733606557377049,
+      "grad_norm": 3.123792886734009,
+      "learning_rate": 8.122796532415724e-06,
+      "loss": 1.054,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5735655737704918,
+      "grad_norm": 5.054446697235107,
+      "learning_rate": 8.11627731185708e-06,
+      "loss": 1.0501,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5737704918032787,
+      "grad_norm": 3.5041120052337646,
+      "learning_rate": 8.109758921232688e-06,
+      "loss": 0.7655,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5739754098360655,
+      "grad_norm": 5.439658164978027,
+      "learning_rate": 8.103241363414432e-06,
+      "loss": 0.9173,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5741803278688524,
+      "grad_norm": 4.11661434173584,
+      "learning_rate": 8.096724641273834e-06,
+      "loss": 0.9596,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5743852459016393,
+      "grad_norm": 5.090843677520752,
+      "learning_rate": 8.090208757682047e-06,
+      "loss": 1.0492,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5745901639344262,
+      "grad_norm": 3.616626739501953,
+      "learning_rate": 8.083693715509843e-06,
+      "loss": 0.977,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5747950819672131,
+      "grad_norm": 10.666855812072754,
+      "learning_rate": 8.077179517627638e-06,
+      "loss": 0.9165,
+      "step": 2805
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 5.020249366760254,
+      "learning_rate": 8.070666166905473e-06,
+      "loss": 1.3629,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5752049180327868,
+      "grad_norm": 6.268934726715088,
+      "learning_rate": 8.064153666213007e-06,
+      "loss": 1.0582,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5754098360655737,
+      "grad_norm": 2.4577977657318115,
+      "learning_rate": 8.05764201841953e-06,
+      "loss": 0.9677,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5756147540983606,
+      "grad_norm": 3.283555269241333,
+      "learning_rate": 8.051131226393958e-06,
+      "loss": 0.8066,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5758196721311475,
+      "grad_norm": 8.772235870361328,
+      "learning_rate": 8.044621293004836e-06,
+      "loss": 0.9924,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5760245901639345,
+      "grad_norm": 3.1046111583709717,
+      "learning_rate": 8.03811222112031e-06,
+      "loss": 1.143,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5762295081967214,
+      "grad_norm": 3.4638099670410156,
+      "learning_rate": 8.03160401360817e-06,
+      "loss": 0.911,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5764344262295082,
+      "grad_norm": 3.0555810928344727,
+      "learning_rate": 8.025096673335812e-06,
+      "loss": 1.1158,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5766393442622951,
+      "grad_norm": 3.308623790740967,
+      "learning_rate": 8.018590203170252e-06,
+      "loss": 0.817,
+      "step": 2814
+    },
+    {
+      "epoch": 0.576844262295082,
+      "grad_norm": 3.7928576469421387,
+      "learning_rate": 8.012084605978121e-06,
+      "loss": 0.8873,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5770491803278689,
+      "grad_norm": 5.405666828155518,
+      "learning_rate": 8.00557988462567e-06,
+      "loss": 1.1655,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5772540983606558,
+      "grad_norm": 3.5454261302948,
+      "learning_rate": 7.999076041978769e-06,
+      "loss": 0.8354,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5774590163934427,
+      "grad_norm": 2.1475627422332764,
+      "learning_rate": 7.992573080902881e-06,
+      "loss": 0.8815,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5776639344262295,
+      "grad_norm": 3.0153775215148926,
+      "learning_rate": 7.986071004263104e-06,
+      "loss": 1.0627,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5778688524590164,
+      "grad_norm": 3.1182193756103516,
+      "learning_rate": 7.979569814924131e-06,
+      "loss": 0.6528,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5780737704918033,
+      "grad_norm": 3.3505334854125977,
+      "learning_rate": 7.973069515750266e-06,
+      "loss": 0.8021,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5782786885245902,
+      "grad_norm": 4.120420932769775,
+      "learning_rate": 7.966570109605429e-06,
+      "loss": 0.6361,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5784836065573771,
+      "grad_norm": 6.235356330871582,
+      "learning_rate": 7.960071599353137e-06,
+      "loss": 1.0278,
+      "step": 2823
+    },
+    {
+      "epoch": 0.578688524590164,
+      "grad_norm": 5.5441203117370605,
+      "learning_rate": 7.953573987856522e-06,
+      "loss": 0.8052,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5788934426229508,
+      "grad_norm": 2.8812525272369385,
+      "learning_rate": 7.94707727797831e-06,
+      "loss": 0.7976,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5790983606557377,
+      "grad_norm": 3.731182336807251,
+      "learning_rate": 7.940581472580831e-06,
+      "loss": 1.2931,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5793032786885246,
+      "grad_norm": 2.946089267730713,
+      "learning_rate": 7.93408657452603e-06,
+      "loss": 1.1763,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5795081967213115,
+      "grad_norm": 4.634861946105957,
+      "learning_rate": 7.927592586675431e-06,
+      "loss": 1.1548,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5797131147540984,
+      "grad_norm": 3.6617560386657715,
+      "learning_rate": 7.921099511890174e-06,
+      "loss": 0.972,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5799180327868853,
+      "grad_norm": 5.692076683044434,
+      "learning_rate": 7.914607353030989e-06,
+      "loss": 1.0674,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5801229508196721,
+      "grad_norm": 2.476396322250366,
+      "learning_rate": 7.908116112958208e-06,
+      "loss": 0.9913,
+      "step": 2831
+    },
+    {
+      "epoch": 0.580327868852459,
+      "grad_norm": 7.419476509094238,
+      "learning_rate": 7.90162579453175e-06,
+      "loss": 0.8247,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5805327868852459,
+      "grad_norm": 4.59544038772583,
+      "learning_rate": 7.895136400611132e-06,
+      "loss": 0.6513,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5807377049180328,
+      "grad_norm": 5.733709812164307,
+      "learning_rate": 7.888647934055467e-06,
+      "loss": 0.9662,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5809426229508197,
+      "grad_norm": 4.093591213226318,
+      "learning_rate": 7.882160397723452e-06,
+      "loss": 0.9312,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5811475409836065,
+      "grad_norm": 4.4393510818481445,
+      "learning_rate": 7.875673794473382e-06,
+      "loss": 1.1618,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5813524590163934,
+      "grad_norm": 7.923176288604736,
+      "learning_rate": 7.869188127163135e-06,
+      "loss": 0.9662,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5815573770491803,
+      "grad_norm": 4.76301908493042,
+      "learning_rate": 7.862703398650185e-06,
+      "loss": 0.8244,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5817622950819672,
+      "grad_norm": 3.505248546600342,
+      "learning_rate": 7.85621961179158e-06,
+      "loss": 1.1644,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5819672131147541,
+      "grad_norm": 3.661006450653076,
+      "learning_rate": 7.84973676944396e-06,
+      "loss": 0.9202,
+      "step": 2840
+    },
+    {
+      "epoch": 0.582172131147541,
+      "grad_norm": 3.8582708835601807,
+      "learning_rate": 7.843254874463554e-06,
+      "loss": 1.1565,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5823770491803278,
+      "grad_norm": 2.475719690322876,
+      "learning_rate": 7.836773929706162e-06,
+      "loss": 0.896,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5825819672131147,
+      "grad_norm": 5.10936975479126,
+      "learning_rate": 7.830293938027174e-06,
+      "loss": 0.9319,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5827868852459016,
+      "grad_norm": 2.8479061126708984,
+      "learning_rate": 7.823814902281553e-06,
+      "loss": 1.0589,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5829918032786885,
+      "grad_norm": 3.914551258087158,
+      "learning_rate": 7.817336825323858e-06,
+      "loss": 0.9621,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5831967213114754,
+      "grad_norm": 3.727564573287964,
+      "learning_rate": 7.810859710008197e-06,
+      "loss": 0.9232,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5834016393442623,
+      "grad_norm": 3.89494252204895,
+      "learning_rate": 7.80438355918828e-06,
+      "loss": 1.0125,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5836065573770491,
+      "grad_norm": 3.992284059524536,
+      "learning_rate": 7.797908375717382e-06,
+      "loss": 1.0088,
+      "step": 2848
+    },
+    {
+      "epoch": 0.583811475409836,
+      "grad_norm": 3.7627346515655518,
+      "learning_rate": 7.791434162448349e-06,
+      "loss": 1.2656,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5840163934426229,
+      "grad_norm": 2.9969613552093506,
+      "learning_rate": 7.784960922233603e-06,
+      "loss": 0.8761,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5842213114754098,
+      "grad_norm": 4.956835746765137,
+      "learning_rate": 7.778488657925134e-06,
+      "loss": 1.1293,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5844262295081967,
+      "grad_norm": 3.1216609477996826,
+      "learning_rate": 7.772017372374516e-06,
+      "loss": 0.6091,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5846311475409836,
+      "grad_norm": 2.4656982421875,
+      "learning_rate": 7.765547068432867e-06,
+      "loss": 1.0656,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5848360655737705,
+      "grad_norm": 3.959662437438965,
+      "learning_rate": 7.759077748950895e-06,
+      "loss": 0.703,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5850409836065574,
+      "grad_norm": 4.265832901000977,
+      "learning_rate": 7.752609416778864e-06,
+      "loss": 1.0419,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5852459016393443,
+      "grad_norm": 3.033813714981079,
+      "learning_rate": 7.746142074766602e-06,
+      "loss": 0.9722,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5854508196721312,
+      "grad_norm": 3.6453499794006348,
+      "learning_rate": 7.739675725763504e-06,
+      "loss": 1.0758,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5856557377049181,
+      "grad_norm": 4.494044303894043,
+      "learning_rate": 7.73321037261853e-06,
+      "loss": 1.1623,
+      "step": 2858
+    },
+    {
+      "epoch": 0.585860655737705,
+      "grad_norm": 5.8452467918396,
+      "learning_rate": 7.726746018180196e-06,
+      "loss": 1.0417,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5860655737704918,
+      "grad_norm": 3.9912312030792236,
+      "learning_rate": 7.720282665296577e-06,
+      "loss": 0.8219,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5862704918032787,
+      "grad_norm": 6.494071960449219,
+      "learning_rate": 7.713820316815312e-06,
+      "loss": 0.9038,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5864754098360656,
+      "grad_norm": 3.4187498092651367,
+      "learning_rate": 7.7073589755836e-06,
+      "loss": 0.9263,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5866803278688525,
+      "grad_norm": 5.317602634429932,
+      "learning_rate": 7.700898644448183e-06,
+      "loss": 0.8985,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5868852459016394,
+      "grad_norm": 6.002202033996582,
+      "learning_rate": 7.694439326255372e-06,
+      "loss": 1.2188,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5870901639344263,
+      "grad_norm": 3.8294575214385986,
+      "learning_rate": 7.687981023851023e-06,
+      "loss": 0.9299,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5872950819672131,
+      "grad_norm": 3.886047601699829,
+      "learning_rate": 7.681523740080554e-06,
+      "loss": 1.1039,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 2.4638774394989014,
+      "learning_rate": 7.675067477788919e-06,
+      "loss": 0.9305,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5877049180327869,
+      "grad_norm": 5.533035755157471,
+      "learning_rate": 7.668612239820636e-06,
+      "loss": 0.9746,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5879098360655738,
+      "grad_norm": 4.783204555511475,
+      "learning_rate": 7.662158029019767e-06,
+      "loss": 0.8713,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5881147540983607,
+      "grad_norm": 4.475332736968994,
+      "learning_rate": 7.655704848229916e-06,
+      "loss": 1.0983,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5883196721311476,
+      "grad_norm": 2.8249828815460205,
+      "learning_rate": 7.64925270029424e-06,
+      "loss": 0.7843,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5885245901639344,
+      "grad_norm": 4.698386192321777,
+      "learning_rate": 7.642801588055441e-06,
+      "loss": 1.1237,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5887295081967213,
+      "grad_norm": 5.255701541900635,
+      "learning_rate": 7.636351514355766e-06,
+      "loss": 0.8356,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5889344262295082,
+      "grad_norm": 5.96900749206543,
+      "learning_rate": 7.629902482036996e-06,
+      "loss": 1.0729,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5891393442622951,
+      "grad_norm": 5.299187183380127,
+      "learning_rate": 7.623454493940457e-06,
+      "loss": 0.8875,
+      "step": 2875
+    },
+    {
+      "epoch": 0.589344262295082,
+      "grad_norm": 2.9992353916168213,
+      "learning_rate": 7.617007552907023e-06,
+      "loss": 0.977,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5895491803278688,
+      "grad_norm": 4.39481258392334,
+      "learning_rate": 7.610561661777094e-06,
+      "loss": 1.2396,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5897540983606557,
+      "grad_norm": 2.617504596710205,
+      "learning_rate": 7.604116823390611e-06,
+      "loss": 0.9958,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5899590163934426,
+      "grad_norm": 6.503381252288818,
+      "learning_rate": 7.597673040587061e-06,
+      "loss": 1.2847,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 3.3820717334747314,
+      "learning_rate": 7.591230316205459e-06,
+      "loss": 0.921,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5903688524590164,
+      "grad_norm": 4.314914226531982,
+      "learning_rate": 7.584788653084346e-06,
+      "loss": 0.7668,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5905737704918033,
+      "grad_norm": 8.153070449829102,
+      "learning_rate": 7.578348054061807e-06,
+      "loss": 1.0101,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5907786885245901,
+      "grad_norm": 3.6003758907318115,
+      "learning_rate": 7.571908521975455e-06,
+      "loss": 0.7304,
+      "step": 2883
+    },
+    {
+      "epoch": 0.590983606557377,
+      "grad_norm": 8.269102096557617,
+      "learning_rate": 7.565470059662428e-06,
+      "loss": 1.158,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5911885245901639,
+      "grad_norm": 2.9810750484466553,
+      "learning_rate": 7.559032669959398e-06,
+      "loss": 0.9692,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5913934426229508,
+      "grad_norm": 9.759629249572754,
+      "learning_rate": 7.5525963557025595e-06,
+      "loss": 0.9275,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5915983606557377,
+      "grad_norm": 4.96049165725708,
+      "learning_rate": 7.546161119727648e-06,
+      "loss": 1.1486,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5918032786885246,
+      "grad_norm": 4.140109062194824,
+      "learning_rate": 7.539726964869898e-06,
+      "loss": 1.0211,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5920081967213114,
+      "grad_norm": 4.790685176849365,
+      "learning_rate": 7.53329389396409e-06,
+      "loss": 0.6852,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5922131147540983,
+      "grad_norm": 3.3261725902557373,
+      "learning_rate": 7.52686190984452e-06,
+      "loss": 1.2232,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5924180327868852,
+      "grad_norm": 3.57787823677063,
+      "learning_rate": 7.520431015345e-06,
+      "loss": 0.9175,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5926229508196721,
+      "grad_norm": 6.94440221786499,
+      "learning_rate": 7.514001213298868e-06,
+      "loss": 1.2865,
+      "step": 2892
+    },
+    {
+      "epoch": 0.592827868852459,
+      "grad_norm": 4.0665998458862305,
+      "learning_rate": 7.50757250653898e-06,
+      "loss": 1.1416,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5930327868852459,
+      "grad_norm": 4.743961334228516,
+      "learning_rate": 7.501144897897707e-06,
+      "loss": 1.1148,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5932377049180327,
+      "grad_norm": 4.554562091827393,
+      "learning_rate": 7.494718390206937e-06,
+      "loss": 1.0704,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5934426229508196,
+      "grad_norm": 3.598200798034668,
+      "learning_rate": 7.488292986298072e-06,
+      "loss": 1.0034,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5936475409836065,
+      "grad_norm": 11.02628231048584,
+      "learning_rate": 7.4818686890020385e-06,
+      "loss": 1.1445,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5938524590163935,
+      "grad_norm": 3.5481443405151367,
+      "learning_rate": 7.47544550114925e-06,
+      "loss": 0.8771,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5940573770491804,
+      "grad_norm": 5.295269966125488,
+      "learning_rate": 7.469023425569661e-06,
+      "loss": 0.8574,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5942622950819673,
+      "grad_norm": 6.84586238861084,
+      "learning_rate": 7.462602465092716e-06,
+      "loss": 0.9977,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5944672131147541,
+      "grad_norm": 3.2961645126342773,
+      "learning_rate": 7.456182622547377e-06,
+      "loss": 1.0012,
+      "step": 2901
+    },
+    {
+      "epoch": 0.594672131147541,
+      "grad_norm": 3.3578672409057617,
+      "learning_rate": 7.449763900762108e-06,
+      "loss": 0.984,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5948770491803279,
+      "grad_norm": 4.331342697143555,
+      "learning_rate": 7.443346302564881e-06,
+      "loss": 1.0076,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5950819672131148,
+      "grad_norm": 3.323050022125244,
+      "learning_rate": 7.4369298307831835e-06,
+      "loss": 0.7904,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5952868852459017,
+      "grad_norm": 6.993074893951416,
+      "learning_rate": 7.430514488243985e-06,
+      "loss": 0.9258,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5954918032786886,
+      "grad_norm": 3.097355842590332,
+      "learning_rate": 7.4241002777737754e-06,
+      "loss": 0.8448,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5956967213114754,
+      "grad_norm": 2.2627415657043457,
+      "learning_rate": 7.417687202198542e-06,
+      "loss": 0.844,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5959016393442623,
+      "grad_norm": 4.495140552520752,
+      "learning_rate": 7.411275264343769e-06,
+      "loss": 0.9314,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5961065573770492,
+      "grad_norm": 3.4113974571228027,
+      "learning_rate": 7.404864467034439e-06,
+      "loss": 0.9948,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5963114754098361,
+      "grad_norm": 5.162919521331787,
+      "learning_rate": 7.398454813095033e-06,
+      "loss": 0.9868,
+      "step": 2910
+    },
+    {
+      "epoch": 0.596516393442623,
+      "grad_norm": 5.080804347991943,
+      "learning_rate": 7.392046305349534e-06,
+      "loss": 0.8085,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5967213114754099,
+      "grad_norm": 4.0576629638671875,
+      "learning_rate": 7.385638946621409e-06,
+      "loss": 1.0584,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5969262295081967,
+      "grad_norm": 4.016748428344727,
+      "learning_rate": 7.379232739733625e-06,
+      "loss": 0.9919,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5971311475409836,
+      "grad_norm": 6.018937110900879,
+      "learning_rate": 7.372827687508646e-06,
+      "loss": 0.6234,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5973360655737705,
+      "grad_norm": 2.9837806224823,
+      "learning_rate": 7.366423792768423e-06,
+      "loss": 0.9874,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5975409836065574,
+      "grad_norm": 5.006044387817383,
+      "learning_rate": 7.360021058334389e-06,
+      "loss": 1.0436,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5977459016393443,
+      "grad_norm": 5.53792667388916,
+      "learning_rate": 7.353619487027479e-06,
+      "loss": 0.7732,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5979508196721312,
+      "grad_norm": 3.018527030944824,
+      "learning_rate": 7.3472190816681114e-06,
+      "loss": 1.1199,
+      "step": 2918
+    },
+    {
+      "epoch": 0.598155737704918,
+      "grad_norm": 4.508271217346191,
+      "learning_rate": 7.340819845076185e-06,
+      "loss": 0.9444,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5983606557377049,
+      "grad_norm": 3.567796230316162,
+      "learning_rate": 7.334421780071089e-06,
+      "loss": 0.8106,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5985655737704918,
+      "grad_norm": 9.131518363952637,
+      "learning_rate": 7.328024889471696e-06,
+      "loss": 1.1493,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5987704918032787,
+      "grad_norm": 8.482489585876465,
+      "learning_rate": 7.3216291760963655e-06,
+      "loss": 1.192,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5989754098360656,
+      "grad_norm": 4.365199089050293,
+      "learning_rate": 7.3152346427629225e-06,
+      "loss": 0.7408,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5991803278688524,
+      "grad_norm": 11.461033821105957,
+      "learning_rate": 7.308841292288693e-06,
+      "loss": 1.1497,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5993852459016393,
+      "grad_norm": 2.0599420070648193,
+      "learning_rate": 7.302449127490471e-06,
+      "loss": 0.7554,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5995901639344262,
+      "grad_norm": 4.0596923828125,
+      "learning_rate": 7.296058151184524e-06,
+      "loss": 1.2251,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5997950819672131,
+      "grad_norm": 4.241912841796875,
+      "learning_rate": 7.289668366186603e-06,
+      "loss": 0.935,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.382880210876465,
+      "learning_rate": 7.283279775311931e-06,
+      "loss": 1.2015,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6002049180327869,
+      "grad_norm": 7.712989807128906,
+      "learning_rate": 7.2768923813752135e-06,
+      "loss": 0.9276,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6004098360655737,
+      "grad_norm": 6.502160549163818,
+      "learning_rate": 7.27050618719061e-06,
+      "loss": 1.0211,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6006147540983606,
+      "grad_norm": 4.3236613273620605,
+      "learning_rate": 7.26412119557177e-06,
+      "loss": 0.9791,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6008196721311475,
+      "grad_norm": 3.690648078918457,
+      "learning_rate": 7.2577374093318045e-06,
+      "loss": 0.7822,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6010245901639344,
+      "grad_norm": 3.673290491104126,
+      "learning_rate": 7.25135483128329e-06,
+      "loss": 1.0407,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6012295081967213,
+      "grad_norm": 3.2589595317840576,
+      "learning_rate": 7.244973464238281e-06,
+      "loss": 0.6199,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6014344262295082,
+      "grad_norm": 6.475401878356934,
+      "learning_rate": 7.23859331100829e-06,
+      "loss": 1.2689,
+      "step": 2935
+    },
+    {
+      "epoch": 0.601639344262295,
+      "grad_norm": 11.945892333984375,
+      "learning_rate": 7.2322143744043005e-06,
+      "loss": 0.9637,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6018442622950819,
+      "grad_norm": 6.000344753265381,
+      "learning_rate": 7.225836657236752e-06,
+      "loss": 1.1673,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6020491803278688,
+      "grad_norm": 4.167516708374023,
+      "learning_rate": 7.219460162315552e-06,
+      "loss": 0.9402,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6022540983606557,
+      "grad_norm": 3.1611194610595703,
+      "learning_rate": 7.213084892450079e-06,
+      "loss": 1.0572,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6024590163934426,
+      "grad_norm": 3.3896913528442383,
+      "learning_rate": 7.206710850449147e-06,
+      "loss": 1.0386,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6026639344262295,
+      "grad_norm": 2.795640230178833,
+      "learning_rate": 7.200338039121053e-06,
+      "loss": 0.7079,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6028688524590164,
+      "grad_norm": 4.578722953796387,
+      "learning_rate": 7.193966461273542e-06,
+      "loss": 0.94,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6030737704918033,
+      "grad_norm": 3.0467495918273926,
+      "learning_rate": 7.187596119713816e-06,
+      "loss": 0.9,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6032786885245902,
+      "grad_norm": 4.37275505065918,
+      "learning_rate": 7.181227017248532e-06,
+      "loss": 0.9562,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6034836065573771,
+      "grad_norm": 2.7493858337402344,
+      "learning_rate": 7.174859156683801e-06,
+      "loss": 1.0486,
+      "step": 2945
+    },
+    {
+      "epoch": 0.603688524590164,
+      "grad_norm": 5.55593204498291,
+      "learning_rate": 7.16849254082519e-06,
+      "loss": 0.8594,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6038934426229509,
+      "grad_norm": 3.9935367107391357,
+      "learning_rate": 7.162127172477713e-06,
+      "loss": 1.0184,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6040983606557377,
+      "grad_norm": 2.319489002227783,
+      "learning_rate": 7.155763054445833e-06,
+      "loss": 0.959,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6043032786885246,
+      "grad_norm": 4.575728893280029,
+      "learning_rate": 7.149400189533473e-06,
+      "loss": 1.2259,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6045081967213115,
+      "grad_norm": 4.295917510986328,
+      "learning_rate": 7.143038580543995e-06,
+      "loss": 0.9397,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6047131147540984,
+      "grad_norm": 3.237816095352173,
+      "learning_rate": 7.136678230280205e-06,
+      "loss": 1.1022,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6049180327868853,
+      "grad_norm": 3.6444404125213623,
+      "learning_rate": 7.130319141544362e-06,
+      "loss": 1.015,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6051229508196722,
+      "grad_norm": 5.008123397827148,
+      "learning_rate": 7.123961317138166e-06,
+      "loss": 1.0807,
+      "step": 2953
+    },
+    {
+      "epoch": 0.605327868852459,
+      "grad_norm": 4.627206802368164,
+      "learning_rate": 7.1176047598627575e-06,
+      "loss": 0.9587,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6055327868852459,
+      "grad_norm": 2.9460976123809814,
+      "learning_rate": 7.111249472518718e-06,
+      "loss": 0.8081,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6057377049180328,
+      "grad_norm": 5.652465343475342,
+      "learning_rate": 7.104895457906082e-06,
+      "loss": 1.0604,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6059426229508197,
+      "grad_norm": 5.8128485679626465,
+      "learning_rate": 7.098542718824308e-06,
+      "loss": 0.9099,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6061475409836066,
+      "grad_norm": 3.0612573623657227,
+      "learning_rate": 7.092191258072298e-06,
+      "loss": 0.9451,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6063524590163935,
+      "grad_norm": 5.709494113922119,
+      "learning_rate": 7.085841078448393e-06,
+      "loss": 1.1969,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6065573770491803,
+      "grad_norm": 4.271442413330078,
+      "learning_rate": 7.079492182750368e-06,
+      "loss": 1.0545,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6067622950819672,
+      "grad_norm": 4.97298526763916,
+      "learning_rate": 7.073144573775428e-06,
+      "loss": 0.9015,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6069672131147541,
+      "grad_norm": 4.896202087402344,
+      "learning_rate": 7.066798254320219e-06,
+      "loss": 1.0863,
+      "step": 2962
+    },
+    {
+      "epoch": 0.607172131147541,
+      "grad_norm": 4.035538673400879,
+      "learning_rate": 7.0604532271808105e-06,
+      "loss": 1.1783,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6073770491803279,
+      "grad_norm": 3.538221836090088,
+      "learning_rate": 7.054109495152717e-06,
+      "loss": 0.8979,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6075819672131147,
+      "grad_norm": 3.274289846420288,
+      "learning_rate": 7.047767061030858e-06,
+      "loss": 0.7037,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6077868852459016,
+      "grad_norm": 2.7325758934020996,
+      "learning_rate": 7.041425927609607e-06,
+      "loss": 1.0248,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6079918032786885,
+      "grad_norm": 2.6730661392211914,
+      "learning_rate": 7.03508609768275e-06,
+      "loss": 0.8994,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6081967213114754,
+      "grad_norm": 3.462167978286743,
+      "learning_rate": 7.0287475740434995e-06,
+      "loss": 1.089,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6084016393442623,
+      "grad_norm": 4.156543731689453,
+      "learning_rate": 7.022410359484495e-06,
+      "loss": 1.048,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6086065573770492,
+      "grad_norm": 5.936999797821045,
+      "learning_rate": 7.016074456797803e-06,
+      "loss": 1.1053,
+      "step": 2970
+    },
+    {
+      "epoch": 0.608811475409836,
+      "grad_norm": 3.6216306686401367,
+      "learning_rate": 7.009739868774905e-06,
+      "loss": 1.1856,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6090163934426229,
+      "grad_norm": 5.764642238616943,
+      "learning_rate": 7.003406598206706e-06,
+      "loss": 1.0465,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6092213114754098,
+      "grad_norm": 4.733697891235352,
+      "learning_rate": 6.997074647883527e-06,
+      "loss": 0.8016,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6094262295081967,
+      "grad_norm": 3.549466133117676,
+      "learning_rate": 6.9907440205951235e-06,
+      "loss": 0.8984,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6096311475409836,
+      "grad_norm": 2.9366295337677,
+      "learning_rate": 6.984414719130641e-06,
+      "loss": 0.712,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6098360655737705,
+      "grad_norm": 2.4208931922912598,
+      "learning_rate": 6.978086746278665e-06,
+      "loss": 0.7133,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6100409836065573,
+      "grad_norm": 3.5475926399230957,
+      "learning_rate": 6.971760104827184e-06,
+      "loss": 0.8432,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6102459016393442,
+      "grad_norm": 4.769879341125488,
+      "learning_rate": 6.9654347975636015e-06,
+      "loss": 1.0105,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6104508196721311,
+      "grad_norm": 3.5416548252105713,
+      "learning_rate": 6.959110827274733e-06,
+      "loss": 0.976,
+      "step": 2979
+    },
+    {
+      "epoch": 0.610655737704918,
+      "grad_norm": 4.904038429260254,
+      "learning_rate": 6.952788196746806e-06,
+      "loss": 0.9985,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6108606557377049,
+      "grad_norm": 3.111579656600952,
+      "learning_rate": 6.946466908765464e-06,
+      "loss": 0.8799,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6110655737704918,
+      "grad_norm": 4.334231376647949,
+      "learning_rate": 6.94014696611574e-06,
+      "loss": 0.9508,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6112704918032786,
+      "grad_norm": 2.868680953979492,
+      "learning_rate": 6.933828371582096e-06,
+      "loss": 0.82,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6114754098360655,
+      "grad_norm": 3.4205567836761475,
+      "learning_rate": 6.9275111279483885e-06,
+      "loss": 1.0455,
+      "step": 2984
+    },
+    {
+      "epoch": 0.6116803278688525,
+      "grad_norm": 3.6585028171539307,
+      "learning_rate": 6.921195237997884e-06,
+      "loss": 1.1324,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6118852459016394,
+      "grad_norm": 4.5409626960754395,
+      "learning_rate": 6.9148807045132435e-06,
+      "loss": 1.1563,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6120901639344263,
+      "grad_norm": 3.1467208862304688,
+      "learning_rate": 6.90856753027654e-06,
+      "loss": 0.9936,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6122950819672132,
+      "grad_norm": 5.919410228729248,
+      "learning_rate": 6.902255718069247e-06,
+      "loss": 0.8004,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 5.569858551025391,
+      "learning_rate": 6.89594527067223e-06,
+      "loss": 1.0962,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6127049180327869,
+      "grad_norm": 5.592573165893555,
+      "learning_rate": 6.889636190865759e-06,
+      "loss": 1.1,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6129098360655738,
+      "grad_norm": 3.3599517345428467,
+      "learning_rate": 6.883328481429505e-06,
+      "loss": 0.7798,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6131147540983607,
+      "grad_norm": 4.424466609954834,
+      "learning_rate": 6.87702214514253e-06,
+      "loss": 1.134,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6133196721311476,
+      "grad_norm": 2.6366467475891113,
+      "learning_rate": 6.8707171847832884e-06,
+      "loss": 1.1399,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6135245901639345,
+      "grad_norm": 3.1947898864746094,
+      "learning_rate": 6.8644136031296335e-06,
+      "loss": 1.1722,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6137295081967213,
+      "grad_norm": 3.4746644496917725,
+      "learning_rate": 6.858111402958815e-06,
+      "loss": 0.757,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6139344262295082,
+      "grad_norm": 3.7095696926116943,
+      "learning_rate": 6.851810587047461e-06,
+      "loss": 0.7948,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6141393442622951,
+      "grad_norm": 3.582946538925171,
+      "learning_rate": 6.8455111581715995e-06,
+      "loss": 0.9411,
+      "step": 2997
+    },
+    {
+      "epoch": 0.614344262295082,
+      "grad_norm": 11.155003547668457,
+      "learning_rate": 6.839213119106644e-06,
+      "loss": 1.0018,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6145491803278689,
+      "grad_norm": 2.30126953125,
+      "learning_rate": 6.832916472627407e-06,
+      "loss": 0.7323,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6147540983606558,
+      "grad_norm": 3.2994024753570557,
+      "learning_rate": 6.826621221508063e-06,
+      "loss": 1.2932,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6149590163934426,
+      "grad_norm": 4.028141021728516,
+      "learning_rate": 6.820327368522196e-06,
+      "loss": 0.9492,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6151639344262295,
+      "grad_norm": 4.360920429229736,
+      "learning_rate": 6.814034916442762e-06,
+      "loss": 0.9819,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6153688524590164,
+      "grad_norm": 4.372984886169434,
+      "learning_rate": 6.807743868042098e-06,
+      "loss": 0.8597,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6155737704918033,
+      "grad_norm": 4.48314905166626,
+      "learning_rate": 6.801454226091932e-06,
+      "loss": 1.0748,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6157786885245902,
+      "grad_norm": 3.41589093208313,
+      "learning_rate": 6.79516599336336e-06,
+      "loss": 1.1039,
+      "step": 3005
+    },
+    {
+      "epoch": 0.615983606557377,
+      "grad_norm": 4.425970077514648,
+      "learning_rate": 6.7888791726268744e-06,
+      "loss": 1.0837,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6161885245901639,
+      "grad_norm": 3.440138816833496,
+      "learning_rate": 6.7825937666523235e-06,
+      "loss": 1.2843,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6163934426229508,
+      "grad_norm": 4.668529510498047,
+      "learning_rate": 6.7763097782089515e-06,
+      "loss": 0.81,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6165983606557377,
+      "grad_norm": 5.042308330535889,
+      "learning_rate": 6.77002721006537e-06,
+      "loss": 0.8486,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6168032786885246,
+      "grad_norm": 6.1208014488220215,
+      "learning_rate": 6.763746064989561e-06,
+      "loss": 1.113,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6170081967213115,
+      "grad_norm": 6.959681510925293,
+      "learning_rate": 6.757466345748887e-06,
+      "loss": 0.8222,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6172131147540983,
+      "grad_norm": 3.5055882930755615,
+      "learning_rate": 6.751188055110078e-06,
+      "loss": 0.9175,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6174180327868852,
+      "grad_norm": 3.1910462379455566,
+      "learning_rate": 6.744911195839237e-06,
+      "loss": 0.7817,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6176229508196721,
+      "grad_norm": 3.460913896560669,
+      "learning_rate": 6.738635770701834e-06,
+      "loss": 1.0123,
+      "step": 3014
+    },
+    {
+      "epoch": 0.617827868852459,
+      "grad_norm": 4.469415664672852,
+      "learning_rate": 6.732361782462704e-06,
+      "loss": 1.1255,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6180327868852459,
+      "grad_norm": 3.082031011581421,
+      "learning_rate": 6.726089233886064e-06,
+      "loss": 0.7713,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6182377049180328,
+      "grad_norm": 3.6253902912139893,
+      "learning_rate": 6.7198181277354734e-06,
+      "loss": 1.0842,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6184426229508196,
+      "grad_norm": 4.07529878616333,
+      "learning_rate": 6.713548466773876e-06,
+      "loss": 0.7773,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6186475409836065,
+      "grad_norm": 5.080194473266602,
+      "learning_rate": 6.707280253763568e-06,
+      "loss": 0.9895,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6188524590163934,
+      "grad_norm": 2.3339734077453613,
+      "learning_rate": 6.701013491466216e-06,
+      "loss": 1.0234,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6190573770491803,
+      "grad_norm": 3.913994073867798,
+      "learning_rate": 6.694748182642836e-06,
+      "loss": 0.8694,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6192622950819672,
+      "grad_norm": 5.195343971252441,
+      "learning_rate": 6.688484330053815e-06,
+      "loss": 1.0509,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6194672131147541,
+      "grad_norm": 2.4994120597839355,
+      "learning_rate": 6.682221936458893e-06,
+      "loss": 0.7966,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6196721311475409,
+      "grad_norm": 3.205369472503662,
+      "learning_rate": 6.675961004617163e-06,
+      "loss": 0.8943,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6198770491803278,
+      "grad_norm": 2.868053674697876,
+      "learning_rate": 6.669701537287081e-06,
+      "loss": 0.8815,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6200819672131147,
+      "grad_norm": 2.0975122451782227,
+      "learning_rate": 6.663443537226458e-06,
+      "loss": 0.8531,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6202868852459016,
+      "grad_norm": 2.721346139907837,
+      "learning_rate": 6.657187007192457e-06,
+      "loss": 0.8509,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6204918032786885,
+      "grad_norm": 6.57705020904541,
+      "learning_rate": 6.650931949941588e-06,
+      "loss": 0.8348,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6206967213114755,
+      "grad_norm": 3.6032891273498535,
+      "learning_rate": 6.6446783682297174e-06,
+      "loss": 1.0105,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6209016393442623,
+      "grad_norm": 3.4750730991363525,
+      "learning_rate": 6.638426264812064e-06,
+      "loss": 0.9199,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6211065573770492,
+      "grad_norm": 5.441538333892822,
+      "learning_rate": 6.632175642443187e-06,
+      "loss": 1.1953,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6213114754098361,
+      "grad_norm": 2.8458447456359863,
+      "learning_rate": 6.625926503876998e-06,
+      "loss": 0.9585,
+      "step": 3032
+    },
+    {
+      "epoch": 0.621516393442623,
+      "grad_norm": 3.3854637145996094,
+      "learning_rate": 6.6196788518667595e-06,
+      "loss": 1.0924,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6217213114754099,
+      "grad_norm": 3.769967794418335,
+      "learning_rate": 6.6134326891650736e-06,
+      "loss": 0.9224,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6219262295081968,
+      "grad_norm": 4.270944118499756,
+      "learning_rate": 6.607188018523883e-06,
+      "loss": 0.7862,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6221311475409836,
+      "grad_norm": 3.923025608062744,
+      "learning_rate": 6.6009448426944785e-06,
+      "loss": 0.9062,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6223360655737705,
+      "grad_norm": 4.866612911224365,
+      "learning_rate": 6.594703164427495e-06,
+      "loss": 0.9317,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6225409836065574,
+      "grad_norm": 4.617237091064453,
+      "learning_rate": 6.588462986472897e-06,
+      "loss": 0.8971,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6227459016393443,
+      "grad_norm": 5.835080623626709,
+      "learning_rate": 6.5822243115799965e-06,
+      "loss": 0.9355,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6229508196721312,
+      "grad_norm": 4.417764186859131,
+      "learning_rate": 6.575987142497441e-06,
+      "loss": 0.9715,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6231557377049181,
+      "grad_norm": 5.3223443031311035,
+      "learning_rate": 6.569751481973221e-06,
+      "loss": 0.8847,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6233606557377049,
+      "grad_norm": 4.828551292419434,
+      "learning_rate": 6.563517332754645e-06,
+      "loss": 1.0985,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6235655737704918,
+      "grad_norm": 2.025726556777954,
+      "learning_rate": 6.557284697588374e-06,
+      "loss": 0.7387,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6237704918032787,
+      "grad_norm": 4.741490364074707,
+      "learning_rate": 6.551053579220396e-06,
+      "loss": 0.8329,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6239754098360656,
+      "grad_norm": 2.516554832458496,
+      "learning_rate": 6.544823980396025e-06,
+      "loss": 1.032,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6241803278688525,
+      "grad_norm": 5.956765651702881,
+      "learning_rate": 6.53859590385991e-06,
+      "loss": 1.0233,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6243852459016394,
+      "grad_norm": 3.71528697013855,
+      "learning_rate": 6.532369352356032e-06,
+      "loss": 1.0334,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6245901639344262,
+      "grad_norm": 2.9537808895111084,
+      "learning_rate": 6.526144328627697e-06,
+      "loss": 1.0309,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6247950819672131,
+      "grad_norm": 5.428108215332031,
+      "learning_rate": 6.519920835417535e-06,
+      "loss": 0.8218,
+      "step": 3049
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 4.132347583770752,
+      "learning_rate": 6.5136988754675065e-06,
+      "loss": 1.0726,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6252049180327869,
+      "grad_norm": 6.492029190063477,
+      "learning_rate": 6.507478451518901e-06,
+      "loss": 0.9848,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6254098360655738,
+      "grad_norm": 3.393329620361328,
+      "learning_rate": 6.501259566312312e-06,
+      "loss": 0.7765,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6256147540983606,
+      "grad_norm": 2.5968616008758545,
+      "learning_rate": 6.4950422225876805e-06,
+      "loss": 0.9509,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6258196721311475,
+      "grad_norm": 3.330024480819702,
+      "learning_rate": 6.488826423084251e-06,
+      "loss": 0.9407,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6260245901639344,
+      "grad_norm": 3.067164421081543,
+      "learning_rate": 6.4826121705405945e-06,
+      "loss": 1.0265,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6262295081967213,
+      "grad_norm": 5.693222522735596,
+      "learning_rate": 6.476399467694595e-06,
+      "loss": 0.8039,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6264344262295082,
+      "grad_norm": 3.914616346359253,
+      "learning_rate": 6.470188317283458e-06,
+      "loss": 1.1227,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6266393442622951,
+      "grad_norm": 2.8386805057525635,
+      "learning_rate": 6.463978722043711e-06,
+      "loss": 0.9681,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6268442622950819,
+      "grad_norm": 3.698997974395752,
+      "learning_rate": 6.45777068471118e-06,
+      "loss": 1.0064,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6270491803278688,
+      "grad_norm": 3.3162684440612793,
+      "learning_rate": 6.451564208021022e-06,
+      "loss": 0.8426,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6272540983606557,
+      "grad_norm": 2.534482479095459,
+      "learning_rate": 6.445359294707694e-06,
+      "loss": 0.8352,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6274590163934426,
+      "grad_norm": 4.4653167724609375,
+      "learning_rate": 6.439155947504973e-06,
+      "loss": 1.0557,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6276639344262295,
+      "grad_norm": 4.072644233703613,
+      "learning_rate": 6.432954169145937e-06,
+      "loss": 0.9656,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6278688524590164,
+      "grad_norm": 3.7390997409820557,
+      "learning_rate": 6.426753962362981e-06,
+      "loss": 1.0493,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6280737704918032,
+      "grad_norm": 4.893221378326416,
+      "learning_rate": 6.420555329887806e-06,
+      "loss": 1.0238,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6282786885245901,
+      "grad_norm": 3.162691831588745,
+      "learning_rate": 6.4143582744514135e-06,
+      "loss": 0.9839,
+      "step": 3066
+    },
+    {
+      "epoch": 0.628483606557377,
+      "grad_norm": 7.287731170654297,
+      "learning_rate": 6.408162798784112e-06,
+      "loss": 0.8017,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6286885245901639,
+      "grad_norm": 3.21976637840271,
+      "learning_rate": 6.401968905615525e-06,
+      "loss": 0.7538,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6288934426229508,
+      "grad_norm": 3.229511260986328,
+      "learning_rate": 6.395776597674568e-06,
+      "loss": 0.8272,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6290983606557377,
+      "grad_norm": 5.254249095916748,
+      "learning_rate": 6.3895858776894544e-06,
+      "loss": 0.9356,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6293032786885245,
+      "grad_norm": 4.787110805511475,
+      "learning_rate": 6.383396748387708e-06,
+      "loss": 0.7672,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6295081967213115,
+      "grad_norm": 6.035008430480957,
+      "learning_rate": 6.377209212496149e-06,
+      "loss": 1.252,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6297131147540984,
+      "grad_norm": 3.594222068786621,
+      "learning_rate": 6.371023272740891e-06,
+      "loss": 1.091,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6299180327868853,
+      "grad_norm": 2.89103627204895,
+      "learning_rate": 6.364838931847347e-06,
+      "loss": 0.8798,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6301229508196722,
+      "grad_norm": 2.6374573707580566,
+      "learning_rate": 6.358656192540227e-06,
+      "loss": 1.1513,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6303278688524591,
+      "grad_norm": 2.988679885864258,
+      "learning_rate": 6.352475057543541e-06,
+      "loss": 0.9625,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6305327868852459,
+      "grad_norm": 3.709951162338257,
+      "learning_rate": 6.346295529580573e-06,
+      "loss": 0.8143,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6307377049180328,
+      "grad_norm": 5.120326042175293,
+      "learning_rate": 6.340117611373922e-06,
+      "loss": 0.9714,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6309426229508197,
+      "grad_norm": 3.4435441493988037,
+      "learning_rate": 6.333941305645466e-06,
+      "loss": 1.1073,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6311475409836066,
+      "grad_norm": 7.174446105957031,
+      "learning_rate": 6.327766615116368e-06,
+      "loss": 1.0661,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6313524590163935,
+      "grad_norm": 2.4811880588531494,
+      "learning_rate": 6.321593542507089e-06,
+      "loss": 0.8124,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6315573770491804,
+      "grad_norm": 3.1231350898742676,
+      "learning_rate": 6.315422090537373e-06,
+      "loss": 0.9648,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6317622950819672,
+      "grad_norm": 7.10088586807251,
+      "learning_rate": 6.309252261926258e-06,
+      "loss": 1.0517,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6319672131147541,
+      "grad_norm": 3.082724094390869,
+      "learning_rate": 6.303084059392044e-06,
+      "loss": 0.7767,
+      "step": 3084
+    },
+    {
+      "epoch": 0.632172131147541,
+      "grad_norm": 2.969245195388794,
+      "learning_rate": 6.29691748565234e-06,
+      "loss": 0.8967,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6323770491803279,
+      "grad_norm": 3.503936529159546,
+      "learning_rate": 6.29075254342403e-06,
+      "loss": 0.7572,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6325819672131148,
+      "grad_norm": 3.58308744430542,
+      "learning_rate": 6.284589235423267e-06,
+      "loss": 1.1157,
+      "step": 3087
+    },
+    {
+      "epoch": 0.6327868852459017,
+      "grad_norm": 4.746630668640137,
+      "learning_rate": 6.278427564365499e-06,
+      "loss": 0.8667,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6329918032786885,
+      "grad_norm": 6.183226585388184,
+      "learning_rate": 6.272267532965448e-06,
+      "loss": 1.0637,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6331967213114754,
+      "grad_norm": 3.377096652984619,
+      "learning_rate": 6.266109143937111e-06,
+      "loss": 1.067,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6334016393442623,
+      "grad_norm": 3.3289737701416016,
+      "learning_rate": 6.259952399993762e-06,
+      "loss": 1.0754,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6336065573770492,
+      "grad_norm": 4.35269832611084,
+      "learning_rate": 6.2537973038479505e-06,
+      "loss": 0.9379,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6338114754098361,
+      "grad_norm": 2.6460795402526855,
+      "learning_rate": 6.247643858211508e-06,
+      "loss": 0.7826,
+      "step": 3093
+    },
+    {
+      "epoch": 0.634016393442623,
+      "grad_norm": 3.5920090675354004,
+      "learning_rate": 6.241492065795522e-06,
+      "loss": 0.8022,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6342213114754098,
+      "grad_norm": 4.744017124176025,
+      "learning_rate": 6.2353419293103675e-06,
+      "loss": 0.9881,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6344262295081967,
+      "grad_norm": 3.3954269886016846,
+      "learning_rate": 6.229193451465682e-06,
+      "loss": 1.0723,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6346311475409836,
+      "grad_norm": 3.7355151176452637,
+      "learning_rate": 6.223046634970378e-06,
+      "loss": 0.9899,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6348360655737705,
+      "grad_norm": 3.22088360786438,
+      "learning_rate": 6.216901482532624e-06,
+      "loss": 1.0081,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6350409836065574,
+      "grad_norm": 2.2783315181732178,
+      "learning_rate": 6.2107579968598685e-06,
+      "loss": 0.7338,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6352459016393442,
+      "grad_norm": 2.6563801765441895,
+      "learning_rate": 6.204616180658822e-06,
+      "loss": 0.8472,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6354508196721311,
+      "grad_norm": 4.473625659942627,
+      "learning_rate": 6.1984760366354525e-06,
+      "loss": 0.7879,
+      "step": 3101
+    },
+    {
+      "epoch": 0.635655737704918,
+      "grad_norm": 4.103723526000977,
+      "learning_rate": 6.192337567494999e-06,
+      "loss": 0.81,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6358606557377049,
+      "grad_norm": 3.087249994277954,
+      "learning_rate": 6.186200775941963e-06,
+      "loss": 1.0663,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6360655737704918,
+      "grad_norm": 3.341689348220825,
+      "learning_rate": 6.180065664680104e-06,
+      "loss": 0.8317,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6362704918032787,
+      "grad_norm": 3.171376943588257,
+      "learning_rate": 6.173932236412439e-06,
+      "loss": 1.0041,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6364754098360655,
+      "grad_norm": 3.13665771484375,
+      "learning_rate": 6.167800493841245e-06,
+      "loss": 0.775,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6366803278688524,
+      "grad_norm": 2.654588222503662,
+      "learning_rate": 6.1616704396680615e-06,
+      "loss": 0.9571,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6368852459016393,
+      "grad_norm": 4.924312114715576,
+      "learning_rate": 6.155542076593676e-06,
+      "loss": 1.1339,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6370901639344262,
+      "grad_norm": 7.829956531524658,
+      "learning_rate": 6.149415407318133e-06,
+      "loss": 0.7711,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6372950819672131,
+      "grad_norm": 3.1782584190368652,
+      "learning_rate": 6.143290434540738e-06,
+      "loss": 0.939,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 3.6460213661193848,
+      "learning_rate": 6.137167160960045e-06,
+      "loss": 0.8227,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6377049180327868,
+      "grad_norm": 3.476234197616577,
+      "learning_rate": 6.1310455892738495e-06,
+      "loss": 0.7935,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6379098360655737,
+      "grad_norm": 3.402555465698242,
+      "learning_rate": 6.124925722179211e-06,
+      "loss": 1.1584,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6381147540983606,
+      "grad_norm": 4.854242324829102,
+      "learning_rate": 6.118807562372434e-06,
+      "loss": 0.8055,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6383196721311475,
+      "grad_norm": 6.391777992248535,
+      "learning_rate": 6.112691112549065e-06,
+      "loss": 1.0056,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6385245901639345,
+      "grad_norm": 6.7494797706604,
+      "learning_rate": 6.106576375403904e-06,
+      "loss": 1.0741,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6387295081967214,
+      "grad_norm": 4.06374979019165,
+      "learning_rate": 6.100463353630992e-06,
+      "loss": 0.8993,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6389344262295082,
+      "grad_norm": 4.789913177490234,
+      "learning_rate": 6.094352049923624e-06,
+      "loss": 1.0798,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6391393442622951,
+      "grad_norm": 3.793696403503418,
+      "learning_rate": 6.088242466974317e-06,
+      "loss": 1.2248,
+      "step": 3119
+    },
+    {
+      "epoch": 0.639344262295082,
+      "grad_norm": 2.9148385524749756,
+      "learning_rate": 6.082134607474852e-06,
+      "loss": 0.9559,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6395491803278689,
+      "grad_norm": 5.667548656463623,
+      "learning_rate": 6.076028474116243e-06,
+      "loss": 0.9983,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6397540983606558,
+      "grad_norm": 5.407018661499023,
+      "learning_rate": 6.069924069588737e-06,
+      "loss": 1.0686,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6399590163934427,
+      "grad_norm": 3.7814934253692627,
+      "learning_rate": 6.063821396581827e-06,
+      "loss": 0.6631,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6401639344262295,
+      "grad_norm": 9.548797607421875,
+      "learning_rate": 6.0577204577842395e-06,
+      "loss": 0.8966,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6403688524590164,
+      "grad_norm": 5.036262035369873,
+      "learning_rate": 6.051621255883942e-06,
+      "loss": 1.1007,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6405737704918033,
+      "grad_norm": 4.666416645050049,
+      "learning_rate": 6.045523793568128e-06,
+      "loss": 1.0121,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6407786885245902,
+      "grad_norm": 6.011282920837402,
+      "learning_rate": 6.039428073523228e-06,
+      "loss": 1.2516,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6409836065573771,
+      "grad_norm": 3.7584850788116455,
+      "learning_rate": 6.033334098434918e-06,
+      "loss": 1.1357,
+      "step": 3128
+    },
+    {
+      "epoch": 0.641188524590164,
+      "grad_norm": 3.5245559215545654,
+      "learning_rate": 6.0272418709880785e-06,
+      "loss": 1.0255,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6413934426229508,
+      "grad_norm": 3.634214162826538,
+      "learning_rate": 6.021151393866843e-06,
+      "loss": 0.8801,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6415983606557377,
+      "grad_norm": 2.063800811767578,
+      "learning_rate": 6.015062669754565e-06,
+      "loss": 0.7968,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6418032786885246,
+      "grad_norm": 4.929571151733398,
+      "learning_rate": 6.00897570133383e-06,
+      "loss": 1.0208,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6420081967213115,
+      "grad_norm": 3.307591199874878,
+      "learning_rate": 6.0028904912864395e-06,
+      "loss": 0.9876,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6422131147540984,
+      "grad_norm": 5.752949237823486,
+      "learning_rate": 5.996807042293429e-06,
+      "loss": 0.9634,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6424180327868853,
+      "grad_norm": 3.4266579151153564,
+      "learning_rate": 5.990725357035064e-06,
+      "loss": 0.8737,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6426229508196721,
+      "grad_norm": 3.1227030754089355,
+      "learning_rate": 5.9846454381908155e-06,
+      "loss": 0.9056,
+      "step": 3136
+    },
+    {
+      "epoch": 0.642827868852459,
+      "grad_norm": 3.495849132537842,
+      "learning_rate": 5.978567288439389e-06,
+      "loss": 0.845,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6430327868852459,
+      "grad_norm": 3.8165531158447266,
+      "learning_rate": 5.97249091045871e-06,
+      "loss": 0.7715,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6432377049180328,
+      "grad_norm": 4.206615924835205,
+      "learning_rate": 5.966416306925923e-06,
+      "loss": 0.8798,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6434426229508197,
+      "grad_norm": 2.722877025604248,
+      "learning_rate": 5.9603434805173835e-06,
+      "loss": 0.6959,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6436475409836065,
+      "grad_norm": 3.8439369201660156,
+      "learning_rate": 5.95427243390867e-06,
+      "loss": 1.0997,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6438524590163934,
+      "grad_norm": 2.7307393550872803,
+      "learning_rate": 5.948203169774581e-06,
+      "loss": 0.7954,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6440573770491803,
+      "grad_norm": 6.420963764190674,
+      "learning_rate": 5.9421356907891195e-06,
+      "loss": 0.9572,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6442622950819672,
+      "grad_norm": 3.4155280590057373,
+      "learning_rate": 5.9360699996255075e-06,
+      "loss": 1.0074,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6444672131147541,
+      "grad_norm": 2.709831714630127,
+      "learning_rate": 5.930006098956184e-06,
+      "loss": 1.012,
+      "step": 3145
+    },
+    {
+      "epoch": 0.644672131147541,
+      "grad_norm": 3.804936647415161,
+      "learning_rate": 5.923943991452792e-06,
+      "loss": 0.8456,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6448770491803278,
+      "grad_norm": 3.8567311763763428,
+      "learning_rate": 5.917883679786185e-06,
+      "loss": 1.2,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6450819672131147,
+      "grad_norm": 3.1079440116882324,
+      "learning_rate": 5.911825166626428e-06,
+      "loss": 1.0544,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6452868852459016,
+      "grad_norm": 3.4262969493865967,
+      "learning_rate": 5.905768454642795e-06,
+      "loss": 0.6626,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6454918032786885,
+      "grad_norm": 4.574223518371582,
+      "learning_rate": 5.8997135465037595e-06,
+      "loss": 0.8779,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6456967213114754,
+      "grad_norm": 3.389404535293579,
+      "learning_rate": 5.893660444877007e-06,
+      "loss": 0.8064,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6459016393442623,
+      "grad_norm": 5.859574794769287,
+      "learning_rate": 5.887609152429424e-06,
+      "loss": 1.1368,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6461065573770491,
+      "grad_norm": 6.824120044708252,
+      "learning_rate": 5.881559671827109e-06,
+      "loss": 1.0536,
+      "step": 3153
+    },
+    {
+      "epoch": 0.646311475409836,
+      "grad_norm": 3.3388702869415283,
+      "learning_rate": 5.87551200573534e-06,
+      "loss": 1.046,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6465163934426229,
+      "grad_norm": 5.180499076843262,
+      "learning_rate": 5.869466156818619e-06,
+      "loss": 1.0596,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6467213114754098,
+      "grad_norm": 5.572572708129883,
+      "learning_rate": 5.863422127740638e-06,
+      "loss": 1.2917,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6469262295081967,
+      "grad_norm": 4.1977386474609375,
+      "learning_rate": 5.857379921164283e-06,
+      "loss": 0.9267,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6471311475409836,
+      "grad_norm": 5.140130996704102,
+      "learning_rate": 5.851339539751644e-06,
+      "loss": 1.124,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6473360655737705,
+      "grad_norm": 3.8683688640594482,
+      "learning_rate": 5.8453009861640006e-06,
+      "loss": 1.1162,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6475409836065574,
+      "grad_norm": 8.430578231811523,
+      "learning_rate": 5.83926426306184e-06,
+      "loss": 1.0067,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6477459016393443,
+      "grad_norm": 4.131682395935059,
+      "learning_rate": 5.833229373104823e-06,
+      "loss": 0.7124,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6479508196721312,
+      "grad_norm": 4.765605449676514,
+      "learning_rate": 5.827196318951818e-06,
+      "loss": 1.0034,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6481557377049181,
+      "grad_norm": 2.575714111328125,
+      "learning_rate": 5.82116510326088e-06,
+      "loss": 0.9499,
+      "step": 3163
+    },
+    {
+      "epoch": 0.648360655737705,
+      "grad_norm": 11.99698543548584,
+      "learning_rate": 5.8151357286892515e-06,
+      "loss": 1.1777,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6485655737704918,
+      "grad_norm": 3.6729753017425537,
+      "learning_rate": 5.809108197893368e-06,
+      "loss": 0.9132,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6487704918032787,
+      "grad_norm": 2.8938217163085938,
+      "learning_rate": 5.803082513528853e-06,
+      "loss": 0.8866,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6489754098360656,
+      "grad_norm": 3.0478568077087402,
+      "learning_rate": 5.79705867825051e-06,
+      "loss": 1.051,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6491803278688525,
+      "grad_norm": 5.799971103668213,
+      "learning_rate": 5.791036694712336e-06,
+      "loss": 0.9038,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6493852459016394,
+      "grad_norm": 3.476093053817749,
+      "learning_rate": 5.785016565567505e-06,
+      "loss": 0.9477,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6495901639344263,
+      "grad_norm": 5.486820220947266,
+      "learning_rate": 5.778998293468384e-06,
+      "loss": 0.7232,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6497950819672131,
+      "grad_norm": 3.064617872238159,
+      "learning_rate": 5.772981881066506e-06,
+      "loss": 1.0458,
+      "step": 3171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.020503044128418,
+      "learning_rate": 5.766967331012598e-06,
+      "loss": 1.0351,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6502049180327869,
+      "grad_norm": 4.9088945388793945,
+      "learning_rate": 5.76095464595656e-06,
+      "loss": 1.0146,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6504098360655738,
+      "grad_norm": 3.521791696548462,
+      "learning_rate": 5.754943828547482e-06,
+      "loss": 1.1235,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6506147540983607,
+      "grad_norm": 4.152431488037109,
+      "learning_rate": 5.748934881433612e-06,
+      "loss": 1.0593,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6508196721311476,
+      "grad_norm": 5.602632522583008,
+      "learning_rate": 5.742927807262385e-06,
+      "loss": 1.0075,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6510245901639344,
+      "grad_norm": 4.131965160369873,
+      "learning_rate": 5.736922608680413e-06,
+      "loss": 0.7724,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6512295081967213,
+      "grad_norm": 6.6191511154174805,
+      "learning_rate": 5.7309192883334745e-06,
+      "loss": 1.0035,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6514344262295082,
+      "grad_norm": 4.852734088897705,
+      "learning_rate": 5.724917848866527e-06,
+      "loss": 0.7961,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6516393442622951,
+      "grad_norm": 3.176023244857788,
+      "learning_rate": 5.7189182929236965e-06,
+      "loss": 0.847,
+      "step": 3180
+    },
+    {
+      "epoch": 0.651844262295082,
+      "grad_norm": 4.522341251373291,
+      "learning_rate": 5.7129206231482815e-06,
+      "loss": 1.1002,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6520491803278688,
+      "grad_norm": 4.149558067321777,
+      "learning_rate": 5.706924842182739e-06,
+      "loss": 1.0059,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6522540983606557,
+      "grad_norm": 2.3053314685821533,
+      "learning_rate": 5.700930952668704e-06,
+      "loss": 0.8445,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6524590163934426,
+      "grad_norm": 3.796177625656128,
+      "learning_rate": 5.694938957246988e-06,
+      "loss": 1.0521,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6526639344262295,
+      "grad_norm": 5.937129974365234,
+      "learning_rate": 5.688948858557542e-06,
+      "loss": 0.7692,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6528688524590164,
+      "grad_norm": 10.294455528259277,
+      "learning_rate": 5.682960659239503e-06,
+      "loss": 1.3307,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6530737704918033,
+      "grad_norm": 2.477484703063965,
+      "learning_rate": 5.676974361931158e-06,
+      "loss": 0.8142,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6532786885245901,
+      "grad_norm": 7.762325286865234,
+      "learning_rate": 5.670989969269966e-06,
+      "loss": 1.0966,
+      "step": 3188
+    },
+    {
+      "epoch": 0.653483606557377,
+      "grad_norm": 3.6905510425567627,
+      "learning_rate": 5.665007483892544e-06,
+      "loss": 0.9735,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6536885245901639,
+      "grad_norm": 3.03662371635437,
+      "learning_rate": 5.659026908434663e-06,
+      "loss": 0.7116,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6538934426229508,
+      "grad_norm": 2.177253484725952,
+      "learning_rate": 5.653048245531263e-06,
+      "loss": 0.6087,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6540983606557377,
+      "grad_norm": 4.95379638671875,
+      "learning_rate": 5.647071497816422e-06,
+      "loss": 1.0978,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6543032786885246,
+      "grad_norm": 4.235795974731445,
+      "learning_rate": 5.6410966679234005e-06,
+      "loss": 0.9495,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6545081967213114,
+      "grad_norm": 4.607889175415039,
+      "learning_rate": 5.635123758484596e-06,
+      "loss": 1.0704,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6547131147540983,
+      "grad_norm": 2.513231039047241,
+      "learning_rate": 5.6291527721315675e-06,
+      "loss": 0.9672,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6549180327868852,
+      "grad_norm": 4.373373508453369,
+      "learning_rate": 5.623183711495019e-06,
+      "loss": 1.0173,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6551229508196721,
+      "grad_norm": 4.640996932983398,
+      "learning_rate": 5.617216579204814e-06,
+      "loss": 1.1207,
+      "step": 3197
+    },
+    {
+      "epoch": 0.655327868852459,
+      "grad_norm": 4.124151229858398,
+      "learning_rate": 5.6112513778899615e-06,
+      "loss": 1.1003,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6555327868852459,
+      "grad_norm": 3.116379499435425,
+      "learning_rate": 5.605288110178621e-06,
+      "loss": 0.8087,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 5.4974284172058105,
+      "learning_rate": 5.599326778698106e-06,
+      "loss": 1.1759,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6559426229508196,
+      "grad_norm": 9.298431396484375,
+      "learning_rate": 5.593367386074867e-06,
+      "loss": 0.8369,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6561475409836065,
+      "grad_norm": 4.913243293762207,
+      "learning_rate": 5.587409934934507e-06,
+      "loss": 1.0759,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6563524590163935,
+      "grad_norm": 4.08642578125,
+      "learning_rate": 5.581454427901769e-06,
+      "loss": 0.9875,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6565573770491804,
+      "grad_norm": 6.540504455566406,
+      "learning_rate": 5.575500867600545e-06,
+      "loss": 0.9836,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6567622950819673,
+      "grad_norm": 3.830272674560547,
+      "learning_rate": 5.569549256653868e-06,
+      "loss": 0.6412,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6569672131147541,
+      "grad_norm": 4.586799144744873,
+      "learning_rate": 5.563599597683905e-06,
+      "loss": 0.8196,
+      "step": 3206
+    },
+    {
+      "epoch": 0.657172131147541,
+      "grad_norm": 4.158080101013184,
+      "learning_rate": 5.557651893311971e-06,
+      "loss": 0.8803,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6573770491803279,
+      "grad_norm": 4.070335865020752,
+      "learning_rate": 5.5517061461585135e-06,
+      "loss": 0.9586,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6575819672131148,
+      "grad_norm": 3.377445697784424,
+      "learning_rate": 5.545762358843132e-06,
+      "loss": 1.062,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6577868852459017,
+      "grad_norm": 4.9805707931518555,
+      "learning_rate": 5.539820533984541e-06,
+      "loss": 0.9307,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6579918032786886,
+      "grad_norm": 5.96992826461792,
+      "learning_rate": 5.5338806742006075e-06,
+      "loss": 0.9503,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6581967213114754,
+      "grad_norm": 14.507769584655762,
+      "learning_rate": 5.527942782108324e-06,
+      "loss": 0.8502,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6584016393442623,
+      "grad_norm": 3.4704368114471436,
+      "learning_rate": 5.52200686032382e-06,
+      "loss": 1.1706,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6586065573770492,
+      "grad_norm": 7.559295177459717,
+      "learning_rate": 5.516072911462356e-06,
+      "loss": 0.9493,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6588114754098361,
+      "grad_norm": 3.020237922668457,
+      "learning_rate": 5.510140938138322e-06,
+      "loss": 1.0437,
+      "step": 3215
+    },
+    {
+      "epoch": 0.659016393442623,
+      "grad_norm": 3.8023593425750732,
+      "learning_rate": 5.504210942965242e-06,
+      "loss": 1.1752,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6592213114754099,
+      "grad_norm": 6.1116719245910645,
+      "learning_rate": 5.498282928555754e-06,
+      "loss": 0.8135,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6594262295081967,
+      "grad_norm": 4.31804084777832,
+      "learning_rate": 5.492356897521644e-06,
+      "loss": 1.2571,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6596311475409836,
+      "grad_norm": 2.383199691772461,
+      "learning_rate": 5.486432852473817e-06,
+      "loss": 0.7989,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6598360655737705,
+      "grad_norm": 2.3827731609344482,
+      "learning_rate": 5.480510796022294e-06,
+      "loss": 0.943,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6600409836065574,
+      "grad_norm": 6.082052707672119,
+      "learning_rate": 5.474590730776226e-06,
+      "loss": 1.0247,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6602459016393443,
+      "grad_norm": 4.581332206726074,
+      "learning_rate": 5.46867265934389e-06,
+      "loss": 0.9201,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6604508196721312,
+      "grad_norm": 3.2866029739379883,
+      "learning_rate": 5.46275658433268e-06,
+      "loss": 0.8781,
+      "step": 3223
+    },
+    {
+      "epoch": 0.660655737704918,
+      "grad_norm": 12.633234977722168,
+      "learning_rate": 5.456842508349112e-06,
+      "loss": 1.1584,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6608606557377049,
+      "grad_norm": 3.640861988067627,
+      "learning_rate": 5.4509304339988225e-06,
+      "loss": 0.8579,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6610655737704918,
+      "grad_norm": 4.30332612991333,
+      "learning_rate": 5.445020363886568e-06,
+      "loss": 0.9531,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6612704918032787,
+      "grad_norm": 6.212997913360596,
+      "learning_rate": 5.4391123006162075e-06,
+      "loss": 1.2411,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6614754098360656,
+      "grad_norm": 3.213942289352417,
+      "learning_rate": 5.4332062467907365e-06,
+      "loss": 0.7864,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6616803278688524,
+      "grad_norm": 3.906261444091797,
+      "learning_rate": 5.427302205012257e-06,
+      "loss": 1.1357,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6618852459016393,
+      "grad_norm": 7.876497268676758,
+      "learning_rate": 5.421400177881983e-06,
+      "loss": 1.3131,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6620901639344262,
+      "grad_norm": 3.4059839248657227,
+      "learning_rate": 5.415500168000234e-06,
+      "loss": 0.9214,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6622950819672131,
+      "grad_norm": 7.83441686630249,
+      "learning_rate": 5.409602177966453e-06,
+      "loss": 0.8674,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 9.862367630004883,
+      "learning_rate": 5.403706210379187e-06,
+      "loss": 1.2314,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6627049180327869,
+      "grad_norm": 5.191277980804443,
+      "learning_rate": 5.397812267836096e-06,
+      "loss": 1.0343,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6629098360655737,
+      "grad_norm": 5.164862632751465,
+      "learning_rate": 5.39192035293394e-06,
+      "loss": 0.7738,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6631147540983606,
+      "grad_norm": 4.787623882293701,
+      "learning_rate": 5.386030468268594e-06,
+      "loss": 0.9129,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6633196721311475,
+      "grad_norm": 5.89495325088501,
+      "learning_rate": 5.3801426164350335e-06,
+      "loss": 1.0115,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6635245901639344,
+      "grad_norm": 3.23479962348938,
+      "learning_rate": 5.374256800027341e-06,
+      "loss": 0.8894,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6637295081967213,
+      "grad_norm": 3.4466514587402344,
+      "learning_rate": 5.3683730216387e-06,
+      "loss": 1.1049,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6639344262295082,
+      "grad_norm": 3.170574903488159,
+      "learning_rate": 5.362491283861402e-06,
+      "loss": 0.9844,
+      "step": 3240
+    },
+    {
+      "epoch": 0.664139344262295,
+      "grad_norm": 5.711082458496094,
+      "learning_rate": 5.356611589286824e-06,
+      "loss": 1.0207,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6643442622950819,
+      "grad_norm": 2.481264114379883,
+      "learning_rate": 5.350733940505458e-06,
+      "loss": 1.0024,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6645491803278688,
+      "grad_norm": 4.727540969848633,
+      "learning_rate": 5.344858340106895e-06,
+      "loss": 1.0811,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6647540983606557,
+      "grad_norm": 7.475618839263916,
+      "learning_rate": 5.338984790679819e-06,
+      "loss": 0.8713,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6649590163934426,
+      "grad_norm": 4.920323371887207,
+      "learning_rate": 5.333113294812002e-06,
+      "loss": 0.8099,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6651639344262295,
+      "grad_norm": 4.5614399909973145,
+      "learning_rate": 5.327243855090324e-06,
+      "loss": 0.9541,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6653688524590164,
+      "grad_norm": 4.975894451141357,
+      "learning_rate": 5.321376474100753e-06,
+      "loss": 1.0556,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6655737704918033,
+      "grad_norm": 5.124459743499756,
+      "learning_rate": 5.315511154428354e-06,
+      "loss": 1.0697,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6657786885245902,
+      "grad_norm": 3.7209672927856445,
+      "learning_rate": 5.309647898657277e-06,
+      "loss": 0.8791,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6659836065573771,
+      "grad_norm": 3.842963218688965,
+      "learning_rate": 5.30378670937077e-06,
+      "loss": 0.927,
+      "step": 3250
+    },
+    {
+      "epoch": 0.666188524590164,
+      "grad_norm": 3.076547861099243,
+      "learning_rate": 5.29792758915117e-06,
+      "loss": 0.769,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6663934426229509,
+      "grad_norm": 2.5298640727996826,
+      "learning_rate": 5.292070540579888e-06,
+      "loss": 0.9127,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6665983606557377,
+      "grad_norm": 4.898927211761475,
+      "learning_rate": 5.286215566237445e-06,
+      "loss": 1.0297,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6668032786885246,
+      "grad_norm": 4.944405555725098,
+      "learning_rate": 5.28036266870344e-06,
+      "loss": 0.8873,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6670081967213115,
+      "grad_norm": 4.79062557220459,
+      "learning_rate": 5.274511850556543e-06,
+      "loss": 0.9519,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6672131147540984,
+      "grad_norm": 3.626382827758789,
+      "learning_rate": 5.268663114374524e-06,
+      "loss": 1.2326,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6674180327868853,
+      "grad_norm": 3.860326051712036,
+      "learning_rate": 5.2628164627342315e-06,
+      "loss": 0.6122,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6676229508196722,
+      "grad_norm": 3.3850696086883545,
+      "learning_rate": 5.256971898211593e-06,
+      "loss": 0.9458,
+      "step": 3258
+    },
+    {
+      "epoch": 0.667827868852459,
+      "grad_norm": 3.657270669937134,
+      "learning_rate": 5.251129423381622e-06,
+      "loss": 0.9006,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6680327868852459,
+      "grad_norm": 3.2005999088287354,
+      "learning_rate": 5.245289040818403e-06,
+      "loss": 0.9638,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6682377049180328,
+      "grad_norm": 3.839258909225464,
+      "learning_rate": 5.23945075309511e-06,
+      "loss": 0.9244,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6684426229508197,
+      "grad_norm": 5.121523857116699,
+      "learning_rate": 5.233614562783975e-06,
+      "loss": 1.1205,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6686475409836066,
+      "grad_norm": 11.140026092529297,
+      "learning_rate": 5.227780472456329e-06,
+      "loss": 0.8805,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6688524590163935,
+      "grad_norm": 5.900184631347656,
+      "learning_rate": 5.221948484682564e-06,
+      "loss": 0.9452,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6690573770491803,
+      "grad_norm": 4.5009074211120605,
+      "learning_rate": 5.216118602032152e-06,
+      "loss": 0.9356,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6692622950819672,
+      "grad_norm": 4.568884372711182,
+      "learning_rate": 5.210290827073627e-06,
+      "loss": 0.8104,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6694672131147541,
+      "grad_norm": 3.4325876235961914,
+      "learning_rate": 5.204465162374603e-06,
+      "loss": 0.8972,
+      "step": 3267
+    },
+    {
+      "epoch": 0.669672131147541,
+      "grad_norm": 2.893563747406006,
+      "learning_rate": 5.198641610501771e-06,
+      "loss": 1.1187,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6698770491803279,
+      "grad_norm": 3.326927900314331,
+      "learning_rate": 5.192820174020874e-06,
+      "loss": 1.0933,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6700819672131147,
+      "grad_norm": 4.45408821105957,
+      "learning_rate": 5.1870008554967356e-06,
+      "loss": 1.1653,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6702868852459016,
+      "grad_norm": 3.243682622909546,
+      "learning_rate": 5.181183657493241e-06,
+      "loss": 0.9098,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6704918032786885,
+      "grad_norm": 2.0776913166046143,
+      "learning_rate": 5.175368582573343e-06,
+      "loss": 0.6594,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6706967213114754,
+      "grad_norm": 4.429205894470215,
+      "learning_rate": 5.16955563329906e-06,
+      "loss": 0.9136,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6709016393442623,
+      "grad_norm": 2.161069869995117,
+      "learning_rate": 5.163744812231473e-06,
+      "loss": 0.6781,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6711065573770492,
+      "grad_norm": 4.35562801361084,
+      "learning_rate": 5.157936121930728e-06,
+      "loss": 1.2449,
+      "step": 3275
+    },
+    {
+      "epoch": 0.671311475409836,
+      "grad_norm": 8.172331809997559,
+      "learning_rate": 5.152129564956021e-06,
+      "loss": 1.0923,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6715163934426229,
+      "grad_norm": 3.3913097381591797,
+      "learning_rate": 5.146325143865617e-06,
+      "loss": 1.024,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6717213114754098,
+      "grad_norm": 2.5244810581207275,
+      "learning_rate": 5.140522861216846e-06,
+      "loss": 0.7154,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6719262295081967,
+      "grad_norm": 5.664602756500244,
+      "learning_rate": 5.134722719566092e-06,
+      "loss": 1.0063,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6721311475409836,
+      "grad_norm": 4.475277423858643,
+      "learning_rate": 5.128924721468782e-06,
+      "loss": 0.8198,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6723360655737705,
+      "grad_norm": 2.7978005409240723,
+      "learning_rate": 5.1231288694794144e-06,
+      "loss": 0.9627,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6725409836065573,
+      "grad_norm": 5.198509693145752,
+      "learning_rate": 5.117335166151536e-06,
+      "loss": 0.9862,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6727459016393442,
+      "grad_norm": 4.5353875160217285,
+      "learning_rate": 5.111543614037751e-06,
+      "loss": 0.8189,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6729508196721311,
+      "grad_norm": 6.88459587097168,
+      "learning_rate": 5.105754215689709e-06,
+      "loss": 1.04,
+      "step": 3284
+    },
+    {
+      "epoch": 0.673155737704918,
+      "grad_norm": 4.0733747482299805,
+      "learning_rate": 5.099966973658117e-06,
+      "loss": 0.8538,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6733606557377049,
+      "grad_norm": 4.571137428283691,
+      "learning_rate": 5.094181890492733e-06,
+      "loss": 0.8681,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6735655737704918,
+      "grad_norm": 4.53250789642334,
+      "learning_rate": 5.08839896874235e-06,
+      "loss": 1.0174,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6737704918032786,
+      "grad_norm": 4.144317150115967,
+      "learning_rate": 5.082618210954827e-06,
+      "loss": 1.2555,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6739754098360655,
+      "grad_norm": 4.276817798614502,
+      "learning_rate": 5.076839619677066e-06,
+      "loss": 0.8707,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6741803278688525,
+      "grad_norm": 8.373600959777832,
+      "learning_rate": 5.071063197454997e-06,
+      "loss": 1.0454,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6743852459016394,
+      "grad_norm": 3.112983226776123,
+      "learning_rate": 5.065288946833618e-06,
+      "loss": 1.2877,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6745901639344263,
+      "grad_norm": 4.505974292755127,
+      "learning_rate": 5.059516870356951e-06,
+      "loss": 1.0107,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6747950819672132,
+      "grad_norm": 4.385852336883545,
+      "learning_rate": 5.053746970568081e-06,
+      "loss": 0.9993,
+      "step": 3293
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 1.8929425477981567,
+      "learning_rate": 5.047979250009112e-06,
+      "loss": 0.7212,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6752049180327869,
+      "grad_norm": 5.105231285095215,
+      "learning_rate": 5.042213711221201e-06,
+      "loss": 0.9968,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6754098360655738,
+      "grad_norm": 5.3026628494262695,
+      "learning_rate": 5.0364503567445435e-06,
+      "loss": 0.9221,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6756147540983607,
+      "grad_norm": 2.810302734375,
+      "learning_rate": 5.030689189118358e-06,
+      "loss": 0.847,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6758196721311476,
+      "grad_norm": 3.6060473918914795,
+      "learning_rate": 5.024930210880924e-06,
+      "loss": 0.8587,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6760245901639345,
+      "grad_norm": 2.559358835220337,
+      "learning_rate": 5.019173424569538e-06,
+      "loss": 0.8262,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6762295081967213,
+      "grad_norm": 2.6443843841552734,
+      "learning_rate": 5.01341883272054e-06,
+      "loss": 0.8624,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6764344262295082,
+      "grad_norm": 2.7621212005615234,
+      "learning_rate": 5.007666437869294e-06,
+      "loss": 0.5976,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6766393442622951,
+      "grad_norm": 2.895416021347046,
+      "learning_rate": 5.0019162425502e-06,
+      "loss": 0.8866,
+      "step": 3302
+    },
+    {
+      "epoch": 0.676844262295082,
+      "grad_norm": 4.098370552062988,
+      "learning_rate": 4.996168249296701e-06,
+      "loss": 0.969,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6770491803278689,
+      "grad_norm": 5.3008646965026855,
+      "learning_rate": 4.99042246064125e-06,
+      "loss": 0.754,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6772540983606558,
+      "grad_norm": 4.393996715545654,
+      "learning_rate": 4.984678879115339e-06,
+      "loss": 0.9084,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6774590163934426,
+      "grad_norm": 3.6576082706451416,
+      "learning_rate": 4.978937507249488e-06,
+      "loss": 1.05,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6776639344262295,
+      "grad_norm": 3.49992036819458,
+      "learning_rate": 4.9731983475732415e-06,
+      "loss": 0.7952,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6778688524590164,
+      "grad_norm": 4.721869468688965,
+      "learning_rate": 4.96746140261517e-06,
+      "loss": 0.9843,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6780737704918033,
+      "grad_norm": 4.357363224029541,
+      "learning_rate": 4.961726674902868e-06,
+      "loss": 0.9985,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6782786885245902,
+      "grad_norm": 3.2776641845703125,
+      "learning_rate": 4.955994166962958e-06,
+      "loss": 1.0667,
+      "step": 3310
+    },
+    {
+      "epoch": 0.678483606557377,
+      "grad_norm": 2.8590707778930664,
+      "learning_rate": 4.950263881321069e-06,
+      "loss": 0.9357,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6786885245901639,
+      "grad_norm": 5.818782329559326,
+      "learning_rate": 4.944535820501864e-06,
+      "loss": 1.0204,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6788934426229508,
+      "grad_norm": 7.178651809692383,
+      "learning_rate": 4.938809987029032e-06,
+      "loss": 0.7165,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6790983606557377,
+      "grad_norm": 2.583544969558716,
+      "learning_rate": 4.933086383425269e-06,
+      "loss": 0.785,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6793032786885246,
+      "grad_norm": 5.848850727081299,
+      "learning_rate": 4.927365012212285e-06,
+      "loss": 0.8404,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6795081967213115,
+      "grad_norm": 3.961108446121216,
+      "learning_rate": 4.921645875910818e-06,
+      "loss": 0.8407,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6797131147540983,
+      "grad_norm": 4.472800254821777,
+      "learning_rate": 4.915928977040615e-06,
+      "loss": 0.9291,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6799180327868852,
+      "grad_norm": 2.6276087760925293,
+      "learning_rate": 4.91021431812044e-06,
+      "loss": 0.8845,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6801229508196721,
+      "grad_norm": 4.290722370147705,
+      "learning_rate": 4.904501901668066e-06,
+      "loss": 1.0262,
+      "step": 3319
+    },
+    {
+      "epoch": 0.680327868852459,
+      "grad_norm": 3.7782955169677734,
+      "learning_rate": 4.898791730200284e-06,
+      "loss": 0.7663,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6805327868852459,
+      "grad_norm": 2.636887550354004,
+      "learning_rate": 4.893083806232893e-06,
+      "loss": 0.9016,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6807377049180328,
+      "grad_norm": 3.938728094100952,
+      "learning_rate": 4.887378132280694e-06,
+      "loss": 0.8392,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6809426229508196,
+      "grad_norm": 5.17773962020874,
+      "learning_rate": 4.881674710857513e-06,
+      "loss": 0.9456,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6811475409836065,
+      "grad_norm": 4.4963531494140625,
+      "learning_rate": 4.875973544476173e-06,
+      "loss": 1.1182,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6813524590163934,
+      "grad_norm": 3.099020481109619,
+      "learning_rate": 4.8702746356485005e-06,
+      "loss": 0.9008,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6815573770491803,
+      "grad_norm": 4.317246437072754,
+      "learning_rate": 4.864577986885335e-06,
+      "loss": 0.8419,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6817622950819672,
+      "grad_norm": 6.04793119430542,
+      "learning_rate": 4.8588836006965135e-06,
+      "loss": 1.1333,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6819672131147541,
+      "grad_norm": 3.8902833461761475,
+      "learning_rate": 4.85319147959089e-06,
+      "loss": 1.0352,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6821721311475409,
+      "grad_norm": 3.3015854358673096,
+      "learning_rate": 4.847501626076301e-06,
+      "loss": 1.1001,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6823770491803278,
+      "grad_norm": 3.494065761566162,
+      "learning_rate": 4.841814042659595e-06,
+      "loss": 0.7776,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6825819672131147,
+      "grad_norm": 8.628152847290039,
+      "learning_rate": 4.83612873184662e-06,
+      "loss": 0.8598,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6827868852459016,
+      "grad_norm": 3.371544599533081,
+      "learning_rate": 4.830445696142222e-06,
+      "loss": 0.7934,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6829918032786885,
+      "grad_norm": 4.757236003875732,
+      "learning_rate": 4.824764938050242e-06,
+      "loss": 1.1177,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6831967213114755,
+      "grad_norm": 6.1230034828186035,
+      "learning_rate": 4.819086460073523e-06,
+      "loss": 1.1163,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6834016393442623,
+      "grad_norm": 2.7277281284332275,
+      "learning_rate": 4.8134102647139e-06,
+      "loss": 0.8962,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6836065573770492,
+      "grad_norm": 6.77896785736084,
+      "learning_rate": 4.807736354472198e-06,
+      "loss": 0.8714,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6838114754098361,
+      "grad_norm": 3.373610496520996,
+      "learning_rate": 4.802064731848237e-06,
+      "loss": 0.9787,
+      "step": 3337
+    },
+    {
+      "epoch": 0.684016393442623,
+      "grad_norm": 5.112484455108643,
+      "learning_rate": 4.796395399340844e-06,
+      "loss": 0.9846,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6842213114754099,
+      "grad_norm": 3.338949680328369,
+      "learning_rate": 4.790728359447813e-06,
+      "loss": 0.879,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6844262295081968,
+      "grad_norm": 2.4970149993896484,
+      "learning_rate": 4.7850636146659454e-06,
+      "loss": 0.8966,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6846311475409836,
+      "grad_norm": 3.437805414199829,
+      "learning_rate": 4.779401167491023e-06,
+      "loss": 1.0472,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6848360655737705,
+      "grad_norm": 4.8879876136779785,
+      "learning_rate": 4.773741020417819e-06,
+      "loss": 1.0174,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6850409836065574,
+      "grad_norm": 3.037790298461914,
+      "learning_rate": 4.768083175940093e-06,
+      "loss": 1.0865,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6852459016393443,
+      "grad_norm": 4.047876834869385,
+      "learning_rate": 4.762427636550586e-06,
+      "loss": 1.2541,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6854508196721312,
+      "grad_norm": 5.238515377044678,
+      "learning_rate": 4.756774404741035e-06,
+      "loss": 0.9031,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6856557377049181,
+      "grad_norm": 3.4126060009002686,
+      "learning_rate": 4.751123483002141e-06,
+      "loss": 0.5987,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6858606557377049,
+      "grad_norm": 4.465988636016846,
+      "learning_rate": 4.7454748738236e-06,
+      "loss": 0.9331,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6860655737704918,
+      "grad_norm": 3.3896453380584717,
+      "learning_rate": 4.739828579694095e-06,
+      "loss": 0.9711,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6862704918032787,
+      "grad_norm": 2.9937899112701416,
+      "learning_rate": 4.73418460310128e-06,
+      "loss": 0.8578,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6864754098360656,
+      "grad_norm": 2.913877010345459,
+      "learning_rate": 4.728542946531784e-06,
+      "loss": 0.7127,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6866803278688525,
+      "grad_norm": 8.79892349243164,
+      "learning_rate": 4.722903612471219e-06,
+      "loss": 0.9187,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6868852459016394,
+      "grad_norm": 4.369035243988037,
+      "learning_rate": 4.717266603404178e-06,
+      "loss": 1.0666,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6870901639344262,
+      "grad_norm": 3.655097246170044,
+      "learning_rate": 4.711631921814224e-06,
+      "loss": 0.827,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6872950819672131,
+      "grad_norm": 3.7337284088134766,
+      "learning_rate": 4.705999570183895e-06,
+      "loss": 1.0167,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 2.7115960121154785,
+      "learning_rate": 4.700369550994707e-06,
+      "loss": 0.9207,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6877049180327869,
+      "grad_norm": 3.8604652881622314,
+      "learning_rate": 4.694741866727141e-06,
+      "loss": 0.7556,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6879098360655738,
+      "grad_norm": 4.130412578582764,
+      "learning_rate": 4.689116519860655e-06,
+      "loss": 0.9837,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6881147540983606,
+      "grad_norm": 4.554333209991455,
+      "learning_rate": 4.683493512873675e-06,
+      "loss": 1.2827,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6883196721311475,
+      "grad_norm": 6.359424591064453,
+      "learning_rate": 4.677872848243601e-06,
+      "loss": 0.8682,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6885245901639344,
+      "grad_norm": 9.780887603759766,
+      "learning_rate": 4.672254528446788e-06,
+      "loss": 0.7213,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6887295081967213,
+      "grad_norm": 3.7541913986206055,
+      "learning_rate": 4.666638555958572e-06,
+      "loss": 0.92,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6889344262295082,
+      "grad_norm": 2.969896078109741,
+      "learning_rate": 4.661024933253242e-06,
+      "loss": 0.8352,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6891393442622951,
+      "grad_norm": 5.145637035369873,
+      "learning_rate": 4.655413662804073e-06,
+      "loss": 0.9145,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6893442622950819,
+      "grad_norm": 3.1089272499084473,
+      "learning_rate": 4.649804747083279e-06,
+      "loss": 0.8047,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6895491803278688,
+      "grad_norm": 4.513894557952881,
+      "learning_rate": 4.644198188562048e-06,
+      "loss": 0.8059,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6897540983606557,
+      "grad_norm": 4.1606059074401855,
+      "learning_rate": 4.638593989710531e-06,
+      "loss": 0.9229,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6899590163934426,
+      "grad_norm": 2.806818723678589,
+      "learning_rate": 4.632992152997836e-06,
+      "loss": 0.8788,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6901639344262295,
+      "grad_norm": 3.623939037322998,
+      "learning_rate": 4.627392680892031e-06,
+      "loss": 0.924,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6903688524590164,
+      "grad_norm": 2.2685680389404297,
+      "learning_rate": 4.6217955758601426e-06,
+      "loss": 0.8098,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6905737704918032,
+      "grad_norm": 5.757803440093994,
+      "learning_rate": 4.616200840368159e-06,
+      "loss": 0.8901,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6907786885245901,
+      "grad_norm": 2.623622417449951,
+      "learning_rate": 4.610608476881012e-06,
+      "loss": 0.7097,
+      "step": 3371
+    },
+    {
+      "epoch": 0.690983606557377,
+      "grad_norm": 6.42322301864624,
+      "learning_rate": 4.605018487862594e-06,
+      "loss": 0.8523,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6911885245901639,
+      "grad_norm": 7.2730393409729,
+      "learning_rate": 4.5994308757757685e-06,
+      "loss": 0.7685,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6913934426229508,
+      "grad_norm": 4.471402645111084,
+      "learning_rate": 4.593845643082323e-06,
+      "loss": 0.9249,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6915983606557377,
+      "grad_norm": 5.4387993812561035,
+      "learning_rate": 4.588262792243013e-06,
+      "loss": 1.0997,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6918032786885245,
+      "grad_norm": 3.601532459259033,
+      "learning_rate": 4.582682325717545e-06,
+      "loss": 0.6775,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6920081967213115,
+      "grad_norm": 3.831787586212158,
+      "learning_rate": 4.577104245964569e-06,
+      "loss": 0.9144,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6922131147540984,
+      "grad_norm": 4.197061061859131,
+      "learning_rate": 4.571528555441687e-06,
+      "loss": 1.1789,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6924180327868853,
+      "grad_norm": 4.516688346862793,
+      "learning_rate": 4.5659552566054485e-06,
+      "loss": 0.8666,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6926229508196722,
+      "grad_norm": 4.549363136291504,
+      "learning_rate": 4.5603843519113525e-06,
+      "loss": 0.7002,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6928278688524591,
+      "grad_norm": 4.860085487365723,
+      "learning_rate": 4.5548158438138275e-06,
+      "loss": 0.7613,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6930327868852459,
+      "grad_norm": 2.837841033935547,
+      "learning_rate": 4.549249734766269e-06,
+      "loss": 0.8667,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6932377049180328,
+      "grad_norm": 3.6493101119995117,
+      "learning_rate": 4.543686027221001e-06,
+      "loss": 1.0278,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6934426229508197,
+      "grad_norm": 5.507307052612305,
+      "learning_rate": 4.538124723629294e-06,
+      "loss": 0.9875,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6936475409836066,
+      "grad_norm": 8.536882400512695,
+      "learning_rate": 4.532565826441354e-06,
+      "loss": 1.1602,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6938524590163935,
+      "grad_norm": 3.6537747383117676,
+      "learning_rate": 4.527009338106333e-06,
+      "loss": 1.0092,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6940573770491804,
+      "grad_norm": 6.259799957275391,
+      "learning_rate": 4.52145526107232e-06,
+      "loss": 0.8127,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6942622950819672,
+      "grad_norm": 6.623767852783203,
+      "learning_rate": 4.515903597786339e-06,
+      "loss": 1.0881,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6944672131147541,
+      "grad_norm": 2.392399549484253,
+      "learning_rate": 4.510354350694357e-06,
+      "loss": 0.8281,
+      "step": 3389
+    },
+    {
+      "epoch": 0.694672131147541,
+      "grad_norm": 2.3204548358917236,
+      "learning_rate": 4.504807522241269e-06,
+      "loss": 0.7144,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6948770491803279,
+      "grad_norm": 4.049571990966797,
+      "learning_rate": 4.499263114870908e-06,
+      "loss": 0.8357,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6950819672131148,
+      "grad_norm": 4.035634994506836,
+      "learning_rate": 4.493721131026042e-06,
+      "loss": 0.9622,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6952868852459017,
+      "grad_norm": 4.951204299926758,
+      "learning_rate": 4.488181573148367e-06,
+      "loss": 1.0531,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6954918032786885,
+      "grad_norm": 3.468881607055664,
+      "learning_rate": 4.482644443678516e-06,
+      "loss": 0.8504,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6956967213114754,
+      "grad_norm": 7.006948947906494,
+      "learning_rate": 4.477109745056043e-06,
+      "loss": 1.0203,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6959016393442623,
+      "grad_norm": 3.220390796661377,
+      "learning_rate": 4.471577479719437e-06,
+      "loss": 1.0608,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6961065573770492,
+      "grad_norm": 5.608388423919678,
+      "learning_rate": 4.466047650106114e-06,
+      "loss": 1.1779,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6963114754098361,
+      "grad_norm": 6.73259162902832,
+      "learning_rate": 4.460520258652426e-06,
+      "loss": 0.9828,
+      "step": 3398
+    },
+    {
+      "epoch": 0.696516393442623,
+      "grad_norm": 3.722245454788208,
+      "learning_rate": 4.4549953077936315e-06,
+      "loss": 1.0418,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6967213114754098,
+      "grad_norm": 4.618396282196045,
+      "learning_rate": 4.449472799963927e-06,
+      "loss": 0.9328,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6969262295081967,
+      "grad_norm": 5.5608601570129395,
+      "learning_rate": 4.443952737596433e-06,
+      "loss": 1.2187,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6971311475409836,
+      "grad_norm": 1.5482349395751953,
+      "learning_rate": 4.438435123123186e-06,
+      "loss": 0.6483,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6973360655737705,
+      "grad_norm": 8.336015701293945,
+      "learning_rate": 4.43291995897515e-06,
+      "loss": 0.8971,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6975409836065574,
+      "grad_norm": 3.4167397022247314,
+      "learning_rate": 4.427407247582205e-06,
+      "loss": 0.7552,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6977459016393442,
+      "grad_norm": 4.175859451293945,
+      "learning_rate": 4.421896991373157e-06,
+      "loss": 1.0455,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6979508196721311,
+      "grad_norm": 4.851931095123291,
+      "learning_rate": 4.416389192775714e-06,
+      "loss": 1.0103,
+      "step": 3406
+    },
+    {
+      "epoch": 0.698155737704918,
+      "grad_norm": 4.267609119415283,
+      "learning_rate": 4.410883854216524e-06,
+      "loss": 0.9759,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6983606557377049,
+      "grad_norm": 1.9007676839828491,
+      "learning_rate": 4.40538097812114e-06,
+      "loss": 0.7231,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6985655737704918,
+      "grad_norm": 2.4242665767669678,
+      "learning_rate": 4.399880566914022e-06,
+      "loss": 0.7777,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6987704918032787,
+      "grad_norm": 3.4125254154205322,
+      "learning_rate": 4.394382623018557e-06,
+      "loss": 1.0441,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6989754098360655,
+      "grad_norm": 4.25399923324585,
+      "learning_rate": 4.38888714885704e-06,
+      "loss": 1.0824,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6991803278688524,
+      "grad_norm": 4.688770294189453,
+      "learning_rate": 4.383394146850674e-06,
+      "loss": 0.9237,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6993852459016393,
+      "grad_norm": 4.4051127433776855,
+      "learning_rate": 4.377903619419583e-06,
+      "loss": 1.1715,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6995901639344262,
+      "grad_norm": 3.209798574447632,
+      "learning_rate": 4.372415568982788e-06,
+      "loss": 0.9173,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6997950819672131,
+      "grad_norm": 4.561551094055176,
+      "learning_rate": 4.366929997958235e-06,
+      "loss": 1.0969,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.142172813415527,
+      "learning_rate": 4.361446908762751e-06,
+      "loss": 1.2334,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7002049180327868,
+      "grad_norm": 4.465369701385498,
+      "learning_rate": 4.3559663038121e-06,
+      "loss": 0.9401,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7004098360655737,
+      "grad_norm": 4.359439373016357,
+      "learning_rate": 4.3504881855209345e-06,
+      "loss": 0.8918,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7006147540983606,
+      "grad_norm": 4.861044406890869,
+      "learning_rate": 4.345012556302818e-06,
+      "loss": 0.7598,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7008196721311475,
+      "grad_norm": 4.774158000946045,
+      "learning_rate": 4.339539418570204e-06,
+      "loss": 0.8868,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7010245901639345,
+      "grad_norm": 6.154555797576904,
+      "learning_rate": 4.334068774734467e-06,
+      "loss": 0.9736,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7012295081967214,
+      "grad_norm": 3.0548150539398193,
+      "learning_rate": 4.32860062720587e-06,
+      "loss": 1.0664,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7014344262295082,
+      "grad_norm": 1.968212366104126,
+      "learning_rate": 4.32313497839358e-06,
+      "loss": 0.5991,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7016393442622951,
+      "grad_norm": 2.997602701187134,
+      "learning_rate": 4.317671830705667e-06,
+      "loss": 0.8626,
+      "step": 3424
+    },
+    {
+      "epoch": 0.701844262295082,
+      "grad_norm": 3.8510773181915283,
+      "learning_rate": 4.312211186549092e-06,
+      "loss": 1.1219,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7020491803278689,
+      "grad_norm": 8.716272354125977,
+      "learning_rate": 4.3067530483297185e-06,
+      "loss": 1.1124,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7022540983606558,
+      "grad_norm": 2.8948864936828613,
+      "learning_rate": 4.3012974184523024e-06,
+      "loss": 1.0315,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7024590163934427,
+      "grad_norm": 3.36983323097229,
+      "learning_rate": 4.295844299320496e-06,
+      "loss": 0.9521,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7026639344262295,
+      "grad_norm": 4.547637462615967,
+      "learning_rate": 4.290393693336849e-06,
+      "loss": 0.8854,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7028688524590164,
+      "grad_norm": 2.982128381729126,
+      "learning_rate": 4.284945602902793e-06,
+      "loss": 0.9047,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7030737704918033,
+      "grad_norm": 3.8159985542297363,
+      "learning_rate": 4.279500030418659e-06,
+      "loss": 1.0903,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7032786885245902,
+      "grad_norm": 5.873689651489258,
+      "learning_rate": 4.274056978283675e-06,
+      "loss": 0.8987,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7034836065573771,
+      "grad_norm": 9.128667831420898,
+      "learning_rate": 4.26861644889595e-06,
+      "loss": 0.9565,
+      "step": 3433
+    },
+    {
+      "epoch": 0.703688524590164,
+      "grad_norm": 3.4027345180511475,
+      "learning_rate": 4.263178444652478e-06,
+      "loss": 0.9128,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7038934426229508,
+      "grad_norm": 3.311066150665283,
+      "learning_rate": 4.257742967949149e-06,
+      "loss": 1.0114,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7040983606557377,
+      "grad_norm": 5.241128444671631,
+      "learning_rate": 4.252310021180735e-06,
+      "loss": 1.1008,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7043032786885246,
+      "grad_norm": 4.021668910980225,
+      "learning_rate": 4.246879606740895e-06,
+      "loss": 0.8982,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7045081967213115,
+      "grad_norm": 4.678068161010742,
+      "learning_rate": 4.241451727022172e-06,
+      "loss": 1.0108,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7047131147540984,
+      "grad_norm": 3.279818296432495,
+      "learning_rate": 4.236026384415992e-06,
+      "loss": 0.7521,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7049180327868853,
+      "grad_norm": 3.5209176540374756,
+      "learning_rate": 4.230603581312665e-06,
+      "loss": 1.1333,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7051229508196721,
+      "grad_norm": 4.465606689453125,
+      "learning_rate": 4.225183320101372e-06,
+      "loss": 0.7062,
+      "step": 3441
+    },
+    {
+      "epoch": 0.705327868852459,
+      "grad_norm": 4.908018112182617,
+      "learning_rate": 4.2197656031701914e-06,
+      "loss": 1.0008,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7055327868852459,
+      "grad_norm": 8.64586067199707,
+      "learning_rate": 4.2143504329060724e-06,
+      "loss": 1.1642,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7057377049180328,
+      "grad_norm": 5.942392349243164,
+      "learning_rate": 4.208937811694834e-06,
+      "loss": 1.0603,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7059426229508197,
+      "grad_norm": 4.1926798820495605,
+      "learning_rate": 4.2035277419211824e-06,
+      "loss": 1.0806,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7061475409836065,
+      "grad_norm": 3.6344411373138428,
+      "learning_rate": 4.1981202259686996e-06,
+      "loss": 0.9099,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7063524590163934,
+      "grad_norm": 3.7397875785827637,
+      "learning_rate": 4.192715266219835e-06,
+      "loss": 1.0775,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7065573770491803,
+      "grad_norm": 4.4500813484191895,
+      "learning_rate": 4.18731286505592e-06,
+      "loss": 1.0408,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7067622950819672,
+      "grad_norm": 2.150831937789917,
+      "learning_rate": 4.181913024857154e-06,
+      "loss": 0.8162,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7069672131147541,
+      "grad_norm": 6.67506742477417,
+      "learning_rate": 4.176515748002613e-06,
+      "loss": 0.8991,
+      "step": 3450
+    },
+    {
+      "epoch": 0.707172131147541,
+      "grad_norm": 3.314061403274536,
+      "learning_rate": 4.171121036870228e-06,
+      "loss": 0.9814,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7073770491803278,
+      "grad_norm": 2.85540509223938,
+      "learning_rate": 4.165728893836825e-06,
+      "loss": 0.7627,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7075819672131147,
+      "grad_norm": 4.731251239776611,
+      "learning_rate": 4.160339321278077e-06,
+      "loss": 0.8253,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7077868852459016,
+      "grad_norm": 3.4687416553497314,
+      "learning_rate": 4.1549523215685406e-06,
+      "loss": 1.0231,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7079918032786885,
+      "grad_norm": 3.839430093765259,
+      "learning_rate": 4.149567897081623e-06,
+      "loss": 0.818,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7081967213114754,
+      "grad_norm": 5.10487174987793,
+      "learning_rate": 4.1441860501896e-06,
+      "loss": 1.083,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7084016393442623,
+      "grad_norm": 4.309447765350342,
+      "learning_rate": 4.138806783263633e-06,
+      "loss": 0.9396,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7086065573770491,
+      "grad_norm": 3.794835329055786,
+      "learning_rate": 4.1334300986737146e-06,
+      "loss": 0.8734,
+      "step": 3458
+    },
+    {
+      "epoch": 0.708811475409836,
+      "grad_norm": 3.8589835166931152,
+      "learning_rate": 4.128055998788722e-06,
+      "loss": 1.2249,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7090163934426229,
+      "grad_norm": 3.4655635356903076,
+      "learning_rate": 4.122684485976384e-06,
+      "loss": 0.8685,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7092213114754098,
+      "grad_norm": 2.5611777305603027,
+      "learning_rate": 4.117315562603295e-06,
+      "loss": 0.8712,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7094262295081967,
+      "grad_norm": 3.849532127380371,
+      "learning_rate": 4.111949231034903e-06,
+      "loss": 1.0914,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7096311475409836,
+      "grad_norm": 5.766934394836426,
+      "learning_rate": 4.106585493635519e-06,
+      "loss": 0.9782,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7098360655737705,
+      "grad_norm": 2.1816649436950684,
+      "learning_rate": 4.101224352768312e-06,
+      "loss": 0.9032,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7100409836065574,
+      "grad_norm": 5.015179634094238,
+      "learning_rate": 4.095865810795297e-06,
+      "loss": 1.1011,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7102459016393443,
+      "grad_norm": 7.834405899047852,
+      "learning_rate": 4.090509870077351e-06,
+      "loss": 1.2149,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7104508196721312,
+      "grad_norm": 3.2510979175567627,
+      "learning_rate": 4.085156532974211e-06,
+      "loss": 0.877,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7106557377049181,
+      "grad_norm": 4.802590847015381,
+      "learning_rate": 4.079805801844463e-06,
+      "loss": 0.9551,
+      "step": 3468
+    },
+    {
+      "epoch": 0.710860655737705,
+      "grad_norm": 7.233163833618164,
+      "learning_rate": 4.074457679045534e-06,
+      "loss": 0.8843,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7110655737704918,
+      "grad_norm": 3.6779117584228516,
+      "learning_rate": 4.069112166933713e-06,
+      "loss": 0.8308,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7112704918032787,
+      "grad_norm": 3.0955448150634766,
+      "learning_rate": 4.063769267864138e-06,
+      "loss": 0.9339,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7114754098360656,
+      "grad_norm": 4.939663410186768,
+      "learning_rate": 4.058428984190794e-06,
+      "loss": 1.0828,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7116803278688525,
+      "grad_norm": 3.6742470264434814,
+      "learning_rate": 4.053091318266512e-06,
+      "loss": 0.8266,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7118852459016394,
+      "grad_norm": 3.1168200969696045,
+      "learning_rate": 4.047756272442974e-06,
+      "loss": 1.2086,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7120901639344263,
+      "grad_norm": 2.777696132659912,
+      "learning_rate": 4.042423849070707e-06,
+      "loss": 0.8927,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7122950819672131,
+      "grad_norm": 2.4502737522125244,
+      "learning_rate": 4.037094050499069e-06,
+      "loss": 0.9311,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 3.1885569095611572,
+      "learning_rate": 4.031766879076285e-06,
+      "loss": 0.8419,
+      "step": 3477
+    },
+    {
+      "epoch": 0.7127049180327869,
+      "grad_norm": 2.6134300231933594,
+      "learning_rate": 4.026442337149412e-06,
+      "loss": 0.8809,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7129098360655738,
+      "grad_norm": 7.858091354370117,
+      "learning_rate": 4.021120427064336e-06,
+      "loss": 1.1516,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7131147540983607,
+      "grad_norm": 3.4918994903564453,
+      "learning_rate": 4.0158011511658035e-06,
+      "loss": 0.9058,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7133196721311476,
+      "grad_norm": 3.8337833881378174,
+      "learning_rate": 4.010484511797384e-06,
+      "loss": 0.9983,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7135245901639344,
+      "grad_norm": 4.433988571166992,
+      "learning_rate": 4.005170511301505e-06,
+      "loss": 1.0854,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7137295081967213,
+      "grad_norm": 3.7134053707122803,
+      "learning_rate": 3.9998591520194086e-06,
+      "loss": 0.8065,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7139344262295082,
+      "grad_norm": 4.000126361846924,
+      "learning_rate": 3.994550436291188e-06,
+      "loss": 0.8325,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7141393442622951,
+      "grad_norm": 2.9750845432281494,
+      "learning_rate": 3.9892443664557665e-06,
+      "loss": 0.9678,
+      "step": 3485
+    },
+    {
+      "epoch": 0.714344262295082,
+      "grad_norm": 3.6671557426452637,
+      "learning_rate": 3.9839409448509034e-06,
+      "loss": 0.8366,
+      "step": 3486
+    },
+    {
+      "epoch": 0.7145491803278688,
+      "grad_norm": 3.676506280899048,
+      "learning_rate": 3.978640173813191e-06,
+      "loss": 1.0209,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7147540983606557,
+      "grad_norm": 4.182314395904541,
+      "learning_rate": 3.973342055678052e-06,
+      "loss": 0.8883,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7149590163934426,
+      "grad_norm": 3.4732611179351807,
+      "learning_rate": 3.968046592779748e-06,
+      "loss": 0.9577,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7151639344262295,
+      "grad_norm": 2.417329788208008,
+      "learning_rate": 3.962753787451353e-06,
+      "loss": 0.8713,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7153688524590164,
+      "grad_norm": 2.968392848968506,
+      "learning_rate": 3.9574636420247866e-06,
+      "loss": 0.8281,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7155737704918033,
+      "grad_norm": 3.449378490447998,
+      "learning_rate": 3.9521761588308e-06,
+      "loss": 1.1095,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7157786885245901,
+      "grad_norm": 4.289618015289307,
+      "learning_rate": 3.946891340198951e-06,
+      "loss": 0.9798,
+      "step": 3493
+    },
+    {
+      "epoch": 0.715983606557377,
+      "grad_norm": 3.0992305278778076,
+      "learning_rate": 3.9416091884576404e-06,
+      "loss": 0.9668,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7161885245901639,
+      "grad_norm": 3.1949141025543213,
+      "learning_rate": 3.936329705934089e-06,
+      "loss": 0.8293,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7163934426229508,
+      "grad_norm": 3.309481382369995,
+      "learning_rate": 3.93105289495434e-06,
+      "loss": 0.8111,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7165983606557377,
+      "grad_norm": 3.119929075241089,
+      "learning_rate": 3.925778757843264e-06,
+      "loss": 0.7711,
+      "step": 3497
+    },
+    {
+      "epoch": 0.7168032786885246,
+      "grad_norm": 3.3462674617767334,
+      "learning_rate": 3.920507296924549e-06,
+      "loss": 0.8329,
+      "step": 3498
+    },
+    {
+      "epoch": 0.7170081967213114,
+      "grad_norm": 4.048936367034912,
+      "learning_rate": 3.915238514520709e-06,
+      "loss": 0.8661,
+      "step": 3499
+    },
+    {
+      "epoch": 0.7172131147540983,
+      "grad_norm": 3.9215407371520996,
+      "learning_rate": 3.909972412953069e-06,
+      "loss": 0.993,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7174180327868852,
+      "grad_norm": 4.349125862121582,
+      "learning_rate": 3.904708994541776e-06,
+      "loss": 0.8973,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7176229508196721,
+      "grad_norm": 4.706342697143555,
+      "learning_rate": 3.899448261605806e-06,
+      "loss": 1.1944,
+      "step": 3502
+    },
+    {
+      "epoch": 0.717827868852459,
+      "grad_norm": 3.5449352264404297,
+      "learning_rate": 3.894190216462943e-06,
+      "loss": 0.8724,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7180327868852459,
+      "grad_norm": 5.444278717041016,
+      "learning_rate": 3.88893486142978e-06,
+      "loss": 0.8707,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7182377049180327,
+      "grad_norm": 6.979868412017822,
+      "learning_rate": 3.8836821988217335e-06,
+      "loss": 1.3854,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7184426229508196,
+      "grad_norm": 4.223506450653076,
+      "learning_rate": 3.8784322309530334e-06,
+      "loss": 0.9955,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7186475409836065,
+      "grad_norm": 2.2630980014801025,
+      "learning_rate": 3.873184960136721e-06,
+      "loss": 0.886,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7188524590163935,
+      "grad_norm": 4.874511241912842,
+      "learning_rate": 3.867940388684648e-06,
+      "loss": 0.9482,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7190573770491804,
+      "grad_norm": 2.9956867694854736,
+      "learning_rate": 3.862698518907479e-06,
+      "loss": 0.8238,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7192622950819673,
+      "grad_norm": 3.6096322536468506,
+      "learning_rate": 3.8574593531146884e-06,
+      "loss": 0.6804,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7194672131147541,
+      "grad_norm": 3.3165712356567383,
+      "learning_rate": 3.852222893614556e-06,
+      "loss": 0.9625,
+      "step": 3511
+    },
+    {
+      "epoch": 0.719672131147541,
+      "grad_norm": 8.006746292114258,
+      "learning_rate": 3.846989142714174e-06,
+      "loss": 1.277,
+      "step": 3512
+    },
+    {
+      "epoch": 0.7198770491803279,
+      "grad_norm": 5.505895137786865,
+      "learning_rate": 3.8417581027194385e-06,
+      "loss": 0.8185,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7200819672131148,
+      "grad_norm": 2.644759178161621,
+      "learning_rate": 3.836529775935047e-06,
+      "loss": 0.9271,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7202868852459017,
+      "grad_norm": 2.5091047286987305,
+      "learning_rate": 3.8313041646645095e-06,
+      "loss": 0.9912,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7204918032786886,
+      "grad_norm": 3.733278751373291,
+      "learning_rate": 3.826081271210131e-06,
+      "loss": 1.0759,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7206967213114754,
+      "grad_norm": 3.7579758167266846,
+      "learning_rate": 3.820861097873038e-06,
+      "loss": 1.0213,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7209016393442623,
+      "grad_norm": 3.774646282196045,
+      "learning_rate": 3.81564364695313e-06,
+      "loss": 0.9061,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7211065573770492,
+      "grad_norm": 3.8422820568084717,
+      "learning_rate": 3.8104289207491285e-06,
+      "loss": 1.0211,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 4.063271999359131,
+      "learning_rate": 3.8052169215585465e-06,
+      "loss": 0.9769,
+      "step": 3520
+    },
+    {
+      "epoch": 0.721516393442623,
+      "grad_norm": 3.7554428577423096,
+      "learning_rate": 3.800007651677696e-06,
+      "loss": 1.0239,
+      "step": 3521
+    },
+    {
+      "epoch": 0.7217213114754099,
+      "grad_norm": 4.255080699920654,
+      "learning_rate": 3.7948011134016903e-06,
+      "loss": 1.1343,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7219262295081967,
+      "grad_norm": 2.852350950241089,
+      "learning_rate": 3.789597309024433e-06,
+      "loss": 1.0242,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7221311475409836,
+      "grad_norm": 2.584596872329712,
+      "learning_rate": 3.7843962408386325e-06,
+      "loss": 0.8277,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7223360655737705,
+      "grad_norm": 3.6651875972747803,
+      "learning_rate": 3.7791979111357767e-06,
+      "loss": 0.9662,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7225409836065574,
+      "grad_norm": 5.890571117401123,
+      "learning_rate": 3.774002322206157e-06,
+      "loss": 0.7369,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7227459016393443,
+      "grad_norm": 3.265932559967041,
+      "learning_rate": 3.7688094763388673e-06,
+      "loss": 0.946,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7229508196721312,
+      "grad_norm": 2.779294490814209,
+      "learning_rate": 3.7636193758217697e-06,
+      "loss": 0.7922,
+      "step": 3528
+    },
+    {
+      "epoch": 0.723155737704918,
+      "grad_norm": 3.015928030014038,
+      "learning_rate": 3.7584320229415337e-06,
+      "loss": 0.7499,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7233606557377049,
+      "grad_norm": 6.146920680999756,
+      "learning_rate": 3.7532474199836133e-06,
+      "loss": 0.7785,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7235655737704918,
+      "grad_norm": 4.349066257476807,
+      "learning_rate": 3.748065569232251e-06,
+      "loss": 1.1942,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7237704918032787,
+      "grad_norm": 3.5740301609039307,
+      "learning_rate": 3.742886472970476e-06,
+      "loss": 0.9753,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7239754098360656,
+      "grad_norm": 3.6038496494293213,
+      "learning_rate": 3.737710133480106e-06,
+      "loss": 0.7812,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7241803278688524,
+      "grad_norm": 2.246250867843628,
+      "learning_rate": 3.7325365530417458e-06,
+      "loss": 0.6901,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7243852459016393,
+      "grad_norm": 3.6907236576080322,
+      "learning_rate": 3.72736573393477e-06,
+      "loss": 0.7952,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7245901639344262,
+      "grad_norm": 4.750281810760498,
+      "learning_rate": 3.7221976784373616e-06,
+      "loss": 0.8882,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7247950819672131,
+      "grad_norm": 6.512576580047607,
+      "learning_rate": 3.717032388826466e-06,
+      "loss": 0.8715,
+      "step": 3537
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 4.610989093780518,
+      "learning_rate": 3.711869867377822e-06,
+      "loss": 0.7684,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7252049180327869,
+      "grad_norm": 4.393749237060547,
+      "learning_rate": 3.7067101163659367e-06,
+      "loss": 0.8917,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7254098360655737,
+      "grad_norm": 5.2195048332214355,
+      "learning_rate": 3.7015531380641055e-06,
+      "loss": 1.3171,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7256147540983606,
+      "grad_norm": 3.1422691345214844,
+      "learning_rate": 3.696398934744402e-06,
+      "loss": 0.8854,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7258196721311475,
+      "grad_norm": 2.6899187564849854,
+      "learning_rate": 3.6912475086776732e-06,
+      "loss": 0.7459,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7260245901639344,
+      "grad_norm": 2.9840753078460693,
+      "learning_rate": 3.686098862133546e-06,
+      "loss": 0.6805,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7262295081967213,
+      "grad_norm": 5.3139119148254395,
+      "learning_rate": 3.680952997380419e-06,
+      "loss": 0.9189,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7264344262295082,
+      "grad_norm": 4.0991082191467285,
+      "learning_rate": 3.675809916685471e-06,
+      "loss": 1.2434,
+      "step": 3545
+    },
+    {
+      "epoch": 0.726639344262295,
+      "grad_norm": 2.6715617179870605,
+      "learning_rate": 3.6706696223146477e-06,
+      "loss": 1.0011,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7268442622950819,
+      "grad_norm": 3.7107889652252197,
+      "learning_rate": 3.665532116532672e-06,
+      "loss": 0.7769,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7270491803278688,
+      "grad_norm": 5.154221057891846,
+      "learning_rate": 3.6603974016030387e-06,
+      "loss": 0.7866,
+      "step": 3548
+    },
+    {
+      "epoch": 0.7272540983606557,
+      "grad_norm": 4.123466968536377,
+      "learning_rate": 3.6552654797880026e-06,
+      "loss": 0.9729,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7274590163934426,
+      "grad_norm": 3.548565626144409,
+      "learning_rate": 3.6501363533486024e-06,
+      "loss": 1.0137,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7276639344262295,
+      "grad_norm": 3.279600143432617,
+      "learning_rate": 3.645010024544633e-06,
+      "loss": 1.0103,
+      "step": 3551
+    },
+    {
+      "epoch": 0.7278688524590164,
+      "grad_norm": 3.306887149810791,
+      "learning_rate": 3.6398864956346724e-06,
+      "loss": 0.9502,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7280737704918033,
+      "grad_norm": 4.213255405426025,
+      "learning_rate": 3.6347657688760475e-06,
+      "loss": 1.0707,
+      "step": 3553
+    },
+    {
+      "epoch": 0.7282786885245902,
+      "grad_norm": 2.2007360458374023,
+      "learning_rate": 3.6296478465248585e-06,
+      "loss": 0.871,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7284836065573771,
+      "grad_norm": 3.099226236343384,
+      "learning_rate": 3.6245327308359712e-06,
+      "loss": 1.0055,
+      "step": 3555
+    },
+    {
+      "epoch": 0.728688524590164,
+      "grad_norm": 2.300431489944458,
+      "learning_rate": 3.6194204240630137e-06,
+      "loss": 0.6237,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7288934426229509,
+      "grad_norm": 3.5555760860443115,
+      "learning_rate": 3.6143109284583754e-06,
+      "loss": 0.6788,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7290983606557377,
+      "grad_norm": 3.84749436378479,
+      "learning_rate": 3.609204246273208e-06,
+      "loss": 0.6166,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7293032786885246,
+      "grad_norm": 3.6451807022094727,
+      "learning_rate": 3.604100379757428e-06,
+      "loss": 0.9959,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7295081967213115,
+      "grad_norm": 4.527796268463135,
+      "learning_rate": 3.598999331159695e-06,
+      "loss": 0.793,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7297131147540984,
+      "grad_norm": 7.602413654327393,
+      "learning_rate": 3.5939011027274505e-06,
+      "loss": 0.9896,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7299180327868853,
+      "grad_norm": 3.8810815811157227,
+      "learning_rate": 3.5888056967068818e-06,
+      "loss": 1.0777,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7301229508196722,
+      "grad_norm": 2.8097410202026367,
+      "learning_rate": 3.5837131153429263e-06,
+      "loss": 0.6782,
+      "step": 3563
+    },
+    {
+      "epoch": 0.730327868852459,
+      "grad_norm": 2.4118943214416504,
+      "learning_rate": 3.5786233608792862e-06,
+      "loss": 0.9123,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7305327868852459,
+      "grad_norm": 4.875748634338379,
+      "learning_rate": 3.573536435558417e-06,
+      "loss": 0.8758,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7307377049180328,
+      "grad_norm": 2.1782431602478027,
+      "learning_rate": 3.5684523416215244e-06,
+      "loss": 0.7857,
+      "step": 3566
+    },
+    {
+      "epoch": 0.7309426229508197,
+      "grad_norm": 4.6156768798828125,
+      "learning_rate": 3.5633710813085687e-06,
+      "loss": 0.6321,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7311475409836066,
+      "grad_norm": 10.132577896118164,
+      "learning_rate": 3.558292656858264e-06,
+      "loss": 0.8593,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7313524590163935,
+      "grad_norm": 4.06773567199707,
+      "learning_rate": 3.5532170705080737e-06,
+      "loss": 0.7267,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7315573770491803,
+      "grad_norm": 2.950862407684326,
+      "learning_rate": 3.5481443244941995e-06,
+      "loss": 0.6894,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7317622950819672,
+      "grad_norm": 4.610589504241943,
+      "learning_rate": 3.5430744210516134e-06,
+      "loss": 0.9268,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7319672131147541,
+      "grad_norm": 4.820957660675049,
+      "learning_rate": 3.5380073624140175e-06,
+      "loss": 1.231,
+      "step": 3572
+    },
+    {
+      "epoch": 0.732172131147541,
+      "grad_norm": 7.148326396942139,
+      "learning_rate": 3.5329431508138734e-06,
+      "loss": 1.1337,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7323770491803279,
+      "grad_norm": 2.6751596927642822,
+      "learning_rate": 3.5278817884823723e-06,
+      "loss": 0.9784,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7325819672131147,
+      "grad_norm": 3.4021549224853516,
+      "learning_rate": 3.522823277649463e-06,
+      "loss": 1.0709,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7327868852459016,
+      "grad_norm": 4.2269978523254395,
+      "learning_rate": 3.517767620543834e-06,
+      "loss": 1.2385,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7329918032786885,
+      "grad_norm": 2.09798264503479,
+      "learning_rate": 3.512714819392916e-06,
+      "loss": 0.7169,
+      "step": 3577
+    },
+    {
+      "epoch": 0.7331967213114754,
+      "grad_norm": 3.5467379093170166,
+      "learning_rate": 3.507664876422885e-06,
+      "loss": 1.1895,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7334016393442623,
+      "grad_norm": 3.8033673763275146,
+      "learning_rate": 3.5026177938586515e-06,
+      "loss": 1.0404,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7336065573770492,
+      "grad_norm": 6.072485446929932,
+      "learning_rate": 3.49757357392387e-06,
+      "loss": 0.8746,
+      "step": 3580
+    },
+    {
+      "epoch": 0.733811475409836,
+      "grad_norm": 2.612391233444214,
+      "learning_rate": 3.492532218840935e-06,
+      "loss": 0.6291,
+      "step": 3581
+    },
+    {
+      "epoch": 0.7340163934426229,
+      "grad_norm": 2.766044855117798,
+      "learning_rate": 3.487493730830973e-06,
+      "loss": 0.9038,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7342213114754098,
+      "grad_norm": 2.7935683727264404,
+      "learning_rate": 3.482458112113858e-06,
+      "loss": 0.653,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7344262295081967,
+      "grad_norm": 3.140185832977295,
+      "learning_rate": 3.4774253649081824e-06,
+      "loss": 0.6582,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7346311475409836,
+      "grad_norm": 3.5312159061431885,
+      "learning_rate": 3.4723954914312874e-06,
+      "loss": 0.9684,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7348360655737705,
+      "grad_norm": 3.3143465518951416,
+      "learning_rate": 3.467368493899249e-06,
+      "loss": 0.9388,
+      "step": 3586
+    },
+    {
+      "epoch": 0.7350409836065573,
+      "grad_norm": 5.525887489318848,
+      "learning_rate": 3.4623443745268724e-06,
+      "loss": 0.7871,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7352459016393442,
+      "grad_norm": 5.101330280303955,
+      "learning_rate": 3.4573231355276872e-06,
+      "loss": 0.902,
+      "step": 3588
+    },
+    {
+      "epoch": 0.7354508196721311,
+      "grad_norm": 4.8076558113098145,
+      "learning_rate": 3.452304779113964e-06,
+      "loss": 1.0696,
+      "step": 3589
+    },
+    {
+      "epoch": 0.735655737704918,
+      "grad_norm": 8.634212493896484,
+      "learning_rate": 3.447289307496701e-06,
+      "loss": 1.1323,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7358606557377049,
+      "grad_norm": 5.594690799713135,
+      "learning_rate": 3.4422767228856235e-06,
+      "loss": 0.9927,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7360655737704918,
+      "grad_norm": 4.6134352684021,
+      "learning_rate": 3.437267027489186e-06,
+      "loss": 1.0181,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7362704918032786,
+      "grad_norm": 4.904750347137451,
+      "learning_rate": 3.4322602235145708e-06,
+      "loss": 0.675,
+      "step": 3593
+    },
+    {
+      "epoch": 0.7364754098360655,
+      "grad_norm": 4.760403156280518,
+      "learning_rate": 3.4272563131676884e-06,
+      "loss": 0.7716,
+      "step": 3594
+    },
+    {
+      "epoch": 0.7366803278688525,
+      "grad_norm": 5.089571475982666,
+      "learning_rate": 3.422255298653161e-06,
+      "loss": 1.0103,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7368852459016394,
+      "grad_norm": 5.905520439147949,
+      "learning_rate": 3.417257182174356e-06,
+      "loss": 1.0958,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7370901639344263,
+      "grad_norm": 3.551583766937256,
+      "learning_rate": 3.412261965933353e-06,
+      "loss": 0.8769,
+      "step": 3597
+    },
+    {
+      "epoch": 0.7372950819672132,
+      "grad_norm": 3.9570934772491455,
+      "learning_rate": 3.407269652130948e-06,
+      "loss": 0.7112,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 4.283629894256592,
+      "learning_rate": 3.4022802429666667e-06,
+      "loss": 1.1306,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7377049180327869,
+      "grad_norm": 3.8943657875061035,
+      "learning_rate": 3.397293740638754e-06,
+      "loss": 0.9235,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7379098360655738,
+      "grad_norm": 3.1678853034973145,
+      "learning_rate": 3.3923101473441734e-06,
+      "loss": 1.1528,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7381147540983607,
+      "grad_norm": 5.226072788238525,
+      "learning_rate": 3.387329465278605e-06,
+      "loss": 1.1445,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7383196721311476,
+      "grad_norm": 3.5224220752716064,
+      "learning_rate": 3.3823516966364467e-06,
+      "loss": 1.059,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7385245901639345,
+      "grad_norm": 3.132953643798828,
+      "learning_rate": 3.3773768436108188e-06,
+      "loss": 1.0786,
+      "step": 3604
+    },
+    {
+      "epoch": 0.7387295081967213,
+      "grad_norm": 3.4377105236053467,
+      "learning_rate": 3.372404908393542e-06,
+      "loss": 0.7251,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7389344262295082,
+      "grad_norm": 2.4356954097747803,
+      "learning_rate": 3.3674358931751705e-06,
+      "loss": 0.8064,
+      "step": 3606
+    },
+    {
+      "epoch": 0.7391393442622951,
+      "grad_norm": 2.287515163421631,
+      "learning_rate": 3.3624698001449584e-06,
+      "loss": 1.0774,
+      "step": 3607
+    },
+    {
+      "epoch": 0.739344262295082,
+      "grad_norm": 5.751395225524902,
+      "learning_rate": 3.3575066314908823e-06,
+      "loss": 1.176,
+      "step": 3608
+    },
+    {
+      "epoch": 0.7395491803278689,
+      "grad_norm": 4.930312633514404,
+      "learning_rate": 3.3525463893996167e-06,
+      "loss": 0.7576,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7397540983606558,
+      "grad_norm": 3.8744585514068604,
+      "learning_rate": 3.3475890760565544e-06,
+      "loss": 1.0949,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7399590163934426,
+      "grad_norm": 5.591696739196777,
+      "learning_rate": 3.34263469364581e-06,
+      "loss": 0.8881,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7401639344262295,
+      "grad_norm": 3.32574200630188,
+      "learning_rate": 3.3376832443501838e-06,
+      "loss": 0.696,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7403688524590164,
+      "grad_norm": 3.9204230308532715,
+      "learning_rate": 3.332734730351199e-06,
+      "loss": 1.0551,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7405737704918033,
+      "grad_norm": 3.8743526935577393,
+      "learning_rate": 3.3277891538290806e-06,
+      "loss": 0.887,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7407786885245902,
+      "grad_norm": 5.2939453125,
+      "learning_rate": 3.322846516962762e-06,
+      "loss": 0.8276,
+      "step": 3615
+    },
+    {
+      "epoch": 0.740983606557377,
+      "grad_norm": 3.322377920150757,
+      "learning_rate": 3.3179068219298792e-06,
+      "loss": 0.9657,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7411885245901639,
+      "grad_norm": 4.787900447845459,
+      "learning_rate": 3.3129700709067724e-06,
+      "loss": 0.9735,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7413934426229508,
+      "grad_norm": 6.3879828453063965,
+      "learning_rate": 3.3080362660684885e-06,
+      "loss": 0.8523,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7415983606557377,
+      "grad_norm": 4.454036235809326,
+      "learning_rate": 3.3031054095887673e-06,
+      "loss": 1.2504,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7418032786885246,
+      "grad_norm": 4.158874988555908,
+      "learning_rate": 3.298177503640053e-06,
+      "loss": 0.9345,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7420081967213115,
+      "grad_norm": 9.5687255859375,
+      "learning_rate": 3.293252550393502e-06,
+      "loss": 1.0486,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7422131147540983,
+      "grad_norm": 3.680778980255127,
+      "learning_rate": 3.288330552018957e-06,
+      "loss": 1.0971,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7424180327868852,
+      "grad_norm": 2.3822028636932373,
+      "learning_rate": 3.2834115106849564e-06,
+      "loss": 1.0086,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7426229508196721,
+      "grad_norm": 5.106513023376465,
+      "learning_rate": 3.2784954285587444e-06,
+      "loss": 0.9657,
+      "step": 3624
+    },
+    {
+      "epoch": 0.742827868852459,
+      "grad_norm": 2.6511456966400146,
+      "learning_rate": 3.273582307806258e-06,
+      "loss": 1.0182,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7430327868852459,
+      "grad_norm": 3.9721217155456543,
+      "learning_rate": 3.268672150592129e-06,
+      "loss": 0.8663,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7432377049180328,
+      "grad_norm": 5.434160232543945,
+      "learning_rate": 3.263764959079686e-06,
+      "loss": 0.6752,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7434426229508196,
+      "grad_norm": 3.8235981464385986,
+      "learning_rate": 3.2588607354309465e-06,
+      "loss": 1.0705,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7436475409836065,
+      "grad_norm": 6.132665157318115,
+      "learning_rate": 3.253959481806629e-06,
+      "loss": 0.8886,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7438524590163934,
+      "grad_norm": 3.092726945877075,
+      "learning_rate": 3.249061200366126e-06,
+      "loss": 0.7776,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7440573770491803,
+      "grad_norm": 4.270736217498779,
+      "learning_rate": 3.2441658932675413e-06,
+      "loss": 0.8399,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7442622950819672,
+      "grad_norm": 3.0863876342773438,
+      "learning_rate": 3.2392735626676607e-06,
+      "loss": 1.0412,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7444672131147541,
+      "grad_norm": 3.066093921661377,
+      "learning_rate": 3.234384210721949e-06,
+      "loss": 0.7909,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7446721311475409,
+      "grad_norm": 3.0942187309265137,
+      "learning_rate": 3.22949783958457e-06,
+      "loss": 0.8143,
+      "step": 3634
+    },
+    {
+      "epoch": 0.7448770491803278,
+      "grad_norm": 2.4851644039154053,
+      "learning_rate": 3.2246144514083677e-06,
+      "loss": 0.8581,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7450819672131147,
+      "grad_norm": 4.381285190582275,
+      "learning_rate": 3.219734048344885e-06,
+      "loss": 0.8022,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7452868852459016,
+      "grad_norm": 2.681527614593506,
+      "learning_rate": 3.2148566325443285e-06,
+      "loss": 0.9992,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7454918032786885,
+      "grad_norm": 3.5137956142425537,
+      "learning_rate": 3.2099822061556053e-06,
+      "loss": 0.962,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7456967213114755,
+      "grad_norm": 3.5121119022369385,
+      "learning_rate": 3.2051107713262998e-06,
+      "loss": 1.2097,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7459016393442623,
+      "grad_norm": 4.1459760665893555,
+      "learning_rate": 3.2002423302026707e-06,
+      "loss": 0.7958,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7461065573770492,
+      "grad_norm": 5.356362819671631,
+      "learning_rate": 3.195376884929675e-06,
+      "loss": 0.8567,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7463114754098361,
+      "grad_norm": 2.8962533473968506,
+      "learning_rate": 3.1905144376509367e-06,
+      "loss": 0.7985,
+      "step": 3642
+    },
+    {
+      "epoch": 0.746516393442623,
+      "grad_norm": 4.153777122497559,
+      "learning_rate": 3.185654990508765e-06,
+      "loss": 1.1583,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7467213114754099,
+      "grad_norm": 6.335258960723877,
+      "learning_rate": 3.1807985456441404e-06,
+      "loss": 0.9107,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7469262295081968,
+      "grad_norm": 2.965120792388916,
+      "learning_rate": 3.1759451051967217e-06,
+      "loss": 0.59,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7471311475409836,
+      "grad_norm": 2.5487184524536133,
+      "learning_rate": 3.1710946713048584e-06,
+      "loss": 0.924,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7473360655737705,
+      "grad_norm": 3.2520713806152344,
+      "learning_rate": 3.166247246105556e-06,
+      "loss": 0.82,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7475409836065574,
+      "grad_norm": 2.7817578315734863,
+      "learning_rate": 3.1614028317345035e-06,
+      "loss": 1.0737,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7477459016393443,
+      "grad_norm": 4.201389312744141,
+      "learning_rate": 3.1565614303260637e-06,
+      "loss": 0.944,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7479508196721312,
+      "grad_norm": 3.0726778507232666,
+      "learning_rate": 3.1517230440132717e-06,
+      "loss": 0.9582,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7481557377049181,
+      "grad_norm": 2.5591046810150146,
+      "learning_rate": 3.146887674927831e-06,
+      "loss": 0.9317,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7483606557377049,
+      "grad_norm": 3.4032492637634277,
+      "learning_rate": 3.142055325200121e-06,
+      "loss": 0.7398,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7485655737704918,
+      "grad_norm": 4.271381378173828,
+      "learning_rate": 3.137225996959188e-06,
+      "loss": 0.7935,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7487704918032787,
+      "grad_norm": 5.654604911804199,
+      "learning_rate": 3.1323996923327417e-06,
+      "loss": 1.3687,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7489754098360656,
+      "grad_norm": 2.478200674057007,
+      "learning_rate": 3.1275764134471654e-06,
+      "loss": 0.7778,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7491803278688525,
+      "grad_norm": 5.091400146484375,
+      "learning_rate": 3.122756162427515e-06,
+      "loss": 0.9157,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7493852459016394,
+      "grad_norm": 3.9727730751037598,
+      "learning_rate": 3.1179389413975057e-06,
+      "loss": 1.1133,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7495901639344262,
+      "grad_norm": 2.1222920417785645,
+      "learning_rate": 3.113124752479513e-06,
+      "loss": 0.7914,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7497950819672131,
+      "grad_norm": 3.514667510986328,
+      "learning_rate": 3.1083135977945834e-06,
+      "loss": 1.0252,
+      "step": 3659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.038805961608887,
+      "learning_rate": 3.1035054794624263e-06,
+      "loss": 0.9294,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7502049180327869,
+      "grad_norm": 3.1217446327209473,
+      "learning_rate": 3.0987003996014107e-06,
+      "loss": 0.7342,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7504098360655738,
+      "grad_norm": 2.9484736919403076,
+      "learning_rate": 3.093898360328569e-06,
+      "loss": 1.0377,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7506147540983606,
+      "grad_norm": 5.425601959228516,
+      "learning_rate": 3.0890993637595934e-06,
+      "loss": 0.8262,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7508196721311475,
+      "grad_norm": 7.76238489151001,
+      "learning_rate": 3.084303412008838e-06,
+      "loss": 1.1249,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7510245901639344,
+      "grad_norm": 2.4006686210632324,
+      "learning_rate": 3.0795105071893017e-06,
+      "loss": 0.8381,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7512295081967213,
+      "grad_norm": 2.882462501525879,
+      "learning_rate": 3.0747206514126647e-06,
+      "loss": 0.8525,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7514344262295082,
+      "grad_norm": 3.3562543392181396,
+      "learning_rate": 3.069933846789248e-06,
+      "loss": 0.7306,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7516393442622951,
+      "grad_norm": 4.321474552154541,
+      "learning_rate": 3.0651500954280267e-06,
+      "loss": 1.0923,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7518442622950819,
+      "grad_norm": 4.36995792388916,
+      "learning_rate": 3.060369399436638e-06,
+      "loss": 0.9725,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7520491803278688,
+      "grad_norm": 3.2703349590301514,
+      "learning_rate": 3.055591760921368e-06,
+      "loss": 0.8877,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7522540983606557,
+      "grad_norm": 2.8494322299957275,
+      "learning_rate": 3.050817181987167e-06,
+      "loss": 0.8413,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7524590163934426,
+      "grad_norm": 3.085181713104248,
+      "learning_rate": 3.046045664737618e-06,
+      "loss": 0.6616,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7526639344262295,
+      "grad_norm": 10.1361722946167,
+      "learning_rate": 3.04127721127497e-06,
+      "loss": 0.832,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7528688524590164,
+      "grad_norm": 2.6421453952789307,
+      "learning_rate": 3.0365118237001166e-06,
+      "loss": 1.0104,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7530737704918032,
+      "grad_norm": 3.084125280380249,
+      "learning_rate": 3.031749504112602e-06,
+      "loss": 0.8652,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7532786885245901,
+      "grad_norm": 4.183011054992676,
+      "learning_rate": 3.02699025461062e-06,
+      "loss": 0.746,
+      "step": 3676
+    },
+    {
+      "epoch": 0.753483606557377,
+      "grad_norm": 3.7950966358184814,
+      "learning_rate": 3.0222340772910074e-06,
+      "loss": 0.8351,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7536885245901639,
+      "grad_norm": 2.9729926586151123,
+      "learning_rate": 3.017480974249255e-06,
+      "loss": 0.6085,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7538934426229508,
+      "grad_norm": 4.621546745300293,
+      "learning_rate": 3.0127309475794875e-06,
+      "loss": 1.0419,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7540983606557377,
+      "grad_norm": 2.4214813709259033,
+      "learning_rate": 3.007983999374481e-06,
+      "loss": 0.7648,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7543032786885245,
+      "grad_norm": 3.1902236938476562,
+      "learning_rate": 3.0032401317256644e-06,
+      "loss": 0.9193,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7545081967213115,
+      "grad_norm": 3.379472017288208,
+      "learning_rate": 2.998499346723092e-06,
+      "loss": 1.2393,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7547131147540984,
+      "grad_norm": 5.147739887237549,
+      "learning_rate": 2.9937616464554696e-06,
+      "loss": 0.8824,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7549180327868853,
+      "grad_norm": 4.19382381439209,
+      "learning_rate": 2.9890270330101447e-06,
+      "loss": 1.144,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7551229508196722,
+      "grad_norm": 3.5749759674072266,
+      "learning_rate": 2.9842955084730994e-06,
+      "loss": 0.954,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7553278688524591,
+      "grad_norm": 4.306379318237305,
+      "learning_rate": 2.979567074928962e-06,
+      "loss": 1.0085,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7555327868852459,
+      "grad_norm": 3.682429790496826,
+      "learning_rate": 2.9748417344609925e-06,
+      "loss": 0.7781,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7557377049180328,
+      "grad_norm": 2.7951717376708984,
+      "learning_rate": 2.970119489151094e-06,
+      "loss": 0.9584,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7559426229508197,
+      "grad_norm": 2.8467698097229004,
+      "learning_rate": 2.9654003410797984e-06,
+      "loss": 0.7832,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7561475409836066,
+      "grad_norm": 2.942859649658203,
+      "learning_rate": 2.9606842923262746e-06,
+      "loss": 0.8659,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7563524590163935,
+      "grad_norm": 3.816351890563965,
+      "learning_rate": 2.9559713449683357e-06,
+      "loss": 1.168,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7565573770491804,
+      "grad_norm": 4.745134353637695,
+      "learning_rate": 2.951261501082423e-06,
+      "loss": 1.0845,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7567622950819672,
+      "grad_norm": 3.183626890182495,
+      "learning_rate": 2.9465547627436e-06,
+      "loss": 0.7277,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7569672131147541,
+      "grad_norm": 5.162381172180176,
+      "learning_rate": 2.941851132025575e-06,
+      "loss": 1.079,
+      "step": 3694
+    },
+    {
+      "epoch": 0.757172131147541,
+      "grad_norm": 3.124962329864502,
+      "learning_rate": 2.937150611000684e-06,
+      "loss": 0.9102,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7573770491803279,
+      "grad_norm": 3.4955573081970215,
+      "learning_rate": 2.93245320173989e-06,
+      "loss": 0.9991,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7575819672131148,
+      "grad_norm": 5.562982559204102,
+      "learning_rate": 2.9277589063127876e-06,
+      "loss": 1.058,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7577868852459017,
+      "grad_norm": 4.6041789054870605,
+      "learning_rate": 2.9230677267875997e-06,
+      "loss": 1.184,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7579918032786885,
+      "grad_norm": 4.655192852020264,
+      "learning_rate": 2.9183796652311736e-06,
+      "loss": 0.9055,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7581967213114754,
+      "grad_norm": 3.3236005306243896,
+      "learning_rate": 2.9136947237089862e-06,
+      "loss": 0.8942,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7584016393442623,
+      "grad_norm": 4.0832624435424805,
+      "learning_rate": 2.9090129042851365e-06,
+      "loss": 0.9183,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7586065573770492,
+      "grad_norm": 5.257983684539795,
+      "learning_rate": 2.9043342090223513e-06,
+      "loss": 1.0204,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7588114754098361,
+      "grad_norm": 4.96785306930542,
+      "learning_rate": 2.8996586399819826e-06,
+      "loss": 0.7924,
+      "step": 3703
+    },
+    {
+      "epoch": 0.759016393442623,
+      "grad_norm": 2.733050584793091,
+      "learning_rate": 2.8949861992239947e-06,
+      "loss": 0.8437,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7592213114754098,
+      "grad_norm": 3.254187822341919,
+      "learning_rate": 2.890316888806981e-06,
+      "loss": 0.9635,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7594262295081967,
+      "grad_norm": 2.9561121463775635,
+      "learning_rate": 2.8856507107881657e-06,
+      "loss": 1.0929,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7596311475409836,
+      "grad_norm": 6.161920547485352,
+      "learning_rate": 2.8809876672233718e-06,
+      "loss": 1.0856,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7598360655737705,
+      "grad_norm": 4.2185893058776855,
+      "learning_rate": 2.876327760167056e-06,
+      "loss": 1.0275,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7600409836065574,
+      "grad_norm": 3.397260904312134,
+      "learning_rate": 2.8716709916722895e-06,
+      "loss": 1.093,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7602459016393442,
+      "grad_norm": 4.256423473358154,
+      "learning_rate": 2.8670173637907605e-06,
+      "loss": 0.9576,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7604508196721311,
+      "grad_norm": 2.987426996231079,
+      "learning_rate": 2.862366878572774e-06,
+      "loss": 0.9299,
+      "step": 3711
+    },
+    {
+      "epoch": 0.760655737704918,
+      "grad_norm": 2.9800913333892822,
+      "learning_rate": 2.8577195380672495e-06,
+      "loss": 0.9069,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7608606557377049,
+      "grad_norm": 4.05445671081543,
+      "learning_rate": 2.853075344321724e-06,
+      "loss": 0.9558,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7610655737704918,
+      "grad_norm": 2.9575464725494385,
+      "learning_rate": 2.84843429938234e-06,
+      "loss": 0.888,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7612704918032787,
+      "grad_norm": 5.627894878387451,
+      "learning_rate": 2.8437964052938584e-06,
+      "loss": 0.7974,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7614754098360655,
+      "grad_norm": 3.4361989498138428,
+      "learning_rate": 2.8391616640996565e-06,
+      "loss": 0.9057,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7616803278688524,
+      "grad_norm": 2.1844546794891357,
+      "learning_rate": 2.8345300778417194e-06,
+      "loss": 0.8936,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7618852459016393,
+      "grad_norm": 5.375782012939453,
+      "learning_rate": 2.8299016485606336e-06,
+      "loss": 0.8715,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7620901639344262,
+      "grad_norm": 2.64579439163208,
+      "learning_rate": 2.825276378295605e-06,
+      "loss": 0.9836,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7622950819672131,
+      "grad_norm": 2.239075183868408,
+      "learning_rate": 2.820654269084444e-06,
+      "loss": 0.7311,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7625,
+      "grad_norm": 3.7062582969665527,
+      "learning_rate": 2.8160353229635684e-06,
+      "loss": 1.0253,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7627049180327868,
+      "grad_norm": 4.36362361907959,
+      "learning_rate": 2.8114195419680023e-06,
+      "loss": 0.8926,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7629098360655737,
+      "grad_norm": 3.0431578159332275,
+      "learning_rate": 2.806806928131377e-06,
+      "loss": 0.8336,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7631147540983606,
+      "grad_norm": 2.6580727100372314,
+      "learning_rate": 2.8021974834859258e-06,
+      "loss": 0.7992,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7633196721311475,
+      "grad_norm": 5.55933141708374,
+      "learning_rate": 2.797591210062487e-06,
+      "loss": 1.0576,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7635245901639345,
+      "grad_norm": 5.112779140472412,
+      "learning_rate": 2.792988109890502e-06,
+      "loss": 1.0825,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7637295081967214,
+      "grad_norm": 2.980485677719116,
+      "learning_rate": 2.7883881849980168e-06,
+      "loss": 0.7946,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7639344262295082,
+      "grad_norm": 2.784369707107544,
+      "learning_rate": 2.7837914374116683e-06,
+      "loss": 0.9603,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7641393442622951,
+      "grad_norm": 11.422462463378906,
+      "learning_rate": 2.7791978691567047e-06,
+      "loss": 1.0354,
+      "step": 3729
+    },
+    {
+      "epoch": 0.764344262295082,
+      "grad_norm": 2.7553253173828125,
+      "learning_rate": 2.774607482256967e-06,
+      "loss": 0.7722,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7645491803278689,
+      "grad_norm": 2.4345507621765137,
+      "learning_rate": 2.770020278734904e-06,
+      "loss": 0.6792,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7647540983606558,
+      "grad_norm": 3.063055992126465,
+      "learning_rate": 2.765436260611547e-06,
+      "loss": 0.9041,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7649590163934427,
+      "grad_norm": 3.163034200668335,
+      "learning_rate": 2.760855429906535e-06,
+      "loss": 1.0024,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7651639344262295,
+      "grad_norm": 3.638094186782837,
+      "learning_rate": 2.756277788638099e-06,
+      "loss": 0.9527,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7653688524590164,
+      "grad_norm": 4.399129867553711,
+      "learning_rate": 2.751703338823064e-06,
+      "loss": 0.9344,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7655737704918033,
+      "grad_norm": 3.6983590126037598,
+      "learning_rate": 2.747132082476852e-06,
+      "loss": 1.1025,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7657786885245902,
+      "grad_norm": 2.8665518760681152,
+      "learning_rate": 2.742564021613475e-06,
+      "loss": 0.6459,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7659836065573771,
+      "grad_norm": 7.115226745605469,
+      "learning_rate": 2.7379991582455434e-06,
+      "loss": 0.8645,
+      "step": 3738
+    },
+    {
+      "epoch": 0.766188524590164,
+      "grad_norm": 5.899470329284668,
+      "learning_rate": 2.7334374943842444e-06,
+      "loss": 1.1772,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7663934426229508,
+      "grad_norm": 2.9014787673950195,
+      "learning_rate": 2.7288790320393667e-06,
+      "loss": 0.862,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7665983606557377,
+      "grad_norm": 3.023167371749878,
+      "learning_rate": 2.724323773219296e-06,
+      "loss": 0.7721,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7668032786885246,
+      "grad_norm": 4.696282386779785,
+      "learning_rate": 2.719771719930987e-06,
+      "loss": 1.052,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7670081967213115,
+      "grad_norm": 2.121035575866699,
+      "learning_rate": 2.715222874179996e-06,
+      "loss": 0.8221,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7672131147540984,
+      "grad_norm": 2.6120052337646484,
+      "learning_rate": 2.7106772379704614e-06,
+      "loss": 0.9444,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7674180327868853,
+      "grad_norm": 4.549228668212891,
+      "learning_rate": 2.706134813305111e-06,
+      "loss": 1.0182,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7676229508196721,
+      "grad_norm": 4.172486305236816,
+      "learning_rate": 2.701595602185253e-06,
+      "loss": 1.0342,
+      "step": 3746
+    },
+    {
+      "epoch": 0.767827868852459,
+      "grad_norm": 5.530179500579834,
+      "learning_rate": 2.6970596066107825e-06,
+      "loss": 0.8392,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7680327868852459,
+      "grad_norm": 5.9540276527404785,
+      "learning_rate": 2.6925268285801807e-06,
+      "loss": 0.9777,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7682377049180328,
+      "grad_norm": 3.039767265319824,
+      "learning_rate": 2.6879972700904987e-06,
+      "loss": 1.0916,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7684426229508197,
+      "grad_norm": 2.485373020172119,
+      "learning_rate": 2.6834709331373863e-06,
+      "loss": 0.8204,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7686475409836065,
+      "grad_norm": 3.0241713523864746,
+      "learning_rate": 2.6789478197150654e-06,
+      "loss": 0.9283,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7688524590163934,
+      "grad_norm": 3.591670513153076,
+      "learning_rate": 2.674427931816339e-06,
+      "loss": 0.9155,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7690573770491803,
+      "grad_norm": 2.115086078643799,
+      "learning_rate": 2.6699112714325827e-06,
+      "loss": 0.7591,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7692622950819672,
+      "grad_norm": 2.4178004264831543,
+      "learning_rate": 2.6653978405537594e-06,
+      "loss": 0.7008,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7694672131147541,
+      "grad_norm": 3.9129247665405273,
+      "learning_rate": 2.6608876411684047e-06,
+      "loss": 0.9459,
+      "step": 3755
+    },
+    {
+      "epoch": 0.769672131147541,
+      "grad_norm": 4.924602031707764,
+      "learning_rate": 2.6563806752636314e-06,
+      "loss": 1.0484,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7698770491803278,
+      "grad_norm": 4.134285926818848,
+      "learning_rate": 2.6518769448251267e-06,
+      "loss": 1.1633,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7700819672131147,
+      "grad_norm": 4.355528354644775,
+      "learning_rate": 2.647376451837155e-06,
+      "loss": 1.0932,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7702868852459016,
+      "grad_norm": 5.044935703277588,
+      "learning_rate": 2.6428791982825497e-06,
+      "loss": 0.9786,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7704918032786885,
+      "grad_norm": 3.925523042678833,
+      "learning_rate": 2.638385186142721e-06,
+      "loss": 0.8186,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7706967213114754,
+      "grad_norm": 2.988187551498413,
+      "learning_rate": 2.63389441739765e-06,
+      "loss": 0.926,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7709016393442623,
+      "grad_norm": 6.291748046875,
+      "learning_rate": 2.629406894025891e-06,
+      "loss": 1.0998,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7711065573770491,
+      "grad_norm": 27.340669631958008,
+      "learning_rate": 2.6249226180045606e-06,
+      "loss": 1.2157,
+      "step": 3763
+    },
+    {
+      "epoch": 0.771311475409836,
+      "grad_norm": 2.816814661026001,
+      "learning_rate": 2.6204415913093505e-06,
+      "loss": 0.8851,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7715163934426229,
+      "grad_norm": 4.476043701171875,
+      "learning_rate": 2.6159638159145205e-06,
+      "loss": 1.2222,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7717213114754098,
+      "grad_norm": 4.753095626831055,
+      "learning_rate": 2.611489293792905e-06,
+      "loss": 0.9345,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7719262295081967,
+      "grad_norm": 2.437514066696167,
+      "learning_rate": 2.6070180269158884e-06,
+      "loss": 0.9928,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7721311475409836,
+      "grad_norm": 3.837632417678833,
+      "learning_rate": 2.6025500172534345e-06,
+      "loss": 0.8837,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7723360655737705,
+      "grad_norm": 2.8172779083251953,
+      "learning_rate": 2.5980852667740664e-06,
+      "loss": 0.9472,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7725409836065574,
+      "grad_norm": 4.333314418792725,
+      "learning_rate": 2.5936237774448738e-06,
+      "loss": 0.8214,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7727459016393443,
+      "grad_norm": 3.7686357498168945,
+      "learning_rate": 2.5891655512315083e-06,
+      "loss": 0.8735,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7729508196721312,
+      "grad_norm": 4.322832107543945,
+      "learning_rate": 2.584710590098184e-06,
+      "loss": 0.8708,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7731557377049181,
+      "grad_norm": 3.5747101306915283,
+      "learning_rate": 2.580258896007678e-06,
+      "loss": 0.9519,
+      "step": 3773
+    },
+    {
+      "epoch": 0.773360655737705,
+      "grad_norm": 7.5024638175964355,
+      "learning_rate": 2.5758104709213195e-06,
+      "loss": 0.7127,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7735655737704918,
+      "grad_norm": 2.6025338172912598,
+      "learning_rate": 2.5713653167990126e-06,
+      "loss": 0.8046,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7737704918032787,
+      "grad_norm": 4.726764678955078,
+      "learning_rate": 2.566923435599212e-06,
+      "loss": 1.0116,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7739754098360656,
+      "grad_norm": 4.53946590423584,
+      "learning_rate": 2.562484829278925e-06,
+      "loss": 0.8938,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7741803278688525,
+      "grad_norm": 3.3911256790161133,
+      "learning_rate": 2.5580494997937232e-06,
+      "loss": 1.0317,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7743852459016394,
+      "grad_norm": 4.305842399597168,
+      "learning_rate": 2.5536174490977337e-06,
+      "loss": 0.8718,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7745901639344263,
+      "grad_norm": 3.131964921951294,
+      "learning_rate": 2.5491886791436384e-06,
+      "loss": 0.8388,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7747950819672131,
+      "grad_norm": 2.912113666534424,
+      "learning_rate": 2.5447631918826733e-06,
+      "loss": 0.9083,
+      "step": 3781
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 3.5107622146606445,
+      "learning_rate": 2.540340989264628e-06,
+      "loss": 0.9332,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7752049180327869,
+      "grad_norm": 4.885110378265381,
+      "learning_rate": 2.5359220732378478e-06,
+      "loss": 1.1718,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7754098360655738,
+      "grad_norm": 5.140381813049316,
+      "learning_rate": 2.531506445749219e-06,
+      "loss": 1.1009,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7756147540983607,
+      "grad_norm": 5.696193218231201,
+      "learning_rate": 2.527094108744197e-06,
+      "loss": 0.7518,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7758196721311476,
+      "grad_norm": 7.382317543029785,
+      "learning_rate": 2.5226850641667745e-06,
+      "loss": 1.0458,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7760245901639344,
+      "grad_norm": 2.5367777347564697,
+      "learning_rate": 2.5182793139594998e-06,
+      "loss": 1.0202,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7762295081967213,
+      "grad_norm": 4.3053436279296875,
+      "learning_rate": 2.51387686006346e-06,
+      "loss": 1.0092,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7764344262295082,
+      "grad_norm": 4.8552422523498535,
+      "learning_rate": 2.509477704418303e-06,
+      "loss": 0.8625,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7766393442622951,
+      "grad_norm": 4.173654556274414,
+      "learning_rate": 2.5050818489622144e-06,
+      "loss": 1.0216,
+      "step": 3790
+    },
+    {
+      "epoch": 0.776844262295082,
+      "grad_norm": 5.115068435668945,
+      "learning_rate": 2.5006892956319296e-06,
+      "loss": 0.9873,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7770491803278688,
+      "grad_norm": 4.422987937927246,
+      "learning_rate": 2.4963000463627297e-06,
+      "loss": 1.1408,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7772540983606557,
+      "grad_norm": 4.193176746368408,
+      "learning_rate": 2.4919141030884377e-06,
+      "loss": 1.1464,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7774590163934426,
+      "grad_norm": 3.770535707473755,
+      "learning_rate": 2.4875314677414196e-06,
+      "loss": 0.9949,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7776639344262295,
+      "grad_norm": 4.351651191711426,
+      "learning_rate": 2.4831521422525895e-06,
+      "loss": 1.0159,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7778688524590164,
+      "grad_norm": 2.017727851867676,
+      "learning_rate": 2.4787761285513957e-06,
+      "loss": 0.8175,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7780737704918033,
+      "grad_norm": 3.0114967823028564,
+      "learning_rate": 2.474403428565836e-06,
+      "loss": 0.7156,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7782786885245901,
+      "grad_norm": 3.103724479675293,
+      "learning_rate": 2.470034044222436e-06,
+      "loss": 0.9162,
+      "step": 3798
+    },
+    {
+      "epoch": 0.778483606557377,
+      "grad_norm": 3.7063612937927246,
+      "learning_rate": 2.465667977446268e-06,
+      "loss": 0.9408,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7786885245901639,
+      "grad_norm": 5.56612491607666,
+      "learning_rate": 2.4613052301609474e-06,
+      "loss": 1.0257,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7788934426229508,
+      "grad_norm": 4.285096645355225,
+      "learning_rate": 2.4569458042886242e-06,
+      "loss": 0.8472,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7790983606557377,
+      "grad_norm": 3.7853803634643555,
+      "learning_rate": 2.4525897017499747e-06,
+      "loss": 0.9971,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7793032786885246,
+      "grad_norm": 3.155500650405884,
+      "learning_rate": 2.4482369244642234e-06,
+      "loss": 0.9044,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7795081967213114,
+      "grad_norm": 5.154503345489502,
+      "learning_rate": 2.4438874743491246e-06,
+      "loss": 0.996,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7797131147540983,
+      "grad_norm": 2.709582567214966,
+      "learning_rate": 2.439541353320969e-06,
+      "loss": 1.0841,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7799180327868852,
+      "grad_norm": 3.4971323013305664,
+      "learning_rate": 2.4351985632945773e-06,
+      "loss": 0.9177,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7801229508196721,
+      "grad_norm": 4.899439334869385,
+      "learning_rate": 2.4308591061833064e-06,
+      "loss": 0.9069,
+      "step": 3807
+    },
+    {
+      "epoch": 0.780327868852459,
+      "grad_norm": 5.428472518920898,
+      "learning_rate": 2.4265229838990433e-06,
+      "loss": 0.8981,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7805327868852459,
+      "grad_norm": 4.347060203552246,
+      "learning_rate": 2.4221901983522e-06,
+      "loss": 0.8442,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7807377049180327,
+      "grad_norm": 4.0643768310546875,
+      "learning_rate": 2.4178607514517292e-06,
+      "loss": 0.9866,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7809426229508196,
+      "grad_norm": 2.9939351081848145,
+      "learning_rate": 2.413534645105109e-06,
+      "loss": 0.8918,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7811475409836065,
+      "grad_norm": 3.0441508293151855,
+      "learning_rate": 2.4092118812183375e-06,
+      "loss": 0.9613,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7813524590163935,
+      "grad_norm": 8.436285972595215,
+      "learning_rate": 2.404892461695949e-06,
+      "loss": 1.0317,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7815573770491804,
+      "grad_norm": 4.154617786407471,
+      "learning_rate": 2.400576388441003e-06,
+      "loss": 0.7911,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7817622950819673,
+      "grad_norm": 5.122048854827881,
+      "learning_rate": 2.3962636633550815e-06,
+      "loss": 0.7023,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7819672131147541,
+      "grad_norm": 2.139031410217285,
+      "learning_rate": 2.3919542883382964e-06,
+      "loss": 0.7744,
+      "step": 3816
+    },
+    {
+      "epoch": 0.782172131147541,
+      "grad_norm": 6.990188121795654,
+      "learning_rate": 2.3876482652892773e-06,
+      "loss": 1.0271,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7823770491803279,
+      "grad_norm": 6.034189224243164,
+      "learning_rate": 2.3833455961051866e-06,
+      "loss": 0.8961,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7825819672131148,
+      "grad_norm": 3.9012484550476074,
+      "learning_rate": 2.379046282681691e-06,
+      "loss": 1.2014,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7827868852459017,
+      "grad_norm": 2.0009379386901855,
+      "learning_rate": 2.3747503269130013e-06,
+      "loss": 0.6577,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7829918032786886,
+      "grad_norm": 5.69130802154541,
+      "learning_rate": 2.3704577306918343e-06,
+      "loss": 1.0904,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7831967213114754,
+      "grad_norm": 3.5424599647521973,
+      "learning_rate": 2.3661684959094335e-06,
+      "loss": 0.9353,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7834016393442623,
+      "grad_norm": 2.6538426876068115,
+      "learning_rate": 2.361882624455553e-06,
+      "loss": 0.8554,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7836065573770492,
+      "grad_norm": 5.277229309082031,
+      "learning_rate": 2.357600118218469e-06,
+      "loss": 1.1052,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7838114754098361,
+      "grad_norm": 3.6538636684417725,
+      "learning_rate": 2.3533209790849874e-06,
+      "loss": 1.0164,
+      "step": 3825
+    },
+    {
+      "epoch": 0.784016393442623,
+      "grad_norm": 6.008054256439209,
+      "learning_rate": 2.3490452089404102e-06,
+      "loss": 1.023,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7842213114754099,
+      "grad_norm": 3.288557291030884,
+      "learning_rate": 2.3447728096685685e-06,
+      "loss": 1.0264,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7844262295081967,
+      "grad_norm": 4.873400688171387,
+      "learning_rate": 2.340503783151802e-06,
+      "loss": 0.7311,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7846311475409836,
+      "grad_norm": 5.986291885375977,
+      "learning_rate": 2.336238131270968e-06,
+      "loss": 1.2247,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7848360655737705,
+      "grad_norm": 7.077229022979736,
+      "learning_rate": 2.3319758559054363e-06,
+      "loss": 0.9142,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7850409836065574,
+      "grad_norm": 2.803020715713501,
+      "learning_rate": 2.327716958933088e-06,
+      "loss": 0.9457,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7852459016393443,
+      "grad_norm": 2.74572491645813,
+      "learning_rate": 2.3234614422303192e-06,
+      "loss": 0.8298,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7854508196721312,
+      "grad_norm": 3.914160966873169,
+      "learning_rate": 2.319209307672029e-06,
+      "loss": 1.0058,
+      "step": 3833
+    },
+    {
+      "epoch": 0.785655737704918,
+      "grad_norm": 3.33358097076416,
+      "learning_rate": 2.3149605571316304e-06,
+      "loss": 1.0351,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7858606557377049,
+      "grad_norm": 3.2236366271972656,
+      "learning_rate": 2.310715192481051e-06,
+      "loss": 1.0524,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7860655737704918,
+      "grad_norm": 5.292206287384033,
+      "learning_rate": 2.3064732155907244e-06,
+      "loss": 1.1835,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7862704918032787,
+      "grad_norm": 2.6093640327453613,
+      "learning_rate": 2.3022346283295814e-06,
+      "loss": 0.9045,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7864754098360656,
+      "grad_norm": 4.780330657958984,
+      "learning_rate": 2.2979994325650713e-06,
+      "loss": 1.0355,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7866803278688524,
+      "grad_norm": 2.824227809906006,
+      "learning_rate": 2.293767630163143e-06,
+      "loss": 0.8343,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 5.012356281280518,
+      "learning_rate": 2.289539222988255e-06,
+      "loss": 0.9554,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7870901639344262,
+      "grad_norm": 3.007261276245117,
+      "learning_rate": 2.2853142129033656e-06,
+      "loss": 0.7631,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7872950819672131,
+      "grad_norm": 4.789451599121094,
+      "learning_rate": 2.2810926017699375e-06,
+      "loss": 1.0306,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7875,
+      "grad_norm": 5.27369499206543,
+      "learning_rate": 2.276874391447941e-06,
+      "loss": 0.831,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7877049180327869,
+      "grad_norm": 4.247889518737793,
+      "learning_rate": 2.2726595837958344e-06,
+      "loss": 0.9811,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7879098360655737,
+      "grad_norm": 3.0738987922668457,
+      "learning_rate": 2.268448180670594e-06,
+      "loss": 0.8932,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7881147540983606,
+      "grad_norm": 3.128412961959839,
+      "learning_rate": 2.2642401839276883e-06,
+      "loss": 0.8164,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7883196721311475,
+      "grad_norm": 6.414145469665527,
+      "learning_rate": 2.2600355954210796e-06,
+      "loss": 0.819,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7885245901639344,
+      "grad_norm": 6.108596324920654,
+      "learning_rate": 2.2558344170032355e-06,
+      "loss": 1.0227,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7887295081967213,
+      "grad_norm": 2.4000988006591797,
+      "learning_rate": 2.2516366505251176e-06,
+      "loss": 1.042,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7889344262295082,
+      "grad_norm": 4.397471904754639,
+      "learning_rate": 2.247442297836194e-06,
+      "loss": 0.7773,
+      "step": 3850
+    },
+    {
+      "epoch": 0.789139344262295,
+      "grad_norm": 3.4341061115264893,
+      "learning_rate": 2.243251360784413e-06,
+      "loss": 0.6877,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7893442622950819,
+      "grad_norm": 6.940758228302002,
+      "learning_rate": 2.239063841216229e-06,
+      "loss": 0.9943,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7895491803278688,
+      "grad_norm": 3.0137858390808105,
+      "learning_rate": 2.2348797409765864e-06,
+      "loss": 0.8908,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7897540983606557,
+      "grad_norm": 3.8107590675354004,
+      "learning_rate": 2.230699061908925e-06,
+      "loss": 0.9736,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7899590163934426,
+      "grad_norm": 3.690504789352417,
+      "learning_rate": 2.2265218058551776e-06,
+      "loss": 0.9738,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7901639344262295,
+      "grad_norm": 3.435117721557617,
+      "learning_rate": 2.222347974655766e-06,
+      "loss": 0.9493,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7903688524590164,
+      "grad_norm": 2.4423346519470215,
+      "learning_rate": 2.218177570149609e-06,
+      "loss": 0.7223,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7905737704918033,
+      "grad_norm": 5.462787628173828,
+      "learning_rate": 2.214010594174105e-06,
+      "loss": 0.9761,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7907786885245902,
+      "grad_norm": 7.981245517730713,
+      "learning_rate": 2.20984704856515e-06,
+      "loss": 0.8867,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7909836065573771,
+      "grad_norm": 3.3156898021698,
+      "learning_rate": 2.2056869351571353e-06,
+      "loss": 1.1253,
+      "step": 3860
+    },
+    {
+      "epoch": 0.791188524590164,
+      "grad_norm": 3.538313388824463,
+      "learning_rate": 2.201530255782922e-06,
+      "loss": 0.9345,
+      "step": 3861
+    },
+    {
+      "epoch": 0.7913934426229509,
+      "grad_norm": 2.731868028640747,
+      "learning_rate": 2.1973770122738715e-06,
+      "loss": 0.9822,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7915983606557377,
+      "grad_norm": 2.899815082550049,
+      "learning_rate": 2.193227206459829e-06,
+      "loss": 0.7366,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7918032786885246,
+      "grad_norm": 4.229341983795166,
+      "learning_rate": 2.189080840169123e-06,
+      "loss": 0.5965,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7920081967213115,
+      "grad_norm": 3.3668127059936523,
+      "learning_rate": 2.184937915228569e-06,
+      "loss": 1.2202,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7922131147540984,
+      "grad_norm": 2.9215023517608643,
+      "learning_rate": 2.1807984334634626e-06,
+      "loss": 0.9552,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7924180327868853,
+      "grad_norm": 2.954551935195923,
+      "learning_rate": 2.1766623966975907e-06,
+      "loss": 0.9158,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7926229508196722,
+      "grad_norm": 3.4394872188568115,
+      "learning_rate": 2.1725298067532087e-06,
+      "loss": 0.7872,
+      "step": 3868
+    },
+    {
+      "epoch": 0.792827868852459,
+      "grad_norm": 2.8714683055877686,
+      "learning_rate": 2.1684006654510616e-06,
+      "loss": 0.9746,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7930327868852459,
+      "grad_norm": 2.9164130687713623,
+      "learning_rate": 2.1642749746103807e-06,
+      "loss": 0.8363,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7932377049180328,
+      "grad_norm": 2.674227714538574,
+      "learning_rate": 2.160152736048872e-06,
+      "loss": 0.8349,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7934426229508197,
+      "grad_norm": 2.825063943862915,
+      "learning_rate": 2.1560339515827123e-06,
+      "loss": 0.9912,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7936475409836066,
+      "grad_norm": 2.6091251373291016,
+      "learning_rate": 2.1519186230265663e-06,
+      "loss": 0.8964,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7938524590163935,
+      "grad_norm": 3.085843801498413,
+      "learning_rate": 2.147806752193574e-06,
+      "loss": 0.8755,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7940573770491803,
+      "grad_norm": 3.6388723850250244,
+      "learning_rate": 2.1436983408953526e-06,
+      "loss": 0.9442,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7942622950819672,
+      "grad_norm": 4.039743423461914,
+      "learning_rate": 2.139593390941994e-06,
+      "loss": 0.5633,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7944672131147541,
+      "grad_norm": 3.170715093612671,
+      "learning_rate": 2.135491904142063e-06,
+      "loss": 0.9541,
+      "step": 3877
+    },
+    {
+      "epoch": 0.794672131147541,
+      "grad_norm": 3.0728464126586914,
+      "learning_rate": 2.1313938823026026e-06,
+      "loss": 0.9118,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7948770491803279,
+      "grad_norm": 5.0182085037231445,
+      "learning_rate": 2.1272993272291276e-06,
+      "loss": 0.923,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7950819672131147,
+      "grad_norm": 3.4777979850769043,
+      "learning_rate": 2.123208240725624e-06,
+      "loss": 0.9382,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7952868852459016,
+      "grad_norm": 3.960207939147949,
+      "learning_rate": 2.119120624594554e-06,
+      "loss": 0.9935,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7954918032786885,
+      "grad_norm": 4.9394636154174805,
+      "learning_rate": 2.1150364806368416e-06,
+      "loss": 0.8516,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7956967213114754,
+      "grad_norm": 3.006330728530884,
+      "learning_rate": 2.1109558106518903e-06,
+      "loss": 0.7785,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7959016393442623,
+      "grad_norm": 2.9636478424072266,
+      "learning_rate": 2.1068786164375667e-06,
+      "loss": 1.0439,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7961065573770492,
+      "grad_norm": 3.1806657314300537,
+      "learning_rate": 2.1028048997902175e-06,
+      "loss": 0.7186,
+      "step": 3885
+    },
+    {
+      "epoch": 0.796311475409836,
+      "grad_norm": 4.217372417449951,
+      "learning_rate": 2.098734662504641e-06,
+      "loss": 0.8881,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7965163934426229,
+      "grad_norm": 4.039029598236084,
+      "learning_rate": 2.0946679063741115e-06,
+      "loss": 0.7791,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7967213114754098,
+      "grad_norm": 2.830998659133911,
+      "learning_rate": 2.0906046331903708e-06,
+      "loss": 0.9041,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7969262295081967,
+      "grad_norm": 3.805790424346924,
+      "learning_rate": 2.086544844743622e-06,
+      "loss": 1.1667,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7971311475409836,
+      "grad_norm": 3.4217689037323,
+      "learning_rate": 2.082488542822535e-06,
+      "loss": 1.0563,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7973360655737705,
+      "grad_norm": 2.81428861618042,
+      "learning_rate": 2.078435729214243e-06,
+      "loss": 1.155,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7975409836065573,
+      "grad_norm": 4.425414085388184,
+      "learning_rate": 2.0743864057043474e-06,
+      "loss": 1.1833,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7977459016393442,
+      "grad_norm": 3.464024305343628,
+      "learning_rate": 2.070340574076899e-06,
+      "loss": 0.9845,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7979508196721311,
+      "grad_norm": 8.330756187438965,
+      "learning_rate": 2.066298236114419e-06,
+      "loss": 1.099,
+      "step": 3894
+    },
+    {
+      "epoch": 0.798155737704918,
+      "grad_norm": 3.578822374343872,
+      "learning_rate": 2.0622593935978974e-06,
+      "loss": 0.8351,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7983606557377049,
+      "grad_norm": 2.40861177444458,
+      "learning_rate": 2.0582240483067674e-06,
+      "loss": 1.028,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7985655737704918,
+      "grad_norm": 2.463801622390747,
+      "learning_rate": 2.054192202018932e-06,
+      "loss": 1.0579,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7987704918032786,
+      "grad_norm": 3.310213327407837,
+      "learning_rate": 2.0501638565107486e-06,
+      "loss": 0.839,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7989754098360655,
+      "grad_norm": 3.6901824474334717,
+      "learning_rate": 2.0461390135570357e-06,
+      "loss": 1.0444,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7991803278688525,
+      "grad_norm": 4.017818927764893,
+      "learning_rate": 2.0421176749310657e-06,
+      "loss": 1.1036,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7993852459016394,
+      "grad_norm": 4.81495475769043,
+      "learning_rate": 2.0380998424045663e-06,
+      "loss": 0.8385,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7995901639344263,
+      "grad_norm": 4.574952602386475,
+      "learning_rate": 2.034085517747727e-06,
+      "loss": 1.0952,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7997950819672132,
+      "grad_norm": 5.078124523162842,
+      "learning_rate": 2.030074702729178e-06,
+      "loss": 0.949,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6065661907196045,
+      "learning_rate": 2.0260673991160206e-06,
+      "loss": 1.0101,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8002049180327869,
+      "grad_norm": 4.248940944671631,
+      "learning_rate": 2.0220636086737967e-06,
+      "loss": 0.8468,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8004098360655738,
+      "grad_norm": 3.6319386959075928,
+      "learning_rate": 2.0180633331665077e-06,
+      "loss": 0.749,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8006147540983607,
+      "grad_norm": 4.974079132080078,
+      "learning_rate": 2.0140665743565966e-06,
+      "loss": 0.8925,
+      "step": 3907
+    },
+    {
+      "epoch": 0.8008196721311476,
+      "grad_norm": 6.187105178833008,
+      "learning_rate": 2.0100733340049683e-06,
+      "loss": 1.192,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8010245901639345,
+      "grad_norm": 4.99165153503418,
+      "learning_rate": 2.0060836138709696e-06,
+      "loss": 0.8887,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8012295081967213,
+      "grad_norm": 6.232083797454834,
+      "learning_rate": 2.0020974157124008e-06,
+      "loss": 1.0374,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8014344262295082,
+      "grad_norm": 3.139096260070801,
+      "learning_rate": 1.9981147412855085e-06,
+      "loss": 1.0929,
+      "step": 3911
+    },
+    {
+      "epoch": 0.8016393442622951,
+      "grad_norm": 4.831723213195801,
+      "learning_rate": 1.994135592344988e-06,
+      "loss": 0.7956,
+      "step": 3912
+    },
+    {
+      "epoch": 0.801844262295082,
+      "grad_norm": 3.4528403282165527,
+      "learning_rate": 1.990159970643979e-06,
+      "loss": 1.0451,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8020491803278689,
+      "grad_norm": 4.2525954246521,
+      "learning_rate": 1.9861878779340703e-06,
+      "loss": 0.9548,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8022540983606558,
+      "grad_norm": 3.551820993423462,
+      "learning_rate": 1.9822193159652937e-06,
+      "loss": 0.5938,
+      "step": 3915
+    },
+    {
+      "epoch": 0.8024590163934426,
+      "grad_norm": 2.5375823974609375,
+      "learning_rate": 1.978254286486129e-06,
+      "loss": 0.6927,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8026639344262295,
+      "grad_norm": 5.935129165649414,
+      "learning_rate": 1.9742927912434907e-06,
+      "loss": 1.0983,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8028688524590164,
+      "grad_norm": 3.0073745250701904,
+      "learning_rate": 1.9703348319827452e-06,
+      "loss": 1.0231,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8030737704918033,
+      "grad_norm": 5.048161029815674,
+      "learning_rate": 1.9663804104476957e-06,
+      "loss": 0.9928,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8032786885245902,
+      "grad_norm": 3.178812265396118,
+      "learning_rate": 1.9624295283805973e-06,
+      "loss": 1.0799,
+      "step": 3920
+    },
+    {
+      "epoch": 0.803483606557377,
+      "grad_norm": 3.5672073364257812,
+      "learning_rate": 1.9584821875221283e-06,
+      "loss": 1.0745,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8036885245901639,
+      "grad_norm": 5.0940022468566895,
+      "learning_rate": 1.9545383896114188e-06,
+      "loss": 1.0618,
+      "step": 3922
+    },
+    {
+      "epoch": 0.8038934426229508,
+      "grad_norm": 3.945957660675049,
+      "learning_rate": 1.950598136386036e-06,
+      "loss": 1.0597,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8040983606557377,
+      "grad_norm": 5.158120632171631,
+      "learning_rate": 1.946661429581983e-06,
+      "loss": 0.9959,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8043032786885246,
+      "grad_norm": 4.695667743682861,
+      "learning_rate": 1.9427282709337025e-06,
+      "loss": 1.0502,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8045081967213115,
+      "grad_norm": 2.2972192764282227,
+      "learning_rate": 1.938798662174075e-06,
+      "loss": 0.7837,
+      "step": 3926
+    },
+    {
+      "epoch": 0.8047131147540983,
+      "grad_norm": 4.403879165649414,
+      "learning_rate": 1.9348726050344145e-06,
+      "loss": 1.072,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8049180327868852,
+      "grad_norm": 2.4578707218170166,
+      "learning_rate": 1.9309501012444664e-06,
+      "loss": 0.6965,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8051229508196721,
+      "grad_norm": 3.821096420288086,
+      "learning_rate": 1.92703115253242e-06,
+      "loss": 0.7328,
+      "step": 3929
+    },
+    {
+      "epoch": 0.805327868852459,
+      "grad_norm": 4.071248531341553,
+      "learning_rate": 1.9231157606248963e-06,
+      "loss": 0.9583,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8055327868852459,
+      "grad_norm": 2.534466028213501,
+      "learning_rate": 1.91920392724694e-06,
+      "loss": 1.0178,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8057377049180328,
+      "grad_norm": 2.509268045425415,
+      "learning_rate": 1.9152956541220357e-06,
+      "loss": 0.833,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8059426229508196,
+      "grad_norm": 3.338730812072754,
+      "learning_rate": 1.9113909429721e-06,
+      "loss": 0.8223,
+      "step": 3933
+    },
+    {
+      "epoch": 0.8061475409836065,
+      "grad_norm": 4.255302429199219,
+      "learning_rate": 1.9074897955174764e-06,
+      "loss": 1.119,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8063524590163934,
+      "grad_norm": 6.021346092224121,
+      "learning_rate": 1.9035922134769425e-06,
+      "loss": 1.0675,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8065573770491803,
+      "grad_norm": 1.9522826671600342,
+      "learning_rate": 1.899698198567701e-06,
+      "loss": 0.8312,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8067622950819672,
+      "grad_norm": 2.585707187652588,
+      "learning_rate": 1.8958077525053863e-06,
+      "loss": 0.7557,
+      "step": 3937
+    },
+    {
+      "epoch": 0.8069672131147541,
+      "grad_norm": 4.391520977020264,
+      "learning_rate": 1.8919208770040531e-06,
+      "loss": 0.7695,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8071721311475409,
+      "grad_norm": 2.1626369953155518,
+      "learning_rate": 1.8880375737761958e-06,
+      "loss": 0.78,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8073770491803278,
+      "grad_norm": 3.9568090438842773,
+      "learning_rate": 1.8841578445327258e-06,
+      "loss": 0.9282,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8075819672131147,
+      "grad_norm": 3.5696446895599365,
+      "learning_rate": 1.8802816909829825e-06,
+      "loss": 1.1269,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8077868852459016,
+      "grad_norm": 5.675047397613525,
+      "learning_rate": 1.8764091148347253e-06,
+      "loss": 0.9693,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8079918032786885,
+      "grad_norm": 2.1556270122528076,
+      "learning_rate": 1.8725401177941449e-06,
+      "loss": 0.6753,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8081967213114755,
+      "grad_norm": 4.001691818237305,
+      "learning_rate": 1.8686747015658492e-06,
+      "loss": 0.9029,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8084016393442623,
+      "grad_norm": 5.004266262054443,
+      "learning_rate": 1.8648128678528731e-06,
+      "loss": 0.9922,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8086065573770492,
+      "grad_norm": 3.3776981830596924,
+      "learning_rate": 1.8609546183566707e-06,
+      "loss": 1.0752,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8088114754098361,
+      "grad_norm": 4.525967121124268,
+      "learning_rate": 1.8570999547771163e-06,
+      "loss": 0.6953,
+      "step": 3947
+    },
+    {
+      "epoch": 0.809016393442623,
+      "grad_norm": 5.42959451675415,
+      "learning_rate": 1.8532488788125057e-06,
+      "loss": 0.8001,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8092213114754099,
+      "grad_norm": 4.73175048828125,
+      "learning_rate": 1.8494013921595533e-06,
+      "loss": 0.8176,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8094262295081968,
+      "grad_norm": 2.495774030685425,
+      "learning_rate": 1.8455574965133927e-06,
+      "loss": 0.9654,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8096311475409836,
+      "grad_norm": 3.127222776412964,
+      "learning_rate": 1.841717193567577e-06,
+      "loss": 1.0027,
+      "step": 3951
+    },
+    {
+      "epoch": 0.8098360655737705,
+      "grad_norm": 2.746709108352661,
+      "learning_rate": 1.8378804850140686e-06,
+      "loss": 0.8233,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8100409836065574,
+      "grad_norm": 2.957859992980957,
+      "learning_rate": 1.8340473725432573e-06,
+      "loss": 0.6705,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8102459016393443,
+      "grad_norm": 3.035686731338501,
+      "learning_rate": 1.8302178578439389e-06,
+      "loss": 0.8832,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8104508196721312,
+      "grad_norm": 3.177036762237549,
+      "learning_rate": 1.826391942603336e-06,
+      "loss": 1.05,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8106557377049181,
+      "grad_norm": 4.403125286102295,
+      "learning_rate": 1.8225696285070704e-06,
+      "loss": 0.7739,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8108606557377049,
+      "grad_norm": 7.5669779777526855,
+      "learning_rate": 1.8187509172391881e-06,
+      "loss": 0.8646,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8110655737704918,
+      "grad_norm": 4.07150936126709,
+      "learning_rate": 1.8149358104821435e-06,
+      "loss": 0.7812,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8112704918032787,
+      "grad_norm": 4.963448524475098,
+      "learning_rate": 1.8111243099168053e-06,
+      "loss": 0.7897,
+      "step": 3959
+    },
+    {
+      "epoch": 0.8114754098360656,
+      "grad_norm": 3.325863838195801,
+      "learning_rate": 1.80731641722245e-06,
+      "loss": 0.9748,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8116803278688525,
+      "grad_norm": 3.382282018661499,
+      "learning_rate": 1.803512134076768e-06,
+      "loss": 0.7949,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8118852459016394,
+      "grad_norm": 3.2779994010925293,
+      "learning_rate": 1.7997114621558597e-06,
+      "loss": 0.8197,
+      "step": 3962
+    },
+    {
+      "epoch": 0.8120901639344262,
+      "grad_norm": 5.185385227203369,
+      "learning_rate": 1.7959144031342245e-06,
+      "loss": 0.8293,
+      "step": 3963
+    },
+    {
+      "epoch": 0.8122950819672131,
+      "grad_norm": 4.23775577545166,
+      "learning_rate": 1.7921209586847865e-06,
+      "loss": 1.046,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 5.488034248352051,
+      "learning_rate": 1.7883311304788698e-06,
+      "loss": 0.9574,
+      "step": 3965
+    },
+    {
+      "epoch": 0.8127049180327869,
+      "grad_norm": 2.8818953037261963,
+      "learning_rate": 1.7845449201861975e-06,
+      "loss": 0.7993,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8129098360655738,
+      "grad_norm": 3.231171131134033,
+      "learning_rate": 1.7807623294749087e-06,
+      "loss": 1.031,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8131147540983606,
+      "grad_norm": 5.489217281341553,
+      "learning_rate": 1.776983360011545e-06,
+      "loss": 0.8432,
+      "step": 3968
+    },
+    {
+      "epoch": 0.8133196721311475,
+      "grad_norm": 3.6749942302703857,
+      "learning_rate": 1.7732080134610519e-06,
+      "loss": 1.0809,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8135245901639344,
+      "grad_norm": 2.756143569946289,
+      "learning_rate": 1.7694362914867803e-06,
+      "loss": 1.0897,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8137295081967213,
+      "grad_norm": 3.017864227294922,
+      "learning_rate": 1.7656681957504807e-06,
+      "loss": 0.8035,
+      "step": 3971
+    },
+    {
+      "epoch": 0.8139344262295082,
+      "grad_norm": 3.676959276199341,
+      "learning_rate": 1.761903727912312e-06,
+      "loss": 0.8442,
+      "step": 3972
+    },
+    {
+      "epoch": 0.8141393442622951,
+      "grad_norm": 2.356153964996338,
+      "learning_rate": 1.758142889630824e-06,
+      "loss": 0.6983,
+      "step": 3973
+    },
+    {
+      "epoch": 0.8143442622950819,
+      "grad_norm": 2.5499680042266846,
+      "learning_rate": 1.754385682562979e-06,
+      "loss": 1.0534,
+      "step": 3974
+    },
+    {
+      "epoch": 0.8145491803278688,
+      "grad_norm": 2.983731746673584,
+      "learning_rate": 1.7506321083641353e-06,
+      "loss": 0.9509,
+      "step": 3975
+    },
+    {
+      "epoch": 0.8147540983606557,
+      "grad_norm": 3.677004814147949,
+      "learning_rate": 1.746882168688051e-06,
+      "loss": 1.0872,
+      "step": 3976
+    },
+    {
+      "epoch": 0.8149590163934426,
+      "grad_norm": 7.775208950042725,
+      "learning_rate": 1.743135865186878e-06,
+      "loss": 0.9927,
+      "step": 3977
+    },
+    {
+      "epoch": 0.8151639344262295,
+      "grad_norm": 6.634044170379639,
+      "learning_rate": 1.7393931995111713e-06,
+      "loss": 0.9039,
+      "step": 3978
+    },
+    {
+      "epoch": 0.8153688524590164,
+      "grad_norm": 3.045027256011963,
+      "learning_rate": 1.735654173309882e-06,
+      "loss": 1.0124,
+      "step": 3979
+    },
+    {
+      "epoch": 0.8155737704918032,
+      "grad_norm": 4.484629154205322,
+      "learning_rate": 1.7319187882303579e-06,
+      "loss": 1.0465,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8157786885245901,
+      "grad_norm": 2.035201072692871,
+      "learning_rate": 1.7281870459183425e-06,
+      "loss": 0.5232,
+      "step": 3981
+    },
+    {
+      "epoch": 0.815983606557377,
+      "grad_norm": 3.628594398498535,
+      "learning_rate": 1.7244589480179742e-06,
+      "loss": 0.8966,
+      "step": 3982
+    },
+    {
+      "epoch": 0.8161885245901639,
+      "grad_norm": 3.9330224990844727,
+      "learning_rate": 1.7207344961717854e-06,
+      "loss": 1.0738,
+      "step": 3983
+    },
+    {
+      "epoch": 0.8163934426229508,
+      "grad_norm": 2.184640645980835,
+      "learning_rate": 1.7170136920207025e-06,
+      "loss": 0.6061,
+      "step": 3984
+    },
+    {
+      "epoch": 0.8165983606557377,
+      "grad_norm": 3.5663366317749023,
+      "learning_rate": 1.7132965372040456e-06,
+      "loss": 0.8044,
+      "step": 3985
+    },
+    {
+      "epoch": 0.8168032786885245,
+      "grad_norm": 3.3200559616088867,
+      "learning_rate": 1.7095830333595277e-06,
+      "loss": 1.1214,
+      "step": 3986
+    },
+    {
+      "epoch": 0.8170081967213115,
+      "grad_norm": 4.794538974761963,
+      "learning_rate": 1.7058731821232466e-06,
+      "loss": 0.9666,
+      "step": 3987
+    },
+    {
+      "epoch": 0.8172131147540984,
+      "grad_norm": 2.440300464630127,
+      "learning_rate": 1.7021669851296962e-06,
+      "loss": 0.7389,
+      "step": 3988
+    },
+    {
+      "epoch": 0.8174180327868853,
+      "grad_norm": 5.2145490646362305,
+      "learning_rate": 1.6984644440117648e-06,
+      "loss": 1.32,
+      "step": 3989
+    },
+    {
+      "epoch": 0.8176229508196722,
+      "grad_norm": 3.912095785140991,
+      "learning_rate": 1.6947655604007252e-06,
+      "loss": 0.8541,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8178278688524591,
+      "grad_norm": 5.461912631988525,
+      "learning_rate": 1.6910703359262338e-06,
+      "loss": 1.0935,
+      "step": 3991
+    },
+    {
+      "epoch": 0.8180327868852459,
+      "grad_norm": 3.528472423553467,
+      "learning_rate": 1.6873787722163426e-06,
+      "loss": 0.9685,
+      "step": 3992
+    },
+    {
+      "epoch": 0.8182377049180328,
+      "grad_norm": 4.2801666259765625,
+      "learning_rate": 1.6836908708974886e-06,
+      "loss": 1.0141,
+      "step": 3993
+    },
+    {
+      "epoch": 0.8184426229508197,
+      "grad_norm": 2.6809420585632324,
+      "learning_rate": 1.6800066335944941e-06,
+      "loss": 0.8938,
+      "step": 3994
+    },
+    {
+      "epoch": 0.8186475409836066,
+      "grad_norm": 4.15730619430542,
+      "learning_rate": 1.6763260619305666e-06,
+      "loss": 0.9558,
+      "step": 3995
+    },
+    {
+      "epoch": 0.8188524590163935,
+      "grad_norm": 5.385621070861816,
+      "learning_rate": 1.672649157527302e-06,
+      "loss": 0.8562,
+      "step": 3996
+    },
+    {
+      "epoch": 0.8190573770491804,
+      "grad_norm": 4.03863525390625,
+      "learning_rate": 1.6689759220046785e-06,
+      "loss": 1.1382,
+      "step": 3997
+    },
+    {
+      "epoch": 0.8192622950819672,
+      "grad_norm": 3.6745975017547607,
+      "learning_rate": 1.6653063569810502e-06,
+      "loss": 0.7905,
+      "step": 3998
+    },
+    {
+      "epoch": 0.8194672131147541,
+      "grad_norm": 3.7668657302856445,
+      "learning_rate": 1.6616404640731698e-06,
+      "loss": 1.0608,
+      "step": 3999
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 3.8348240852355957,
+      "learning_rate": 1.6579782448961622e-06,
+      "loss": 1.1046,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8198770491803279,
+      "grad_norm": 2.8567495346069336,
+      "learning_rate": 1.6543197010635315e-06,
+      "loss": 0.879,
+      "step": 4001
+    },
+    {
+      "epoch": 0.8200819672131148,
+      "grad_norm": 15.576361656188965,
+      "learning_rate": 1.6506648341871679e-06,
+      "loss": 1.0216,
+      "step": 4002
+    },
+    {
+      "epoch": 0.8202868852459017,
+      "grad_norm": 3.4510462284088135,
+      "learning_rate": 1.6470136458773412e-06,
+      "loss": 0.8128,
+      "step": 4003
+    },
+    {
+      "epoch": 0.8204918032786885,
+      "grad_norm": 4.886542797088623,
+      "learning_rate": 1.6433661377426968e-06,
+      "loss": 0.8664,
+      "step": 4004
+    },
+    {
+      "epoch": 0.8206967213114754,
+      "grad_norm": 2.790527105331421,
+      "learning_rate": 1.6397223113902627e-06,
+      "loss": 0.7641,
+      "step": 4005
+    },
+    {
+      "epoch": 0.8209016393442623,
+      "grad_norm": 2.984952211380005,
+      "learning_rate": 1.6360821684254424e-06,
+      "loss": 1.0243,
+      "step": 4006
+    },
+    {
+      "epoch": 0.8211065573770492,
+      "grad_norm": 4.396744251251221,
+      "learning_rate": 1.6324457104520198e-06,
+      "loss": 0.787,
+      "step": 4007
+    },
+    {
+      "epoch": 0.8213114754098361,
+      "grad_norm": 4.372068405151367,
+      "learning_rate": 1.6288129390721452e-06,
+      "loss": 1.2927,
+      "step": 4008
+    },
+    {
+      "epoch": 0.821516393442623,
+      "grad_norm": 3.5737905502319336,
+      "learning_rate": 1.6251838558863597e-06,
+      "loss": 1.0816,
+      "step": 4009
+    },
+    {
+      "epoch": 0.8217213114754098,
+      "grad_norm": 5.513006687164307,
+      "learning_rate": 1.621558462493571e-06,
+      "loss": 0.8658,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8219262295081967,
+      "grad_norm": 5.285542964935303,
+      "learning_rate": 1.6179367604910612e-06,
+      "loss": 0.6447,
+      "step": 4011
+    },
+    {
+      "epoch": 0.8221311475409836,
+      "grad_norm": 4.663794994354248,
+      "learning_rate": 1.614318751474485e-06,
+      "loss": 0.8953,
+      "step": 4012
+    },
+    {
+      "epoch": 0.8223360655737705,
+      "grad_norm": 7.307450294494629,
+      "learning_rate": 1.610704437037871e-06,
+      "loss": 0.862,
+      "step": 4013
+    },
+    {
+      "epoch": 0.8225409836065574,
+      "grad_norm": 3.7162275314331055,
+      "learning_rate": 1.6070938187736274e-06,
+      "loss": 0.7173,
+      "step": 4014
+    },
+    {
+      "epoch": 0.8227459016393442,
+      "grad_norm": 6.111292839050293,
+      "learning_rate": 1.6034868982725215e-06,
+      "loss": 0.9495,
+      "step": 4015
+    },
+    {
+      "epoch": 0.8229508196721311,
+      "grad_norm": 7.411086559295654,
+      "learning_rate": 1.599883677123699e-06,
+      "loss": 1.2085,
+      "step": 4016
+    },
+    {
+      "epoch": 0.823155737704918,
+      "grad_norm": 2.88989520072937,
+      "learning_rate": 1.5962841569146747e-06,
+      "loss": 1.0965,
+      "step": 4017
+    },
+    {
+      "epoch": 0.8233606557377049,
+      "grad_norm": 2.9683258533477783,
+      "learning_rate": 1.5926883392313319e-06,
+      "loss": 0.7565,
+      "step": 4018
+    },
+    {
+      "epoch": 0.8235655737704918,
+      "grad_norm": 7.057868480682373,
+      "learning_rate": 1.5890962256579234e-06,
+      "loss": 0.8325,
+      "step": 4019
+    },
+    {
+      "epoch": 0.8237704918032787,
+      "grad_norm": 3.90085506439209,
+      "learning_rate": 1.58550781777707e-06,
+      "loss": 0.798,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8239754098360655,
+      "grad_norm": 2.668856143951416,
+      "learning_rate": 1.5819231171697602e-06,
+      "loss": 0.8901,
+      "step": 4021
+    },
+    {
+      "epoch": 0.8241803278688524,
+      "grad_norm": 2.633272171020508,
+      "learning_rate": 1.5783421254153452e-06,
+      "loss": 0.8298,
+      "step": 4022
+    },
+    {
+      "epoch": 0.8243852459016393,
+      "grad_norm": 2.720069646835327,
+      "learning_rate": 1.5747648440915453e-06,
+      "loss": 0.9942,
+      "step": 4023
+    },
+    {
+      "epoch": 0.8245901639344262,
+      "grad_norm": 3.4883620738983154,
+      "learning_rate": 1.5711912747744517e-06,
+      "loss": 0.8913,
+      "step": 4024
+    },
+    {
+      "epoch": 0.8247950819672131,
+      "grad_norm": 3.130647897720337,
+      "learning_rate": 1.567621419038512e-06,
+      "loss": 0.8925,
+      "step": 4025
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 2.9510891437530518,
+      "learning_rate": 1.5640552784565378e-06,
+      "loss": 0.9207,
+      "step": 4026
+    },
+    {
+      "epoch": 0.8252049180327868,
+      "grad_norm": 4.295780181884766,
+      "learning_rate": 1.5604928545997078e-06,
+      "loss": 0.8057,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8254098360655737,
+      "grad_norm": 2.2888617515563965,
+      "learning_rate": 1.5569341490375623e-06,
+      "loss": 0.9328,
+      "step": 4028
+    },
+    {
+      "epoch": 0.8256147540983606,
+      "grad_norm": 4.345190048217773,
+      "learning_rate": 1.5533791633380025e-06,
+      "loss": 0.6773,
+      "step": 4029
+    },
+    {
+      "epoch": 0.8258196721311475,
+      "grad_norm": 5.452577114105225,
+      "learning_rate": 1.5498278990672911e-06,
+      "loss": 0.9619,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8260245901639345,
+      "grad_norm": 2.6540069580078125,
+      "learning_rate": 1.5462803577900508e-06,
+      "loss": 0.681,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8262295081967214,
+      "grad_norm": 3.675565481185913,
+      "learning_rate": 1.5427365410692663e-06,
+      "loss": 1.0762,
+      "step": 4032
+    },
+    {
+      "epoch": 0.8264344262295082,
+      "grad_norm": 2.4947218894958496,
+      "learning_rate": 1.5391964504662737e-06,
+      "loss": 0.9848,
+      "step": 4033
+    },
+    {
+      "epoch": 0.8266393442622951,
+      "grad_norm": 5.696742057800293,
+      "learning_rate": 1.535660087540779e-06,
+      "loss": 0.8507,
+      "step": 4034
+    },
+    {
+      "epoch": 0.826844262295082,
+      "grad_norm": 5.274368762969971,
+      "learning_rate": 1.5321274538508402e-06,
+      "loss": 1.0248,
+      "step": 4035
+    },
+    {
+      "epoch": 0.8270491803278689,
+      "grad_norm": 4.029303073883057,
+      "learning_rate": 1.5285985509528666e-06,
+      "loss": 1.0947,
+      "step": 4036
+    },
+    {
+      "epoch": 0.8272540983606558,
+      "grad_norm": 4.274089813232422,
+      "learning_rate": 1.5250733804016338e-06,
+      "loss": 0.8092,
+      "step": 4037
+    },
+    {
+      "epoch": 0.8274590163934427,
+      "grad_norm": 3.1778650283813477,
+      "learning_rate": 1.5215519437502635e-06,
+      "loss": 0.7489,
+      "step": 4038
+    },
+    {
+      "epoch": 0.8276639344262295,
+      "grad_norm": 3.1036338806152344,
+      "learning_rate": 1.5180342425502447e-06,
+      "loss": 1.0223,
+      "step": 4039
+    },
+    {
+      "epoch": 0.8278688524590164,
+      "grad_norm": 5.202639579772949,
+      "learning_rate": 1.5145202783514079e-06,
+      "loss": 0.7905,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8280737704918033,
+      "grad_norm": 4.151036262512207,
+      "learning_rate": 1.5110100527019432e-06,
+      "loss": 0.9898,
+      "step": 4041
+    },
+    {
+      "epoch": 0.8282786885245902,
+      "grad_norm": 3.009491205215454,
+      "learning_rate": 1.5075035671483929e-06,
+      "loss": 0.7562,
+      "step": 4042
+    },
+    {
+      "epoch": 0.8284836065573771,
+      "grad_norm": 4.358429908752441,
+      "learning_rate": 1.5040008232356518e-06,
+      "loss": 1.0036,
+      "step": 4043
+    },
+    {
+      "epoch": 0.828688524590164,
+      "grad_norm": 3.933199167251587,
+      "learning_rate": 1.5005018225069646e-06,
+      "loss": 1.1319,
+      "step": 4044
+    },
+    {
+      "epoch": 0.8288934426229508,
+      "grad_norm": 4.474752426147461,
+      "learning_rate": 1.497006566503929e-06,
+      "loss": 0.9573,
+      "step": 4045
+    },
+    {
+      "epoch": 0.8290983606557377,
+      "grad_norm": 6.250091552734375,
+      "learning_rate": 1.4935150567664936e-06,
+      "loss": 0.9767,
+      "step": 4046
+    },
+    {
+      "epoch": 0.8293032786885246,
+      "grad_norm": 2.8639285564422607,
+      "learning_rate": 1.4900272948329509e-06,
+      "loss": 1.0602,
+      "step": 4047
+    },
+    {
+      "epoch": 0.8295081967213115,
+      "grad_norm": 2.715355396270752,
+      "learning_rate": 1.4865432822399439e-06,
+      "loss": 0.899,
+      "step": 4048
+    },
+    {
+      "epoch": 0.8297131147540984,
+      "grad_norm": 2.394219160079956,
+      "learning_rate": 1.483063020522475e-06,
+      "loss": 0.9824,
+      "step": 4049
+    },
+    {
+      "epoch": 0.8299180327868853,
+      "grad_norm": 2.3651156425476074,
+      "learning_rate": 1.4795865112138763e-06,
+      "loss": 0.7665,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8301229508196721,
+      "grad_norm": 3.5672595500946045,
+      "learning_rate": 1.4761137558458382e-06,
+      "loss": 1.0155,
+      "step": 4051
+    },
+    {
+      "epoch": 0.830327868852459,
+      "grad_norm": 7.19058895111084,
+      "learning_rate": 1.4726447559483937e-06,
+      "loss": 1.0296,
+      "step": 4052
+    },
+    {
+      "epoch": 0.8305327868852459,
+      "grad_norm": 3.3648524284362793,
+      "learning_rate": 1.469179513049921e-06,
+      "loss": 0.8546,
+      "step": 4053
+    },
+    {
+      "epoch": 0.8307377049180328,
+      "grad_norm": 2.0833475589752197,
+      "learning_rate": 1.4657180286771454e-06,
+      "loss": 0.955,
+      "step": 4054
+    },
+    {
+      "epoch": 0.8309426229508197,
+      "grad_norm": 6.537900447845459,
+      "learning_rate": 1.4622603043551331e-06,
+      "loss": 1.1956,
+      "step": 4055
+    },
+    {
+      "epoch": 0.8311475409836065,
+      "grad_norm": 2.457509994506836,
+      "learning_rate": 1.4588063416072984e-06,
+      "loss": 0.8556,
+      "step": 4056
+    },
+    {
+      "epoch": 0.8313524590163934,
+      "grad_norm": 2.8449268341064453,
+      "learning_rate": 1.4553561419553885e-06,
+      "loss": 0.8462,
+      "step": 4057
+    },
+    {
+      "epoch": 0.8315573770491803,
+      "grad_norm": 5.2807111740112305,
+      "learning_rate": 1.451909706919501e-06,
+      "loss": 1.0156,
+      "step": 4058
+    },
+    {
+      "epoch": 0.8317622950819672,
+      "grad_norm": 3.3012213706970215,
+      "learning_rate": 1.4484670380180776e-06,
+      "loss": 0.9706,
+      "step": 4059
+    },
+    {
+      "epoch": 0.8319672131147541,
+      "grad_norm": 3.5946993827819824,
+      "learning_rate": 1.4450281367678954e-06,
+      "loss": 0.7719,
+      "step": 4060
+    },
+    {
+      "epoch": 0.832172131147541,
+      "grad_norm": 5.907283306121826,
+      "learning_rate": 1.4415930046840675e-06,
+      "loss": 0.9463,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8323770491803278,
+      "grad_norm": 5.490918159484863,
+      "learning_rate": 1.4381616432800538e-06,
+      "loss": 0.8912,
+      "step": 4062
+    },
+    {
+      "epoch": 0.8325819672131147,
+      "grad_norm": 6.569730758666992,
+      "learning_rate": 1.4347340540676513e-06,
+      "loss": 1.2028,
+      "step": 4063
+    },
+    {
+      "epoch": 0.8327868852459016,
+      "grad_norm": 3.0083420276641846,
+      "learning_rate": 1.4313102385569922e-06,
+      "loss": 0.8896,
+      "step": 4064
+    },
+    {
+      "epoch": 0.8329918032786885,
+      "grad_norm": 5.503916263580322,
+      "learning_rate": 1.4278901982565484e-06,
+      "loss": 0.9211,
+      "step": 4065
+    },
+    {
+      "epoch": 0.8331967213114754,
+      "grad_norm": 2.876513719558716,
+      "learning_rate": 1.4244739346731296e-06,
+      "loss": 0.8526,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8334016393442623,
+      "grad_norm": 2.6975088119506836,
+      "learning_rate": 1.421061449311878e-06,
+      "loss": 0.9319,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8336065573770491,
+      "grad_norm": 5.493238925933838,
+      "learning_rate": 1.4176527436762732e-06,
+      "loss": 0.8795,
+      "step": 4068
+    },
+    {
+      "epoch": 0.833811475409836,
+      "grad_norm": 5.756675720214844,
+      "learning_rate": 1.414247819268131e-06,
+      "loss": 0.8106,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8340163934426229,
+      "grad_norm": 2.695922374725342,
+      "learning_rate": 1.4108466775876007e-06,
+      "loss": 0.818,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8342213114754098,
+      "grad_norm": 4.249372959136963,
+      "learning_rate": 1.4074493201331595e-06,
+      "loss": 0.9099,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8344262295081967,
+      "grad_norm": 2.951392889022827,
+      "learning_rate": 1.404055748401625e-06,
+      "loss": 0.9734,
+      "step": 4072
+    },
+    {
+      "epoch": 0.8346311475409836,
+      "grad_norm": 2.4206743240356445,
+      "learning_rate": 1.4006659638881404e-06,
+      "loss": 0.7158,
+      "step": 4073
+    },
+    {
+      "epoch": 0.8348360655737705,
+      "grad_norm": 3.092503786087036,
+      "learning_rate": 1.3972799680861904e-06,
+      "loss": 0.7052,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8350409836065574,
+      "grad_norm": 3.1200785636901855,
+      "learning_rate": 1.3938977624875783e-06,
+      "loss": 0.9096,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8352459016393443,
+      "grad_norm": 6.584498882293701,
+      "learning_rate": 1.390519348582443e-06,
+      "loss": 1.1013,
+      "step": 4076
+    },
+    {
+      "epoch": 0.8354508196721312,
+      "grad_norm": 4.165837287902832,
+      "learning_rate": 1.3871447278592554e-06,
+      "loss": 0.8923,
+      "step": 4077
+    },
+    {
+      "epoch": 0.8356557377049181,
+      "grad_norm": 2.5938525199890137,
+      "learning_rate": 1.3837739018048101e-06,
+      "loss": 0.785,
+      "step": 4078
+    },
+    {
+      "epoch": 0.835860655737705,
+      "grad_norm": 2.743640422821045,
+      "learning_rate": 1.3804068719042341e-06,
+      "loss": 0.8336,
+      "step": 4079
+    },
+    {
+      "epoch": 0.8360655737704918,
+      "grad_norm": 4.0274481773376465,
+      "learning_rate": 1.3770436396409781e-06,
+      "loss": 0.8492,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8362704918032787,
+      "grad_norm": 3.509920835494995,
+      "learning_rate": 1.3736842064968259e-06,
+      "loss": 1.0141,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8364754098360656,
+      "grad_norm": 4.382757663726807,
+      "learning_rate": 1.3703285739518779e-06,
+      "loss": 0.9002,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8366803278688525,
+      "grad_norm": 3.118889331817627,
+      "learning_rate": 1.366976743484565e-06,
+      "loss": 0.9787,
+      "step": 4083
+    },
+    {
+      "epoch": 0.8368852459016394,
+      "grad_norm": 3.9362759590148926,
+      "learning_rate": 1.36362871657165e-06,
+      "loss": 1.059,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8370901639344263,
+      "grad_norm": 4.2612833976745605,
+      "learning_rate": 1.3602844946882077e-06,
+      "loss": 1.0611,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8372950819672131,
+      "grad_norm": 2.839545488357544,
+      "learning_rate": 1.3569440793076415e-06,
+      "loss": 0.7194,
+      "step": 4086
+    },
+    {
+      "epoch": 0.8375,
+      "grad_norm": 6.562082290649414,
+      "learning_rate": 1.3536074719016811e-06,
+      "loss": 1.1499,
+      "step": 4087
+    },
+    {
+      "epoch": 0.8377049180327869,
+      "grad_norm": 5.631674766540527,
+      "learning_rate": 1.3502746739403738e-06,
+      "loss": 1.0138,
+      "step": 4088
+    },
+    {
+      "epoch": 0.8379098360655738,
+      "grad_norm": 2.460554599761963,
+      "learning_rate": 1.3469456868920916e-06,
+      "loss": 0.8183,
+      "step": 4089
+    },
+    {
+      "epoch": 0.8381147540983607,
+      "grad_norm": 2.8312649726867676,
+      "learning_rate": 1.3436205122235246e-06,
+      "loss": 0.8059,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8383196721311476,
+      "grad_norm": 3.9526145458221436,
+      "learning_rate": 1.340299151399689e-06,
+      "loss": 0.8445,
+      "step": 4091
+    },
+    {
+      "epoch": 0.8385245901639344,
+      "grad_norm": 4.753930568695068,
+      "learning_rate": 1.336981605883907e-06,
+      "loss": 0.9591,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8387295081967213,
+      "grad_norm": 8.88511848449707,
+      "learning_rate": 1.3336678771378386e-06,
+      "loss": 1.0876,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8389344262295082,
+      "grad_norm": 5.798031806945801,
+      "learning_rate": 1.3303579666214494e-06,
+      "loss": 1.1171,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8391393442622951,
+      "grad_norm": 3.6403894424438477,
+      "learning_rate": 1.3270518757930295e-06,
+      "loss": 0.8362,
+      "step": 4095
+    },
+    {
+      "epoch": 0.839344262295082,
+      "grad_norm": 6.810196876525879,
+      "learning_rate": 1.3237496061091771e-06,
+      "loss": 0.977,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8395491803278688,
+      "grad_norm": 3.666501760482788,
+      "learning_rate": 1.320451159024817e-06,
+      "loss": 0.8155,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8397540983606557,
+      "grad_norm": 2.211045026779175,
+      "learning_rate": 1.3171565359931826e-06,
+      "loss": 0.8074,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8399590163934426,
+      "grad_norm": 2.083226203918457,
+      "learning_rate": 1.3138657384658281e-06,
+      "loss": 0.695,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8401639344262295,
+      "grad_norm": 4.736085891723633,
+      "learning_rate": 1.3105787678926175e-06,
+      "loss": 0.7288,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8403688524590164,
+      "grad_norm": 5.597240924835205,
+      "learning_rate": 1.307295625721733e-06,
+      "loss": 0.9203,
+      "step": 4101
+    },
+    {
+      "epoch": 0.8405737704918033,
+      "grad_norm": 2.289820671081543,
+      "learning_rate": 1.3040163133996652e-06,
+      "loss": 0.8053,
+      "step": 4102
+    },
+    {
+      "epoch": 0.8407786885245901,
+      "grad_norm": 3.8666608333587646,
+      "learning_rate": 1.3007408323712223e-06,
+      "loss": 1.0774,
+      "step": 4103
+    },
+    {
+      "epoch": 0.840983606557377,
+      "grad_norm": 3.3604941368103027,
+      "learning_rate": 1.2974691840795206e-06,
+      "loss": 0.9708,
+      "step": 4104
+    },
+    {
+      "epoch": 0.8411885245901639,
+      "grad_norm": 5.808337688446045,
+      "learning_rate": 1.2942013699659928e-06,
+      "loss": 0.8921,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8413934426229508,
+      "grad_norm": 3.5047223567962646,
+      "learning_rate": 1.2909373914703737e-06,
+      "loss": 0.9683,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8415983606557377,
+      "grad_norm": 3.1922216415405273,
+      "learning_rate": 1.2876772500307156e-06,
+      "loss": 0.5854,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8418032786885246,
+      "grad_norm": 3.947108745574951,
+      "learning_rate": 1.2844209470833758e-06,
+      "loss": 0.7823,
+      "step": 4108
+    },
+    {
+      "epoch": 0.8420081967213114,
+      "grad_norm": 2.9682624340057373,
+      "learning_rate": 1.281168484063029e-06,
+      "loss": 0.8152,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8422131147540983,
+      "grad_norm": 3.7374470233917236,
+      "learning_rate": 1.2779198624026468e-06,
+      "loss": 0.8419,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8424180327868852,
+      "grad_norm": 5.640323162078857,
+      "learning_rate": 1.274675083533513e-06,
+      "loss": 1.1156,
+      "step": 4111
+    },
+    {
+      "epoch": 0.8426229508196721,
+      "grad_norm": 3.605867385864258,
+      "learning_rate": 1.271434148885221e-06,
+      "loss": 0.9538,
+      "step": 4112
+    },
+    {
+      "epoch": 0.842827868852459,
+      "grad_norm": 3.4609262943267822,
+      "learning_rate": 1.2681970598856674e-06,
+      "loss": 0.8265,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8430327868852459,
+      "grad_norm": 2.9238991737365723,
+      "learning_rate": 1.2649638179610547e-06,
+      "loss": 0.8891,
+      "step": 4114
+    },
+    {
+      "epoch": 0.8432377049180327,
+      "grad_norm": 3.1056485176086426,
+      "learning_rate": 1.261734424535891e-06,
+      "loss": 1.0279,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8434426229508196,
+      "grad_norm": 6.52197790145874,
+      "learning_rate": 1.2585088810329915e-06,
+      "loss": 1.025,
+      "step": 4116
+    },
+    {
+      "epoch": 0.8436475409836065,
+      "grad_norm": 3.413065195083618,
+      "learning_rate": 1.2552871888734652e-06,
+      "loss": 0.8337,
+      "step": 4117
+    },
+    {
+      "epoch": 0.8438524590163935,
+      "grad_norm": 3.6824684143066406,
+      "learning_rate": 1.25206934947674e-06,
+      "loss": 0.9303,
+      "step": 4118
+    },
+    {
+      "epoch": 0.8440573770491804,
+      "grad_norm": 3.072525978088379,
+      "learning_rate": 1.2488553642605362e-06,
+      "loss": 0.7438,
+      "step": 4119
+    },
+    {
+      "epoch": 0.8442622950819673,
+      "grad_norm": 2.7666165828704834,
+      "learning_rate": 1.2456452346408731e-06,
+      "loss": 0.7533,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8444672131147541,
+      "grad_norm": 2.2164878845214844,
+      "learning_rate": 1.2424389620320797e-06,
+      "loss": 0.8529,
+      "step": 4121
+    },
+    {
+      "epoch": 0.844672131147541,
+      "grad_norm": 4.405210494995117,
+      "learning_rate": 1.2392365478467806e-06,
+      "loss": 0.7025,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8448770491803279,
+      "grad_norm": 3.574648380279541,
+      "learning_rate": 1.2360379934959022e-06,
+      "loss": 0.9943,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8450819672131148,
+      "grad_norm": 2.0514490604400635,
+      "learning_rate": 1.2328433003886698e-06,
+      "loss": 0.7278,
+      "step": 4124
+    },
+    {
+      "epoch": 0.8452868852459017,
+      "grad_norm": 3.3228611946105957,
+      "learning_rate": 1.2296524699326073e-06,
+      "loss": 0.8568,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8454918032786886,
+      "grad_norm": 2.6846065521240234,
+      "learning_rate": 1.2264655035335405e-06,
+      "loss": 0.8333,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8456967213114754,
+      "grad_norm": 4.050583839416504,
+      "learning_rate": 1.2232824025955803e-06,
+      "loss": 0.5258,
+      "step": 4127
+    },
+    {
+      "epoch": 0.8459016393442623,
+      "grad_norm": 4.288610458374023,
+      "learning_rate": 1.220103168521153e-06,
+      "loss": 0.8079,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8461065573770492,
+      "grad_norm": 2.8755109310150146,
+      "learning_rate": 1.2169278027109688e-06,
+      "loss": 1.0296,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8463114754098361,
+      "grad_norm": 4.518649101257324,
+      "learning_rate": 1.2137563065640367e-06,
+      "loss": 0.7806,
+      "step": 4130
+    },
+    {
+      "epoch": 0.846516393442623,
+      "grad_norm": 5.472248077392578,
+      "learning_rate": 1.210588681477659e-06,
+      "loss": 0.7171,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8467213114754099,
+      "grad_norm": 10.693553924560547,
+      "learning_rate": 1.207424928847436e-06,
+      "loss": 1.1285,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8469262295081967,
+      "grad_norm": 4.795588970184326,
+      "learning_rate": 1.2042650500672581e-06,
+      "loss": 1.1007,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8471311475409836,
+      "grad_norm": 3.259746789932251,
+      "learning_rate": 1.201109046529313e-06,
+      "loss": 1.0703,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8473360655737705,
+      "grad_norm": 5.529443264007568,
+      "learning_rate": 1.1979569196240781e-06,
+      "loss": 0.7912,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8475409836065574,
+      "grad_norm": 3.229245901107788,
+      "learning_rate": 1.1948086707403238e-06,
+      "loss": 1.0832,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8477459016393443,
+      "grad_norm": 3.5877585411071777,
+      "learning_rate": 1.191664301265112e-06,
+      "loss": 0.9319,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8479508196721312,
+      "grad_norm": 2.311918020248413,
+      "learning_rate": 1.188523812583795e-06,
+      "loss": 0.8438,
+      "step": 4138
+    },
+    {
+      "epoch": 0.848155737704918,
+      "grad_norm": 5.847475051879883,
+      "learning_rate": 1.1853872060800153e-06,
+      "loss": 1.0024,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8483606557377049,
+      "grad_norm": 3.879202365875244,
+      "learning_rate": 1.1822544831357074e-06,
+      "loss": 0.7239,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8485655737704918,
+      "grad_norm": 6.533285140991211,
+      "learning_rate": 1.1791256451310883e-06,
+      "loss": 1.02,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8487704918032787,
+      "grad_norm": 5.150876998901367,
+      "learning_rate": 1.1760006934446676e-06,
+      "loss": 0.8981,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8489754098360656,
+      "grad_norm": 3.919325828552246,
+      "learning_rate": 1.1728796294532475e-06,
+      "loss": 0.9635,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8491803278688524,
+      "grad_norm": 2.387570381164551,
+      "learning_rate": 1.1697624545319108e-06,
+      "loss": 0.7352,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8493852459016393,
+      "grad_norm": 3.652445077896118,
+      "learning_rate": 1.1666491700540272e-06,
+      "loss": 0.8704,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8495901639344262,
+      "grad_norm": 6.381997108459473,
+      "learning_rate": 1.1635397773912538e-06,
+      "loss": 1.1481,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8497950819672131,
+      "grad_norm": 5.255242824554443,
+      "learning_rate": 1.1604342779135325e-06,
+      "loss": 0.8729,
+      "step": 4147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.879337787628174,
+      "learning_rate": 1.157332672989092e-06,
+      "loss": 0.9385,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8502049180327869,
+      "grad_norm": 4.586965084075928,
+      "learning_rate": 1.154234963984442e-06,
+      "loss": 0.9255,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8504098360655737,
+      "grad_norm": 3.699953556060791,
+      "learning_rate": 1.151141152264379e-06,
+      "loss": 0.8983,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8506147540983606,
+      "grad_norm": 3.1830549240112305,
+      "learning_rate": 1.1480512391919808e-06,
+      "loss": 0.9955,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8508196721311475,
+      "grad_norm": 3.025721788406372,
+      "learning_rate": 1.1449652261286026e-06,
+      "loss": 1.0131,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8510245901639344,
+      "grad_norm": 3.3369204998016357,
+      "learning_rate": 1.1418831144338915e-06,
+      "loss": 0.85,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8512295081967213,
+      "grad_norm": 3.020482301712036,
+      "learning_rate": 1.1388049054657702e-06,
+      "loss": 1.1245,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8514344262295082,
+      "grad_norm": 5.343222141265869,
+      "learning_rate": 1.135730600580439e-06,
+      "loss": 0.9133,
+      "step": 4155
+    },
+    {
+      "epoch": 0.851639344262295,
+      "grad_norm": 3.541775703430176,
+      "learning_rate": 1.1326602011323816e-06,
+      "loss": 0.9189,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8518442622950819,
+      "grad_norm": 4.2086005210876465,
+      "learning_rate": 1.1295937084743636e-06,
+      "loss": 0.7454,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8520491803278688,
+      "grad_norm": 5.733921527862549,
+      "learning_rate": 1.1265311239574228e-06,
+      "loss": 0.8692,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8522540983606557,
+      "grad_norm": 5.8909454345703125,
+      "learning_rate": 1.1234724489308813e-06,
+      "loss": 0.9357,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 4.0233001708984375,
+      "learning_rate": 1.1204176847423343e-06,
+      "loss": 0.8345,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8526639344262295,
+      "grad_norm": 3.792832136154175,
+      "learning_rate": 1.1173668327376597e-06,
+      "loss": 1.0872,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8528688524590164,
+      "grad_norm": 3.329162359237671,
+      "learning_rate": 1.114319894261e-06,
+      "loss": 0.885,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8530737704918033,
+      "grad_norm": 3.1484475135803223,
+      "learning_rate": 1.1112768706547893e-06,
+      "loss": 0.9729,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8532786885245902,
+      "grad_norm": 5.513585567474365,
+      "learning_rate": 1.1082377632597252e-06,
+      "loss": 1.1539,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8534836065573771,
+      "grad_norm": 2.805753707885742,
+      "learning_rate": 1.1052025734147865e-06,
+      "loss": 0.7415,
+      "step": 4165
+    },
+    {
+      "epoch": 0.853688524590164,
+      "grad_norm": 5.4556565284729,
+      "learning_rate": 1.102171302457219e-06,
+      "loss": 1.1939,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8538934426229509,
+      "grad_norm": 2.3507258892059326,
+      "learning_rate": 1.0991439517225467e-06,
+      "loss": 0.8358,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8540983606557377,
+      "grad_norm": 2.8390090465545654,
+      "learning_rate": 1.0961205225445703e-06,
+      "loss": 0.8257,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8543032786885246,
+      "grad_norm": 4.246749401092529,
+      "learning_rate": 1.0931010162553534e-06,
+      "loss": 0.7372,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8545081967213115,
+      "grad_norm": 3.0351626873016357,
+      "learning_rate": 1.0900854341852375e-06,
+      "loss": 0.9542,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8547131147540984,
+      "grad_norm": 4.98389196395874,
+      "learning_rate": 1.0870737776628327e-06,
+      "loss": 0.8941,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8549180327868853,
+      "grad_norm": 7.35919189453125,
+      "learning_rate": 1.0840660480150234e-06,
+      "loss": 1.1623,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8551229508196722,
+      "grad_norm": 5.4004597663879395,
+      "learning_rate": 1.0810622465669574e-06,
+      "loss": 1.1947,
+      "step": 4173
+    },
+    {
+      "epoch": 0.855327868852459,
+      "grad_norm": 2.4218039512634277,
+      "learning_rate": 1.0780623746420592e-06,
+      "loss": 0.9034,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8555327868852459,
+      "grad_norm": 3.250845432281494,
+      "learning_rate": 1.0750664335620176e-06,
+      "loss": 1.1073,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8557377049180328,
+      "grad_norm": 5.125436305999756,
+      "learning_rate": 1.0720744246467862e-06,
+      "loss": 1.077,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8559426229508197,
+      "grad_norm": 5.197737693786621,
+      "learning_rate": 1.069086349214592e-06,
+      "loss": 1.078,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8561475409836066,
+      "grad_norm": 3.3494086265563965,
+      "learning_rate": 1.0661022085819305e-06,
+      "loss": 1.0547,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8563524590163935,
+      "grad_norm": 2.9569246768951416,
+      "learning_rate": 1.0631220040635593e-06,
+      "loss": 0.9565,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8565573770491803,
+      "grad_norm": 3.4068689346313477,
+      "learning_rate": 1.0601457369724988e-06,
+      "loss": 0.9154,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8567622950819672,
+      "grad_norm": 2.630748748779297,
+      "learning_rate": 1.0571734086200424e-06,
+      "loss": 0.9184,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8569672131147541,
+      "grad_norm": 3.9669153690338135,
+      "learning_rate": 1.0542050203157428e-06,
+      "loss": 0.9039,
+      "step": 4182
+    },
+    {
+      "epoch": 0.857172131147541,
+      "grad_norm": 4.605662822723389,
+      "learning_rate": 1.051240573367419e-06,
+      "loss": 0.9036,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8573770491803279,
+      "grad_norm": 2.4805221557617188,
+      "learning_rate": 1.0482800690811523e-06,
+      "loss": 0.9778,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8575819672131147,
+      "grad_norm": 3.671537399291992,
+      "learning_rate": 1.045323508761288e-06,
+      "loss": 1.0236,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8577868852459016,
+      "grad_norm": 2.403724193572998,
+      "learning_rate": 1.042370893710436e-06,
+      "loss": 1.0394,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8579918032786885,
+      "grad_norm": 3.0152883529663086,
+      "learning_rate": 1.0394222252294572e-06,
+      "loss": 0.7938,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8581967213114754,
+      "grad_norm": 4.705810546875,
+      "learning_rate": 1.036477504617489e-06,
+      "loss": 0.8376,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8584016393442623,
+      "grad_norm": 3.537546157836914,
+      "learning_rate": 1.0335367331719226e-06,
+      "loss": 0.9382,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8586065573770492,
+      "grad_norm": 4.423977851867676,
+      "learning_rate": 1.0305999121884047e-06,
+      "loss": 0.9518,
+      "step": 4190
+    },
+    {
+      "epoch": 0.858811475409836,
+      "grad_norm": 2.1691551208496094,
+      "learning_rate": 1.027667042960847e-06,
+      "loss": 0.703,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8590163934426229,
+      "grad_norm": 3.274482011795044,
+      "learning_rate": 1.0247381267814183e-06,
+      "loss": 0.995,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8592213114754098,
+      "grad_norm": 3.9726579189300537,
+      "learning_rate": 1.0218131649405494e-06,
+      "loss": 0.9629,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8594262295081967,
+      "grad_norm": 7.894799709320068,
+      "learning_rate": 1.0188921587269229e-06,
+      "loss": 1.1603,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8596311475409836,
+      "grad_norm": 7.897376537322998,
+      "learning_rate": 1.0159751094274818e-06,
+      "loss": 0.9409,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8598360655737705,
+      "grad_norm": 3.150796890258789,
+      "learning_rate": 1.0130620183274254e-06,
+      "loss": 1.1218,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8600409836065573,
+      "grad_norm": 3.3063602447509766,
+      "learning_rate": 1.0101528867102095e-06,
+      "loss": 0.8783,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8602459016393442,
+      "grad_norm": 4.93207311630249,
+      "learning_rate": 1.0072477158575455e-06,
+      "loss": 1.0268,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8604508196721311,
+      "grad_norm": 3.879265308380127,
+      "learning_rate": 1.0043465070493984e-06,
+      "loss": 1.0653,
+      "step": 4199
+    },
+    {
+      "epoch": 0.860655737704918,
+      "grad_norm": 3.279940128326416,
+      "learning_rate": 1.0014492615639905e-06,
+      "loss": 1.1503,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8608606557377049,
+      "grad_norm": 2.3536229133605957,
+      "learning_rate": 9.985559806777933e-07,
+      "loss": 1.069,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8610655737704918,
+      "grad_norm": 4.679298400878906,
+      "learning_rate": 9.956666656655324e-07,
+      "loss": 0.953,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8612704918032786,
+      "grad_norm": 3.4801876544952393,
+      "learning_rate": 9.927813178001956e-07,
+      "loss": 0.8765,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8614754098360655,
+      "grad_norm": 5.493312358856201,
+      "learning_rate": 9.898999383530072e-07,
+      "loss": 0.8629,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8616803278688525,
+      "grad_norm": 3.486377716064453,
+      "learning_rate": 9.870225285934542e-07,
+      "loss": 0.7568,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8618852459016394,
+      "grad_norm": 1.8946114778518677,
+      "learning_rate": 9.841490897892713e-07,
+      "loss": 0.7495,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8620901639344263,
+      "grad_norm": 4.952028751373291,
+      "learning_rate": 9.812796232064447e-07,
+      "loss": 1.0214,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8622950819672132,
+      "grad_norm": 6.680344104766846,
+      "learning_rate": 9.784141301092086e-07,
+      "loss": 0.9701,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8625,
+      "grad_norm": 2.030212640762329,
+      "learning_rate": 9.755526117600477e-07,
+      "loss": 0.8073,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8627049180327869,
+      "grad_norm": 3.5559685230255127,
+      "learning_rate": 9.726950694196968e-07,
+      "loss": 0.7875,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8629098360655738,
+      "grad_norm": 3.5589301586151123,
+      "learning_rate": 9.698415043471355e-07,
+      "loss": 0.9124,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8631147540983607,
+      "grad_norm": 3.2628114223480225,
+      "learning_rate": 9.669919177995913e-07,
+      "loss": 1.1284,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8633196721311476,
+      "grad_norm": 7.01269006729126,
+      "learning_rate": 9.641463110325445e-07,
+      "loss": 1.1604,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8635245901639345,
+      "grad_norm": 4.8604207038879395,
+      "learning_rate": 9.613046852997198e-07,
+      "loss": 1.2153,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8637295081967213,
+      "grad_norm": 2.2574055194854736,
+      "learning_rate": 9.584670418530805e-07,
+      "loss": 0.7371,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8639344262295082,
+      "grad_norm": 4.213574409484863,
+      "learning_rate": 9.556333819428443e-07,
+      "loss": 1.1434,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8641393442622951,
+      "grad_norm": 3.915079355239868,
+      "learning_rate": 9.5280370681747e-07,
+      "loss": 0.7769,
+      "step": 4217
+    },
+    {
+      "epoch": 0.864344262295082,
+      "grad_norm": 2.938736915588379,
+      "learning_rate": 9.499780177236628e-07,
+      "loss": 0.8907,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8645491803278689,
+      "grad_norm": 3.599963665008545,
+      "learning_rate": 9.47156315906369e-07,
+      "loss": 1.0417,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8647540983606558,
+      "grad_norm": 5.782448768615723,
+      "learning_rate": 9.443386026087786e-07,
+      "loss": 0.9515,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8649590163934426,
+      "grad_norm": 2.7817325592041016,
+      "learning_rate": 9.415248790723274e-07,
+      "loss": 0.8535,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8651639344262295,
+      "grad_norm": 3.0262818336486816,
+      "learning_rate": 9.387151465366906e-07,
+      "loss": 0.8827,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8653688524590164,
+      "grad_norm": 2.8930373191833496,
+      "learning_rate": 9.359094062397855e-07,
+      "loss": 0.9734,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8655737704918033,
+      "grad_norm": 2.761871099472046,
+      "learning_rate": 9.331076594177724e-07,
+      "loss": 0.8605,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8657786885245902,
+      "grad_norm": 3.389500617980957,
+      "learning_rate": 9.30309907305047e-07,
+      "loss": 0.7809,
+      "step": 4225
+    },
+    {
+      "epoch": 0.865983606557377,
+      "grad_norm": 3.6373066902160645,
+      "learning_rate": 9.275161511342501e-07,
+      "loss": 0.8659,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8661885245901639,
+      "grad_norm": 4.36216926574707,
+      "learning_rate": 9.247263921362581e-07,
+      "loss": 1.0504,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8663934426229508,
+      "grad_norm": 4.095118999481201,
+      "learning_rate": 9.219406315401957e-07,
+      "loss": 0.8105,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8665983606557377,
+      "grad_norm": 3.006742000579834,
+      "learning_rate": 9.191588705734111e-07,
+      "loss": 0.7262,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8668032786885246,
+      "grad_norm": 5.805240631103516,
+      "learning_rate": 9.163811104615006e-07,
+      "loss": 0.9855,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8670081967213115,
+      "grad_norm": 2.338980197906494,
+      "learning_rate": 9.136073524282951e-07,
+      "loss": 0.7336,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8672131147540983,
+      "grad_norm": 2.6258912086486816,
+      "learning_rate": 9.108375976958639e-07,
+      "loss": 0.807,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8674180327868852,
+      "grad_norm": 6.099117279052734,
+      "learning_rate": 9.080718474845085e-07,
+      "loss": 1.0402,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8676229508196721,
+      "grad_norm": 2.639775037765503,
+      "learning_rate": 9.053101030127709e-07,
+      "loss": 0.7178,
+      "step": 4234
+    },
+    {
+      "epoch": 0.867827868852459,
+      "grad_norm": 4.8590087890625,
+      "learning_rate": 9.025523654974266e-07,
+      "loss": 0.9694,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8680327868852459,
+      "grad_norm": 3.278620719909668,
+      "learning_rate": 8.997986361534805e-07,
+      "loss": 0.8827,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8682377049180328,
+      "grad_norm": 3.011056661605835,
+      "learning_rate": 8.970489161941786e-07,
+      "loss": 0.7792,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8684426229508196,
+      "grad_norm": 4.57463264465332,
+      "learning_rate": 8.943032068310009e-07,
+      "loss": 0.9906,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8686475409836065,
+      "grad_norm": 2.712808847427368,
+      "learning_rate": 8.915615092736529e-07,
+      "loss": 0.9796,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8688524590163934,
+      "grad_norm": 3.7064337730407715,
+      "learning_rate": 8.888238247300796e-07,
+      "loss": 1.1257,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8690573770491803,
+      "grad_norm": 2.8232269287109375,
+      "learning_rate": 8.860901544064538e-07,
+      "loss": 0.7667,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8692622950819672,
+      "grad_norm": 5.074721813201904,
+      "learning_rate": 8.833604995071832e-07,
+      "loss": 1.0309,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8694672131147541,
+      "grad_norm": 3.6922755241394043,
+      "learning_rate": 8.806348612349036e-07,
+      "loss": 1.0711,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8696721311475409,
+      "grad_norm": 2.622828245162964,
+      "learning_rate": 8.779132407904822e-07,
+      "loss": 0.8534,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8698770491803278,
+      "grad_norm": 3.587320327758789,
+      "learning_rate": 8.751956393730177e-07,
+      "loss": 0.8871,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8700819672131147,
+      "grad_norm": 3.6165947914123535,
+      "learning_rate": 8.724820581798332e-07,
+      "loss": 0.9597,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8702868852459016,
+      "grad_norm": 2.491147041320801,
+      "learning_rate": 8.697724984064826e-07,
+      "loss": 0.7683,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8704918032786885,
+      "grad_norm": 3.6740379333496094,
+      "learning_rate": 8.67066961246753e-07,
+      "loss": 0.8269,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8706967213114755,
+      "grad_norm": 7.578855514526367,
+      "learning_rate": 8.643654478926556e-07,
+      "loss": 1.0865,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8709016393442623,
+      "grad_norm": 4.20536994934082,
+      "learning_rate": 8.616679595344258e-07,
+      "loss": 1.0236,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8711065573770492,
+      "grad_norm": 3.846184253692627,
+      "learning_rate": 8.589744973605285e-07,
+      "loss": 0.5509,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8713114754098361,
+      "grad_norm": 2.7244415283203125,
+      "learning_rate": 8.562850625576546e-07,
+      "loss": 0.9576,
+      "step": 4252
+    },
+    {
+      "epoch": 0.871516393442623,
+      "grad_norm": 4.407885551452637,
+      "learning_rate": 8.535996563107219e-07,
+      "loss": 1.0219,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8717213114754099,
+      "grad_norm": 4.787663459777832,
+      "learning_rate": 8.509182798028703e-07,
+      "loss": 0.9203,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8719262295081968,
+      "grad_norm": 12.138749122619629,
+      "learning_rate": 8.482409342154685e-07,
+      "loss": 1.2774,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8721311475409836,
+      "grad_norm": 3.947551965713501,
+      "learning_rate": 8.455676207281039e-07,
+      "loss": 0.8254,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8723360655737705,
+      "grad_norm": 6.859382629394531,
+      "learning_rate": 8.42898340518592e-07,
+      "loss": 0.6723,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8725409836065574,
+      "grad_norm": 3.0152511596679688,
+      "learning_rate": 8.402330947629688e-07,
+      "loss": 1.0353,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8727459016393443,
+      "grad_norm": 3.4101946353912354,
+      "learning_rate": 8.375718846354953e-07,
+      "loss": 0.8304,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8729508196721312,
+      "grad_norm": 2.8838725090026855,
+      "learning_rate": 8.349147113086487e-07,
+      "loss": 1.1457,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8731557377049181,
+      "grad_norm": 3.9371414184570312,
+      "learning_rate": 8.32261575953135e-07,
+      "loss": 0.9359,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8733606557377049,
+      "grad_norm": 3.9242265224456787,
+      "learning_rate": 8.296124797378747e-07,
+      "loss": 1.0202,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8735655737704918,
+      "grad_norm": 2.897430419921875,
+      "learning_rate": 8.269674238300174e-07,
+      "loss": 0.8878,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8737704918032787,
+      "grad_norm": 2.9408156871795654,
+      "learning_rate": 8.243264093949222e-07,
+      "loss": 0.7844,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8739754098360656,
+      "grad_norm": 10.044285774230957,
+      "learning_rate": 8.216894375961737e-07,
+      "loss": 0.9048,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8741803278688525,
+      "grad_norm": 2.898003101348877,
+      "learning_rate": 8.190565095955749e-07,
+      "loss": 0.7926,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8743852459016394,
+      "grad_norm": 5.022505283355713,
+      "learning_rate": 8.164276265531446e-07,
+      "loss": 0.9444,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8745901639344262,
+      "grad_norm": 5.474678993225098,
+      "learning_rate": 8.138027896271228e-07,
+      "loss": 0.9707,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8747950819672131,
+      "grad_norm": 3.4213764667510986,
+      "learning_rate": 8.111819999739656e-07,
+      "loss": 0.7875,
+      "step": 4269
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 2.2522640228271484,
+      "learning_rate": 8.085652587483461e-07,
+      "loss": 0.9577,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8752049180327869,
+      "grad_norm": 3.551923990249634,
+      "learning_rate": 8.059525671031509e-07,
+      "loss": 0.9066,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8754098360655738,
+      "grad_norm": 2.727128267288208,
+      "learning_rate": 8.033439261894849e-07,
+      "loss": 0.7594,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8756147540983606,
+      "grad_norm": 4.4722185134887695,
+      "learning_rate": 8.007393371566729e-07,
+      "loss": 0.9496,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8758196721311475,
+      "grad_norm": 3.1959474086761475,
+      "learning_rate": 7.98138801152244e-07,
+      "loss": 0.7102,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8760245901639344,
+      "grad_norm": 4.204996109008789,
+      "learning_rate": 7.955423193219503e-07,
+      "loss": 0.7447,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8762295081967213,
+      "grad_norm": 3.2371630668640137,
+      "learning_rate": 7.929498928097545e-07,
+      "loss": 0.9473,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8764344262295082,
+      "grad_norm": 3.7872681617736816,
+      "learning_rate": 7.903615227578331e-07,
+      "loss": 0.918,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8766393442622951,
+      "grad_norm": 3.449249505996704,
+      "learning_rate": 7.877772103065762e-07,
+      "loss": 0.9531,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8768442622950819,
+      "grad_norm": 2.4141488075256348,
+      "learning_rate": 7.851969565945827e-07,
+      "loss": 0.756,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8770491803278688,
+      "grad_norm": 3.2171998023986816,
+      "learning_rate": 7.826207627586701e-07,
+      "loss": 0.7009,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8772540983606557,
+      "grad_norm": 4.011334419250488,
+      "learning_rate": 7.800486299338561e-07,
+      "loss": 0.8991,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8774590163934426,
+      "grad_norm": 2.847374200820923,
+      "learning_rate": 7.77480559253383e-07,
+      "loss": 0.8384,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8776639344262295,
+      "grad_norm": 3.1740026473999023,
+      "learning_rate": 7.749165518486934e-07,
+      "loss": 0.6,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8778688524590164,
+      "grad_norm": 2.0363032817840576,
+      "learning_rate": 7.723566088494449e-07,
+      "loss": 0.7677,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8780737704918032,
+      "grad_norm": 5.897881984710693,
+      "learning_rate": 7.698007313834977e-07,
+      "loss": 0.6919,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8782786885245901,
+      "grad_norm": 2.862131118774414,
+      "learning_rate": 7.672489205769285e-07,
+      "loss": 0.9888,
+      "step": 4286
+    },
+    {
+      "epoch": 0.878483606557377,
+      "grad_norm": 4.9559550285339355,
+      "learning_rate": 7.647011775540192e-07,
+      "loss": 0.9698,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8786885245901639,
+      "grad_norm": 2.1168811321258545,
+      "learning_rate": 7.621575034372586e-07,
+      "loss": 0.6343,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8788934426229508,
+      "grad_norm": 5.097147464752197,
+      "learning_rate": 7.596178993473436e-07,
+      "loss": 0.8289,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8790983606557377,
+      "grad_norm": 4.8860063552856445,
+      "learning_rate": 7.57082366403179e-07,
+      "loss": 1.1769,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8793032786885245,
+      "grad_norm": 3.514216184616089,
+      "learning_rate": 7.545509057218736e-07,
+      "loss": 0.7606,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8795081967213115,
+      "grad_norm": 3.6102278232574463,
+      "learning_rate": 7.52023518418743e-07,
+      "loss": 0.7733,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8797131147540984,
+      "grad_norm": 3.828488349914551,
+      "learning_rate": 7.495002056073109e-07,
+      "loss": 0.9723,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8799180327868853,
+      "grad_norm": 5.068022727966309,
+      "learning_rate": 7.469809683993034e-07,
+      "loss": 1.0301,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8801229508196722,
+      "grad_norm": 2.542539596557617,
+      "learning_rate": 7.444658079046474e-07,
+      "loss": 0.7819,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8803278688524591,
+      "grad_norm": 3.769735336303711,
+      "learning_rate": 7.419547252314796e-07,
+      "loss": 0.9021,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8805327868852459,
+      "grad_norm": 4.089027404785156,
+      "learning_rate": 7.39447721486135e-07,
+      "loss": 0.7995,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8807377049180328,
+      "grad_norm": 2.9396820068359375,
+      "learning_rate": 7.369447977731614e-07,
+      "loss": 0.7519,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8809426229508197,
+      "grad_norm": 4.529244899749756,
+      "learning_rate": 7.344459551952954e-07,
+      "loss": 0.9411,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8811475409836066,
+      "grad_norm": 2.232880115509033,
+      "learning_rate": 7.31951194853483e-07,
+      "loss": 0.8091,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8813524590163935,
+      "grad_norm": 2.7286107540130615,
+      "learning_rate": 7.294605178468727e-07,
+      "loss": 0.9581,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8815573770491804,
+      "grad_norm": 3.438443183898926,
+      "learning_rate": 7.269739252728091e-07,
+      "loss": 0.991,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8817622950819672,
+      "grad_norm": 4.313810348510742,
+      "learning_rate": 7.244914182268425e-07,
+      "loss": 0.8372,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8819672131147541,
+      "grad_norm": 6.869193077087402,
+      "learning_rate": 7.220129978027202e-07,
+      "loss": 0.9145,
+      "step": 4304
+    },
+    {
+      "epoch": 0.882172131147541,
+      "grad_norm": 3.00911545753479,
+      "learning_rate": 7.195386650923908e-07,
+      "loss": 0.6856,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8823770491803279,
+      "grad_norm": 4.769862174987793,
+      "learning_rate": 7.170684211859947e-07,
+      "loss": 0.9138,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8825819672131148,
+      "grad_norm": 4.719786643981934,
+      "learning_rate": 7.146022671718833e-07,
+      "loss": 0.8224,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8827868852459017,
+      "grad_norm": 5.779906272888184,
+      "learning_rate": 7.121402041365999e-07,
+      "loss": 1.028,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8829918032786885,
+      "grad_norm": 3.017225980758667,
+      "learning_rate": 7.0968223316488e-07,
+      "loss": 0.803,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8831967213114754,
+      "grad_norm": 3.4387359619140625,
+      "learning_rate": 7.072283553396653e-07,
+      "loss": 0.9348,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8834016393442623,
+      "grad_norm": 4.548895835876465,
+      "learning_rate": 7.047785717420875e-07,
+      "loss": 1.1133,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8836065573770492,
+      "grad_norm": 3.3295810222625732,
+      "learning_rate": 7.023328834514786e-07,
+      "loss": 0.9787,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8838114754098361,
+      "grad_norm": 2.312019109725952,
+      "learning_rate": 6.998912915453659e-07,
+      "loss": 0.663,
+      "step": 4313
+    },
+    {
+      "epoch": 0.884016393442623,
+      "grad_norm": 4.496081829071045,
+      "learning_rate": 6.974537970994677e-07,
+      "loss": 0.7685,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8842213114754098,
+      "grad_norm": 2.897260904312134,
+      "learning_rate": 6.950204011877038e-07,
+      "loss": 0.8555,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8844262295081967,
+      "grad_norm": 3.3040060997009277,
+      "learning_rate": 6.925911048821798e-07,
+      "loss": 0.7354,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8846311475409836,
+      "grad_norm": 2.4401540756225586,
+      "learning_rate": 6.901659092532043e-07,
+      "loss": 0.8742,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8848360655737705,
+      "grad_norm": 3.698988437652588,
+      "learning_rate": 6.877448153692734e-07,
+      "loss": 1.0359,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8850409836065574,
+      "grad_norm": 2.8651652336120605,
+      "learning_rate": 6.853278242970784e-07,
+      "loss": 1.0493,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8852459016393442,
+      "grad_norm": 7.138483047485352,
+      "learning_rate": 6.829149371014987e-07,
+      "loss": 0.7578,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8854508196721311,
+      "grad_norm": 2.231654167175293,
+      "learning_rate": 6.805061548456104e-07,
+      "loss": 0.7535,
+      "step": 4321
+    },
+    {
+      "epoch": 0.885655737704918,
+      "grad_norm": 2.38155460357666,
+      "learning_rate": 6.781014785906815e-07,
+      "loss": 0.7218,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8858606557377049,
+      "grad_norm": 11.319297790527344,
+      "learning_rate": 6.757009093961664e-07,
+      "loss": 1.1866,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8860655737704918,
+      "grad_norm": 2.7098586559295654,
+      "learning_rate": 6.733044483197148e-07,
+      "loss": 1.0164,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8862704918032787,
+      "grad_norm": 5.368964672088623,
+      "learning_rate": 6.709120964171623e-07,
+      "loss": 1.1346,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8864754098360655,
+      "grad_norm": 5.857633113861084,
+      "learning_rate": 6.685238547425376e-07,
+      "loss": 0.8748,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8866803278688524,
+      "grad_norm": 2.406712293624878,
+      "learning_rate": 6.66139724348055e-07,
+      "loss": 0.9114,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8868852459016393,
+      "grad_norm": 3.9701149463653564,
+      "learning_rate": 6.637597062841217e-07,
+      "loss": 1.0617,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8870901639344262,
+      "grad_norm": 3.2293195724487305,
+      "learning_rate": 6.613838015993312e-07,
+      "loss": 1.0097,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8872950819672131,
+      "grad_norm": 6.343040943145752,
+      "learning_rate": 6.590120113404608e-07,
+      "loss": 0.8771,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8875,
+      "grad_norm": 4.930423736572266,
+      "learning_rate": 6.566443365524788e-07,
+      "loss": 0.8296,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8877049180327868,
+      "grad_norm": 2.96486496925354,
+      "learning_rate": 6.542807782785432e-07,
+      "loss": 1.1239,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8879098360655737,
+      "grad_norm": 10.225021362304688,
+      "learning_rate": 6.519213375599954e-07,
+      "loss": 1.0628,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8881147540983606,
+      "grad_norm": 2.4407777786254883,
+      "learning_rate": 6.495660154363592e-07,
+      "loss": 0.7158,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8883196721311475,
+      "grad_norm": 2.3587865829467773,
+      "learning_rate": 6.472148129453481e-07,
+      "loss": 0.6862,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8885245901639345,
+      "grad_norm": 4.163129806518555,
+      "learning_rate": 6.448677311228601e-07,
+      "loss": 1.0634,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8887295081967214,
+      "grad_norm": 2.2439463138580322,
+      "learning_rate": 6.42524771002977e-07,
+      "loss": 0.8873,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8889344262295082,
+      "grad_norm": 4.323290824890137,
+      "learning_rate": 6.401859336179639e-07,
+      "loss": 0.9376,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8891393442622951,
+      "grad_norm": 2.71176815032959,
+      "learning_rate": 6.378512199982723e-07,
+      "loss": 0.8092,
+      "step": 4339
+    },
+    {
+      "epoch": 0.889344262295082,
+      "grad_norm": 5.902782440185547,
+      "learning_rate": 6.355206311725359e-07,
+      "loss": 0.9131,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8895491803278689,
+      "grad_norm": 3.8830666542053223,
+      "learning_rate": 6.331941681675635e-07,
+      "loss": 0.9927,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8897540983606558,
+      "grad_norm": 5.710979461669922,
+      "learning_rate": 6.308718320083595e-07,
+      "loss": 0.9431,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8899590163934427,
+      "grad_norm": 5.031424045562744,
+      "learning_rate": 6.285536237181033e-07,
+      "loss": 0.9237,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8901639344262295,
+      "grad_norm": 5.557590007781982,
+      "learning_rate": 6.262395443181513e-07,
+      "loss": 0.6876,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8903688524590164,
+      "grad_norm": 3.155324935913086,
+      "learning_rate": 6.239295948280477e-07,
+      "loss": 1.137,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8905737704918033,
+      "grad_norm": 4.413278102874756,
+      "learning_rate": 6.216237762655142e-07,
+      "loss": 1.2103,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8907786885245902,
+      "grad_norm": 5.456820011138916,
+      "learning_rate": 6.193220896464536e-07,
+      "loss": 0.8833,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8909836065573771,
+      "grad_norm": 2.7438526153564453,
+      "learning_rate": 6.170245359849458e-07,
+      "loss": 0.8141,
+      "step": 4348
+    },
+    {
+      "epoch": 0.891188524590164,
+      "grad_norm": 3.6431968212127686,
+      "learning_rate": 6.147311162932534e-07,
+      "loss": 1.1196,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8913934426229508,
+      "grad_norm": 5.8861308097839355,
+      "learning_rate": 6.124418315818159e-07,
+      "loss": 0.9818,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8915983606557377,
+      "grad_norm": 2.5056824684143066,
+      "learning_rate": 6.101566828592476e-07,
+      "loss": 0.8007,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8918032786885246,
+      "grad_norm": 4.241052627563477,
+      "learning_rate": 6.078756711323475e-07,
+      "loss": 1.0991,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8920081967213115,
+      "grad_norm": 2.982138156890869,
+      "learning_rate": 6.055987974060873e-07,
+      "loss": 0.7553,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8922131147540984,
+      "grad_norm": 2.9501729011535645,
+      "learning_rate": 6.033260626836168e-07,
+      "loss": 0.8102,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8924180327868853,
+      "grad_norm": 3.2668159008026123,
+      "learning_rate": 6.010574679662606e-07,
+      "loss": 0.7738,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8926229508196721,
+      "grad_norm": 4.526737213134766,
+      "learning_rate": 5.987930142535192e-07,
+      "loss": 0.8909,
+      "step": 4356
+    },
+    {
+      "epoch": 0.892827868852459,
+      "grad_norm": 3.394904375076294,
+      "learning_rate": 5.965327025430756e-07,
+      "loss": 0.9567,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8930327868852459,
+      "grad_norm": 6.604520320892334,
+      "learning_rate": 5.942765338307776e-07,
+      "loss": 0.9995,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8932377049180328,
+      "grad_norm": 2.8316850662231445,
+      "learning_rate": 5.920245091106536e-07,
+      "loss": 0.919,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8934426229508197,
+      "grad_norm": 4.8346428871154785,
+      "learning_rate": 5.897766293749052e-07,
+      "loss": 0.7606,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8936475409836065,
+      "grad_norm": 2.749054193496704,
+      "learning_rate": 5.875328956139082e-07,
+      "loss": 0.7929,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8938524590163934,
+      "grad_norm": 4.311408042907715,
+      "learning_rate": 5.852933088162105e-07,
+      "loss": 1.0562,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8940573770491803,
+      "grad_norm": 5.221662521362305,
+      "learning_rate": 5.83057869968534e-07,
+      "loss": 0.8406,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8942622950819672,
+      "grad_norm": 3.3485958576202393,
+      "learning_rate": 5.808265800557744e-07,
+      "loss": 0.8596,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8944672131147541,
+      "grad_norm": 3.5793607234954834,
+      "learning_rate": 5.78599440060994e-07,
+      "loss": 0.9212,
+      "step": 4365
+    },
+    {
+      "epoch": 0.894672131147541,
+      "grad_norm": 3.658355712890625,
+      "learning_rate": 5.763764509654324e-07,
+      "loss": 0.9539,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8948770491803278,
+      "grad_norm": 2.9834511280059814,
+      "learning_rate": 5.741576137484994e-07,
+      "loss": 0.8969,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8950819672131147,
+      "grad_norm": 2.561755657196045,
+      "learning_rate": 5.719429293877755e-07,
+      "loss": 1.0379,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8952868852459016,
+      "grad_norm": 5.059043884277344,
+      "learning_rate": 5.697323988590086e-07,
+      "loss": 0.9236,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8954918032786885,
+      "grad_norm": 8.002593994140625,
+      "learning_rate": 5.675260231361191e-07,
+      "loss": 1.1244,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8956967213114754,
+      "grad_norm": 4.277309417724609,
+      "learning_rate": 5.653238031911956e-07,
+      "loss": 0.9296,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8959016393442623,
+      "grad_norm": 3.817937135696411,
+      "learning_rate": 5.631257399944989e-07,
+      "loss": 1.0605,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8961065573770491,
+      "grad_norm": 4.319692611694336,
+      "learning_rate": 5.609318345144533e-07,
+      "loss": 0.704,
+      "step": 4373
+    },
+    {
+      "epoch": 0.896311475409836,
+      "grad_norm": 3.980341911315918,
+      "learning_rate": 5.587420877176563e-07,
+      "loss": 0.8411,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8965163934426229,
+      "grad_norm": 3.5681779384613037,
+      "learning_rate": 5.5655650056887e-07,
+      "loss": 0.798,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8967213114754098,
+      "grad_norm": 4.095805644989014,
+      "learning_rate": 5.543750740310227e-07,
+      "loss": 0.6008,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8969262295081967,
+      "grad_norm": 3.07763409614563,
+      "learning_rate": 5.521978090652147e-07,
+      "loss": 0.6998,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8971311475409836,
+      "grad_norm": 2.4831175804138184,
+      "learning_rate": 5.500247066307096e-07,
+      "loss": 0.8413,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8973360655737705,
+      "grad_norm": 5.26639986038208,
+      "learning_rate": 5.478557676849339e-07,
+      "loss": 0.6437,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8975409836065574,
+      "grad_norm": 4.110195636749268,
+      "learning_rate": 5.456909931834853e-07,
+      "loss": 0.7662,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8977459016393443,
+      "grad_norm": 3.9909448623657227,
+      "learning_rate": 5.435303840801231e-07,
+      "loss": 0.9432,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8979508196721312,
+      "grad_norm": 6.648638725280762,
+      "learning_rate": 5.413739413267771e-07,
+      "loss": 1.0093,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8981557377049181,
+      "grad_norm": 3.7598612308502197,
+      "learning_rate": 5.392216658735316e-07,
+      "loss": 0.8636,
+      "step": 4383
+    },
+    {
+      "epoch": 0.898360655737705,
+      "grad_norm": 3.8490476608276367,
+      "learning_rate": 5.370735586686426e-07,
+      "loss": 0.7303,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8985655737704918,
+      "grad_norm": 2.691554546356201,
+      "learning_rate": 5.349296206585275e-07,
+      "loss": 0.9456,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8987704918032787,
+      "grad_norm": 3.178399085998535,
+      "learning_rate": 5.327898527877662e-07,
+      "loss": 0.9697,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8989754098360656,
+      "grad_norm": 3.0258851051330566,
+      "learning_rate": 5.306542559991035e-07,
+      "loss": 0.8399,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8991803278688525,
+      "grad_norm": 7.1988654136657715,
+      "learning_rate": 5.285228312334423e-07,
+      "loss": 0.8545,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8993852459016394,
+      "grad_norm": 4.80950403213501,
+      "learning_rate": 5.263955794298525e-07,
+      "loss": 0.8389,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8995901639344263,
+      "grad_norm": 4.240326881408691,
+      "learning_rate": 5.242725015255601e-07,
+      "loss": 0.7997,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8997950819672131,
+      "grad_norm": 3.1725289821624756,
+      "learning_rate": 5.221535984559556e-07,
+      "loss": 0.697,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.581289768218994,
+      "learning_rate": 5.200388711545934e-07,
+      "loss": 0.9052,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9002049180327869,
+      "grad_norm": 3.1451563835144043,
+      "learning_rate": 5.179283205531793e-07,
+      "loss": 0.6814,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9004098360655738,
+      "grad_norm": 2.9250080585479736,
+      "learning_rate": 5.158219475815851e-07,
+      "loss": 0.8535,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9006147540983607,
+      "grad_norm": 4.648228168487549,
+      "learning_rate": 5.137197531678417e-07,
+      "loss": 0.8591,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9008196721311476,
+      "grad_norm": 2.714620590209961,
+      "learning_rate": 5.116217382381373e-07,
+      "loss": 0.8764,
+      "step": 4396
+    },
+    {
+      "epoch": 0.9010245901639344,
+      "grad_norm": 3.1173088550567627,
+      "learning_rate": 5.095279037168211e-07,
+      "loss": 0.8289,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9012295081967213,
+      "grad_norm": 2.52093505859375,
+      "learning_rate": 5.074382505263964e-07,
+      "loss": 0.7233,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9014344262295082,
+      "grad_norm": 6.174352169036865,
+      "learning_rate": 5.05352779587529e-07,
+      "loss": 0.8708,
+      "step": 4399
+    },
+    {
+      "epoch": 0.9016393442622951,
+      "grad_norm": 3.6812257766723633,
+      "learning_rate": 5.032714918190384e-07,
+      "loss": 1.0108,
+      "step": 4400
+    },
+    {
+      "epoch": 0.901844262295082,
+      "grad_norm": 2.0849123001098633,
+      "learning_rate": 5.011943881379e-07,
+      "loss": 0.8228,
+      "step": 4401
+    },
+    {
+      "epoch": 0.9020491803278688,
+      "grad_norm": 3.871182918548584,
+      "learning_rate": 4.991214694592539e-07,
+      "loss": 0.8352,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9022540983606557,
+      "grad_norm": 4.464075565338135,
+      "learning_rate": 4.970527366963884e-07,
+      "loss": 0.9526,
+      "step": 4403
+    },
+    {
+      "epoch": 0.9024590163934426,
+      "grad_norm": 3.5963573455810547,
+      "learning_rate": 4.949881907607468e-07,
+      "loss": 0.8404,
+      "step": 4404
+    },
+    {
+      "epoch": 0.9026639344262295,
+      "grad_norm": 2.740741491317749,
+      "learning_rate": 4.929278325619347e-07,
+      "loss": 0.8377,
+      "step": 4405
+    },
+    {
+      "epoch": 0.9028688524590164,
+      "grad_norm": 4.1629462242126465,
+      "learning_rate": 4.908716630077071e-07,
+      "loss": 0.8722,
+      "step": 4406
+    },
+    {
+      "epoch": 0.9030737704918033,
+      "grad_norm": 4.666125774383545,
+      "learning_rate": 4.888196830039748e-07,
+      "loss": 0.8584,
+      "step": 4407
+    },
+    {
+      "epoch": 0.9032786885245901,
+      "grad_norm": 1.854633092880249,
+      "learning_rate": 4.867718934548038e-07,
+      "loss": 0.6216,
+      "step": 4408
+    },
+    {
+      "epoch": 0.903483606557377,
+      "grad_norm": 2.937845468521118,
+      "learning_rate": 4.847282952624143e-07,
+      "loss": 0.8916,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9036885245901639,
+      "grad_norm": 4.094014644622803,
+      "learning_rate": 4.826888893271764e-07,
+      "loss": 1.0708,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9038934426229508,
+      "grad_norm": 3.280808210372925,
+      "learning_rate": 4.806536765476177e-07,
+      "loss": 0.8739,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9040983606557377,
+      "grad_norm": 3.6552088260650635,
+      "learning_rate": 4.786226578204157e-07,
+      "loss": 0.9702,
+      "step": 4412
+    },
+    {
+      "epoch": 0.9043032786885246,
+      "grad_norm": 1.7652102708816528,
+      "learning_rate": 4.76595834040402e-07,
+      "loss": 0.746,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9045081967213114,
+      "grad_norm": 2.903916835784912,
+      "learning_rate": 4.745732061005548e-07,
+      "loss": 0.7879,
+      "step": 4414
+    },
+    {
+      "epoch": 0.9047131147540983,
+      "grad_norm": 5.96388053894043,
+      "learning_rate": 4.725547748920112e-07,
+      "loss": 0.9512,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9049180327868852,
+      "grad_norm": 4.104185104370117,
+      "learning_rate": 4.7054054130405204e-07,
+      "loss": 1.1325,
+      "step": 4416
+    },
+    {
+      "epoch": 0.9051229508196721,
+      "grad_norm": 3.638434648513794,
+      "learning_rate": 4.6853050622411635e-07,
+      "loss": 0.8355,
+      "step": 4417
+    },
+    {
+      "epoch": 0.905327868852459,
+      "grad_norm": 3.6936254501342773,
+      "learning_rate": 4.6652467053778685e-07,
+      "loss": 1.0549,
+      "step": 4418
+    },
+    {
+      "epoch": 0.9055327868852459,
+      "grad_norm": 2.5396084785461426,
+      "learning_rate": 4.645230351287999e-07,
+      "loss": 0.9117,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9057377049180327,
+      "grad_norm": 3.4736039638519287,
+      "learning_rate": 4.6252560087903795e-07,
+      "loss": 0.9644,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9059426229508196,
+      "grad_norm": 5.019174575805664,
+      "learning_rate": 4.6053236866853703e-07,
+      "loss": 0.9946,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9061475409836065,
+      "grad_norm": 4.458375930786133,
+      "learning_rate": 4.585433393754779e-07,
+      "loss": 0.8759,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9063524590163935,
+      "grad_norm": 4.1388020515441895,
+      "learning_rate": 4.5655851387619075e-07,
+      "loss": 0.8796,
+      "step": 4423
+    },
+    {
+      "epoch": 0.9065573770491804,
+      "grad_norm": 5.267240524291992,
+      "learning_rate": 4.545778930451561e-07,
+      "loss": 1.0231,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9067622950819673,
+      "grad_norm": 2.6500368118286133,
+      "learning_rate": 4.5260147775499586e-07,
+      "loss": 1.0582,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9069672131147541,
+      "grad_norm": 4.607022762298584,
+      "learning_rate": 4.506292688764846e-07,
+      "loss": 0.7304,
+      "step": 4426
+    },
+    {
+      "epoch": 0.907172131147541,
+      "grad_norm": 3.7789947986602783,
+      "learning_rate": 4.486612672785451e-07,
+      "loss": 0.9297,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9073770491803279,
+      "grad_norm": 2.318162441253662,
+      "learning_rate": 4.4669747382823923e-07,
+      "loss": 0.8392,
+      "step": 4428
+    },
+    {
+      "epoch": 0.9075819672131148,
+      "grad_norm": 3.29514741897583,
+      "learning_rate": 4.447378893907817e-07,
+      "loss": 0.9714,
+      "step": 4429
+    },
+    {
+      "epoch": 0.9077868852459017,
+      "grad_norm": 6.369581699371338,
+      "learning_rate": 4.427825148295273e-07,
+      "loss": 0.9302,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9079918032786886,
+      "grad_norm": 3.181537628173828,
+      "learning_rate": 4.408313510059814e-07,
+      "loss": 1.1174,
+      "step": 4431
+    },
+    {
+      "epoch": 0.9081967213114754,
+      "grad_norm": 2.2482476234436035,
+      "learning_rate": 4.3888439877979086e-07,
+      "loss": 0.6856,
+      "step": 4432
+    },
+    {
+      "epoch": 0.9084016393442623,
+      "grad_norm": 2.057892322540283,
+      "learning_rate": 4.369416590087472e-07,
+      "loss": 0.7875,
+      "step": 4433
+    },
+    {
+      "epoch": 0.9086065573770492,
+      "grad_norm": 2.9503884315490723,
+      "learning_rate": 4.35003132548788e-07,
+      "loss": 0.9357,
+      "step": 4434
+    },
+    {
+      "epoch": 0.9088114754098361,
+      "grad_norm": 3.752005100250244,
+      "learning_rate": 4.330688202539901e-07,
+      "loss": 0.9389,
+      "step": 4435
+    },
+    {
+      "epoch": 0.909016393442623,
+      "grad_norm": 2.577587366104126,
+      "learning_rate": 4.3113872297657955e-07,
+      "loss": 0.9722,
+      "step": 4436
+    },
+    {
+      "epoch": 0.9092213114754099,
+      "grad_norm": 2.4445502758026123,
+      "learning_rate": 4.292128415669206e-07,
+      "loss": 0.691,
+      "step": 4437
+    },
+    {
+      "epoch": 0.9094262295081967,
+      "grad_norm": 3.19499135017395,
+      "learning_rate": 4.272911768735244e-07,
+      "loss": 1.0153,
+      "step": 4438
+    },
+    {
+      "epoch": 0.9096311475409836,
+      "grad_norm": 3.620959758758545,
+      "learning_rate": 4.2537372974303824e-07,
+      "loss": 0.9755,
+      "step": 4439
+    },
+    {
+      "epoch": 0.9098360655737705,
+      "grad_norm": 4.954711437225342,
+      "learning_rate": 4.234605010202575e-07,
+      "loss": 0.98,
+      "step": 4440
+    },
+    {
+      "epoch": 0.9100409836065574,
+      "grad_norm": 3.5385518074035645,
+      "learning_rate": 4.2155149154811446e-07,
+      "loss": 0.8957,
+      "step": 4441
+    },
+    {
+      "epoch": 0.9102459016393443,
+      "grad_norm": 3.6490910053253174,
+      "learning_rate": 4.1964670216768534e-07,
+      "loss": 0.7452,
+      "step": 4442
+    },
+    {
+      "epoch": 0.9104508196721312,
+      "grad_norm": 3.8589324951171875,
+      "learning_rate": 4.177461337181854e-07,
+      "loss": 0.6876,
+      "step": 4443
+    },
+    {
+      "epoch": 0.910655737704918,
+      "grad_norm": 2.916764497756958,
+      "learning_rate": 4.1584978703697154e-07,
+      "loss": 0.8119,
+      "step": 4444
+    },
+    {
+      "epoch": 0.9108606557377049,
+      "grad_norm": 3.6920530796051025,
+      "learning_rate": 4.1395766295953875e-07,
+      "loss": 0.9529,
+      "step": 4445
+    },
+    {
+      "epoch": 0.9110655737704918,
+      "grad_norm": 6.170463562011719,
+      "learning_rate": 4.1206976231952464e-07,
+      "loss": 0.9955,
+      "step": 4446
+    },
+    {
+      "epoch": 0.9112704918032787,
+      "grad_norm": 6.263040542602539,
+      "learning_rate": 4.1018608594870277e-07,
+      "loss": 0.9947,
+      "step": 4447
+    },
+    {
+      "epoch": 0.9114754098360656,
+      "grad_norm": 4.8011603355407715,
+      "learning_rate": 4.0830663467698817e-07,
+      "loss": 1.0522,
+      "step": 4448
+    },
+    {
+      "epoch": 0.9116803278688524,
+      "grad_norm": 2.8222429752349854,
+      "learning_rate": 4.064314093324306e-07,
+      "loss": 0.7586,
+      "step": 4449
+    },
+    {
+      "epoch": 0.9118852459016393,
+      "grad_norm": 5.497125625610352,
+      "learning_rate": 4.045604107412238e-07,
+      "loss": 0.921,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9120901639344262,
+      "grad_norm": 3.24994158744812,
+      "learning_rate": 4.026936397276926e-07,
+      "loss": 0.9401,
+      "step": 4451
+    },
+    {
+      "epoch": 0.9122950819672131,
+      "grad_norm": 2.8184304237365723,
+      "learning_rate": 4.00831097114307e-07,
+      "loss": 0.6581,
+      "step": 4452
+    },
+    {
+      "epoch": 0.9125,
+      "grad_norm": 3.371417999267578,
+      "learning_rate": 3.9897278372166505e-07,
+      "loss": 0.893,
+      "step": 4453
+    },
+    {
+      "epoch": 0.9127049180327869,
+      "grad_norm": 3.633577585220337,
+      "learning_rate": 3.971187003685095e-07,
+      "loss": 1.0697,
+      "step": 4454
+    },
+    {
+      "epoch": 0.9129098360655737,
+      "grad_norm": 3.6913952827453613,
+      "learning_rate": 3.952688478717148e-07,
+      "loss": 0.9153,
+      "step": 4455
+    },
+    {
+      "epoch": 0.9131147540983606,
+      "grad_norm": 3.951828718185425,
+      "learning_rate": 3.934232270462923e-07,
+      "loss": 0.8397,
+      "step": 4456
+    },
+    {
+      "epoch": 0.9133196721311475,
+      "grad_norm": 5.955436706542969,
+      "learning_rate": 3.9158183870538936e-07,
+      "loss": 0.8842,
+      "step": 4457
+    },
+    {
+      "epoch": 0.9135245901639344,
+      "grad_norm": 2.7566213607788086,
+      "learning_rate": 3.8974468366028917e-07,
+      "loss": 0.9904,
+      "step": 4458
+    },
+    {
+      "epoch": 0.9137295081967213,
+      "grad_norm": 4.221458911895752,
+      "learning_rate": 3.8791176272040986e-07,
+      "loss": 0.6624,
+      "step": 4459
+    },
+    {
+      "epoch": 0.9139344262295082,
+      "grad_norm": 2.4845428466796875,
+      "learning_rate": 3.860830766932999e-07,
+      "loss": 0.9809,
+      "step": 4460
+    },
+    {
+      "epoch": 0.914139344262295,
+      "grad_norm": 4.698200225830078,
+      "learning_rate": 3.8425862638464816e-07,
+      "loss": 0.8853,
+      "step": 4461
+    },
+    {
+      "epoch": 0.9143442622950819,
+      "grad_norm": 3.2094428539276123,
+      "learning_rate": 3.824384125982761e-07,
+      "loss": 0.9788,
+      "step": 4462
+    },
+    {
+      "epoch": 0.9145491803278688,
+      "grad_norm": 4.219266891479492,
+      "learning_rate": 3.8062243613613437e-07,
+      "loss": 0.8547,
+      "step": 4463
+    },
+    {
+      "epoch": 0.9147540983606557,
+      "grad_norm": 3.227492332458496,
+      "learning_rate": 3.7881069779830857e-07,
+      "loss": 1.0022,
+      "step": 4464
+    },
+    {
+      "epoch": 0.9149590163934426,
+      "grad_norm": 4.257427215576172,
+      "learning_rate": 3.7700319838302026e-07,
+      "loss": 1.0562,
+      "step": 4465
+    },
+    {
+      "epoch": 0.9151639344262295,
+      "grad_norm": 4.543182849884033,
+      "learning_rate": 3.7519993868662117e-07,
+      "loss": 1.0664,
+      "step": 4466
+    },
+    {
+      "epoch": 0.9153688524590164,
+      "grad_norm": 4.272963047027588,
+      "learning_rate": 3.734009195035926e-07,
+      "loss": 0.905,
+      "step": 4467
+    },
+    {
+      "epoch": 0.9155737704918033,
+      "grad_norm": 3.5559399127960205,
+      "learning_rate": 3.716061416265526e-07,
+      "loss": 0.9995,
+      "step": 4468
+    },
+    {
+      "epoch": 0.9157786885245902,
+      "grad_norm": 3.785698175430298,
+      "learning_rate": 3.6981560584624765e-07,
+      "loss": 0.8548,
+      "step": 4469
+    },
+    {
+      "epoch": 0.9159836065573771,
+      "grad_norm": 2.314222574234009,
+      "learning_rate": 3.680293129515533e-07,
+      "loss": 0.9098,
+      "step": 4470
+    },
+    {
+      "epoch": 0.916188524590164,
+      "grad_norm": 2.5633578300476074,
+      "learning_rate": 3.6624726372948003e-07,
+      "loss": 0.9656,
+      "step": 4471
+    },
+    {
+      "epoch": 0.9163934426229509,
+      "grad_norm": 6.531082630157471,
+      "learning_rate": 3.644694589651654e-07,
+      "loss": 1.1393,
+      "step": 4472
+    },
+    {
+      "epoch": 0.9165983606557377,
+      "grad_norm": 6.357630252838135,
+      "learning_rate": 3.6269589944188056e-07,
+      "loss": 0.9358,
+      "step": 4473
+    },
+    {
+      "epoch": 0.9168032786885246,
+      "grad_norm": 4.618860721588135,
+      "learning_rate": 3.6092658594102157e-07,
+      "loss": 1.2475,
+      "step": 4474
+    },
+    {
+      "epoch": 0.9170081967213115,
+      "grad_norm": 3.6219661235809326,
+      "learning_rate": 3.5916151924211606e-07,
+      "loss": 0.9269,
+      "step": 4475
+    },
+    {
+      "epoch": 0.9172131147540984,
+      "grad_norm": 2.32619571685791,
+      "learning_rate": 3.574007001228208e-07,
+      "loss": 0.8642,
+      "step": 4476
+    },
+    {
+      "epoch": 0.9174180327868853,
+      "grad_norm": 3.4860427379608154,
+      "learning_rate": 3.556441293589208e-07,
+      "loss": 0.713,
+      "step": 4477
+    },
+    {
+      "epoch": 0.9176229508196722,
+      "grad_norm": 2.799196720123291,
+      "learning_rate": 3.538918077243303e-07,
+      "loss": 0.69,
+      "step": 4478
+    },
+    {
+      "epoch": 0.917827868852459,
+      "grad_norm": 2.658271551132202,
+      "learning_rate": 3.521437359910906e-07,
+      "loss": 0.7164,
+      "step": 4479
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 4.650444507598877,
+      "learning_rate": 3.503999149293702e-07,
+      "loss": 0.9676,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9182377049180328,
+      "grad_norm": 6.008047580718994,
+      "learning_rate": 3.4866034530746553e-07,
+      "loss": 0.8893,
+      "step": 4481
+    },
+    {
+      "epoch": 0.9184426229508197,
+      "grad_norm": 2.5898609161376953,
+      "learning_rate": 3.4692502789179796e-07,
+      "loss": 0.8364,
+      "step": 4482
+    },
+    {
+      "epoch": 0.9186475409836066,
+      "grad_norm": 3.621811628341675,
+      "learning_rate": 3.451939634469215e-07,
+      "loss": 0.8781,
+      "step": 4483
+    },
+    {
+      "epoch": 0.9188524590163935,
+      "grad_norm": 2.6498959064483643,
+      "learning_rate": 3.43467152735506e-07,
+      "loss": 0.8048,
+      "step": 4484
+    },
+    {
+      "epoch": 0.9190573770491803,
+      "grad_norm": 3.845932960510254,
+      "learning_rate": 3.4174459651835524e-07,
+      "loss": 0.9918,
+      "step": 4485
+    },
+    {
+      "epoch": 0.9192622950819672,
+      "grad_norm": 1.844369649887085,
+      "learning_rate": 3.400262955543987e-07,
+      "loss": 0.7198,
+      "step": 4486
+    },
+    {
+      "epoch": 0.9194672131147541,
+      "grad_norm": 2.5590286254882812,
+      "learning_rate": 3.3831225060068863e-07,
+      "loss": 0.9855,
+      "step": 4487
+    },
+    {
+      "epoch": 0.919672131147541,
+      "grad_norm": 2.78468918800354,
+      "learning_rate": 3.366024624123987e-07,
+      "loss": 1.0242,
+      "step": 4488
+    },
+    {
+      "epoch": 0.9198770491803279,
+      "grad_norm": 5.274322509765625,
+      "learning_rate": 3.3489693174283413e-07,
+      "loss": 0.8586,
+      "step": 4489
+    },
+    {
+      "epoch": 0.9200819672131147,
+      "grad_norm": 6.9044189453125,
+      "learning_rate": 3.3319565934341937e-07,
+      "loss": 1.17,
+      "step": 4490
+    },
+    {
+      "epoch": 0.9202868852459016,
+      "grad_norm": 2.6276257038116455,
+      "learning_rate": 3.3149864596370483e-07,
+      "loss": 0.7498,
+      "step": 4491
+    },
+    {
+      "epoch": 0.9204918032786885,
+      "grad_norm": 2.4452266693115234,
+      "learning_rate": 3.298058923513647e-07,
+      "loss": 0.8823,
+      "step": 4492
+    },
+    {
+      "epoch": 0.9206967213114754,
+      "grad_norm": 3.6582906246185303,
+      "learning_rate": 3.281173992521947e-07,
+      "loss": 0.8747,
+      "step": 4493
+    },
+    {
+      "epoch": 0.9209016393442623,
+      "grad_norm": 3.0947391986846924,
+      "learning_rate": 3.264331674101151e-07,
+      "loss": 0.8578,
+      "step": 4494
+    },
+    {
+      "epoch": 0.9211065573770492,
+      "grad_norm": 3.7377142906188965,
+      "learning_rate": 3.2475319756716585e-07,
+      "loss": 0.8952,
+      "step": 4495
+    },
+    {
+      "epoch": 0.921311475409836,
+      "grad_norm": 2.414957284927368,
+      "learning_rate": 3.2307749046351475e-07,
+      "loss": 0.7162,
+      "step": 4496
+    },
+    {
+      "epoch": 0.9215163934426229,
+      "grad_norm": 5.142862796783447,
+      "learning_rate": 3.2140604683744805e-07,
+      "loss": 1.0647,
+      "step": 4497
+    },
+    {
+      "epoch": 0.9217213114754098,
+      "grad_norm": 4.998627662658691,
+      "learning_rate": 3.19738867425371e-07,
+      "loss": 1.0335,
+      "step": 4498
+    },
+    {
+      "epoch": 0.9219262295081967,
+      "grad_norm": 4.888106822967529,
+      "learning_rate": 3.1807595296181383e-07,
+      "loss": 1.0848,
+      "step": 4499
+    },
+    {
+      "epoch": 0.9221311475409836,
+      "grad_norm": 5.446826934814453,
+      "learning_rate": 3.164173041794272e-07,
+      "loss": 1.0132,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9223360655737705,
+      "grad_norm": 6.7133941650390625,
+      "learning_rate": 3.1476292180898206e-07,
+      "loss": 0.774,
+      "step": 4501
+    },
+    {
+      "epoch": 0.9225409836065573,
+      "grad_norm": 3.51406192779541,
+      "learning_rate": 3.131128065793676e-07,
+      "loss": 1.0802,
+      "step": 4502
+    },
+    {
+      "epoch": 0.9227459016393442,
+      "grad_norm": 3.4922678470611572,
+      "learning_rate": 3.1146695921759784e-07,
+      "loss": 0.9637,
+      "step": 4503
+    },
+    {
+      "epoch": 0.9229508196721311,
+      "grad_norm": 4.873763084411621,
+      "learning_rate": 3.098253804488016e-07,
+      "loss": 0.9191,
+      "step": 4504
+    },
+    {
+      "epoch": 0.923155737704918,
+      "grad_norm": 4.548062801361084,
+      "learning_rate": 3.0818807099622816e-07,
+      "loss": 1.1469,
+      "step": 4505
+    },
+    {
+      "epoch": 0.9233606557377049,
+      "grad_norm": 6.122345447540283,
+      "learning_rate": 3.065550315812471e-07,
+      "loss": 0.8698,
+      "step": 4506
+    },
+    {
+      "epoch": 0.9235655737704918,
+      "grad_norm": 3.507765054702759,
+      "learning_rate": 3.049262629233474e-07,
+      "loss": 0.8672,
+      "step": 4507
+    },
+    {
+      "epoch": 0.9237704918032786,
+      "grad_norm": 6.623987197875977,
+      "learning_rate": 3.0330176574013603e-07,
+      "loss": 1.1298,
+      "step": 4508
+    },
+    {
+      "epoch": 0.9239754098360655,
+      "grad_norm": 2.0997440814971924,
+      "learning_rate": 3.016815407473339e-07,
+      "loss": 0.893,
+      "step": 4509
+    },
+    {
+      "epoch": 0.9241803278688525,
+      "grad_norm": 3.097933769226074,
+      "learning_rate": 3.000655886587833e-07,
+      "loss": 0.6546,
+      "step": 4510
+    },
+    {
+      "epoch": 0.9243852459016394,
+      "grad_norm": 3.5387954711914062,
+      "learning_rate": 2.9845391018644696e-07,
+      "loss": 0.7973,
+      "step": 4511
+    },
+    {
+      "epoch": 0.9245901639344263,
+      "grad_norm": 2.776463270187378,
+      "learning_rate": 2.96846506040398e-07,
+      "loss": 0.9274,
+      "step": 4512
+    },
+    {
+      "epoch": 0.9247950819672132,
+      "grad_norm": 2.792287826538086,
+      "learning_rate": 2.95243376928831e-07,
+      "loss": 0.8588,
+      "step": 4513
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 3.602358102798462,
+      "learning_rate": 2.936445235580554e-07,
+      "loss": 0.8163,
+      "step": 4514
+    },
+    {
+      "epoch": 0.9252049180327869,
+      "grad_norm": 3.6945977210998535,
+      "learning_rate": 2.9204994663249664e-07,
+      "loss": 1.0857,
+      "step": 4515
+    },
+    {
+      "epoch": 0.9254098360655738,
+      "grad_norm": 3.5412230491638184,
+      "learning_rate": 2.904596468546983e-07,
+      "loss": 0.9878,
+      "step": 4516
+    },
+    {
+      "epoch": 0.9256147540983607,
+      "grad_norm": 3.544734001159668,
+      "learning_rate": 2.8887362492531656e-07,
+      "loss": 0.9287,
+      "step": 4517
+    },
+    {
+      "epoch": 0.9258196721311476,
+      "grad_norm": 6.865377426147461,
+      "learning_rate": 2.872918815431258e-07,
+      "loss": 0.7806,
+      "step": 4518
+    },
+    {
+      "epoch": 0.9260245901639345,
+      "grad_norm": 4.418528079986572,
+      "learning_rate": 2.8571441740501083e-07,
+      "loss": 0.6751,
+      "step": 4519
+    },
+    {
+      "epoch": 0.9262295081967213,
+      "grad_norm": 4.14500093460083,
+      "learning_rate": 2.841412332059734e-07,
+      "loss": 0.7086,
+      "step": 4520
+    },
+    {
+      "epoch": 0.9264344262295082,
+      "grad_norm": 7.506219863891602,
+      "learning_rate": 2.8257232963913363e-07,
+      "loss": 0.9953,
+      "step": 4521
+    },
+    {
+      "epoch": 0.9266393442622951,
+      "grad_norm": 7.611302852630615,
+      "learning_rate": 2.810077073957218e-07,
+      "loss": 1.0354,
+      "step": 4522
+    },
+    {
+      "epoch": 0.926844262295082,
+      "grad_norm": 4.15239953994751,
+      "learning_rate": 2.794473671650799e-07,
+      "loss": 0.999,
+      "step": 4523
+    },
+    {
+      "epoch": 0.9270491803278689,
+      "grad_norm": 3.839245080947876,
+      "learning_rate": 2.778913096346658e-07,
+      "loss": 1.0868,
+      "step": 4524
+    },
+    {
+      "epoch": 0.9272540983606558,
+      "grad_norm": 3.579706907272339,
+      "learning_rate": 2.763395354900522e-07,
+      "loss": 0.9302,
+      "step": 4525
+    },
+    {
+      "epoch": 0.9274590163934426,
+      "grad_norm": 3.3503947257995605,
+      "learning_rate": 2.747920454149222e-07,
+      "loss": 0.7797,
+      "step": 4526
+    },
+    {
+      "epoch": 0.9276639344262295,
+      "grad_norm": 4.7286152839660645,
+      "learning_rate": 2.7324884009107046e-07,
+      "loss": 0.911,
+      "step": 4527
+    },
+    {
+      "epoch": 0.9278688524590164,
+      "grad_norm": 5.791797637939453,
+      "learning_rate": 2.717099201984075e-07,
+      "loss": 1.0557,
+      "step": 4528
+    },
+    {
+      "epoch": 0.9280737704918033,
+      "grad_norm": 5.659786701202393,
+      "learning_rate": 2.701752864149543e-07,
+      "loss": 1.0569,
+      "step": 4529
+    },
+    {
+      "epoch": 0.9282786885245902,
+      "grad_norm": 3.8405568599700928,
+      "learning_rate": 2.686449394168378e-07,
+      "loss": 0.9838,
+      "step": 4530
+    },
+    {
+      "epoch": 0.928483606557377,
+      "grad_norm": 2.2700932025909424,
+      "learning_rate": 2.6711887987830756e-07,
+      "loss": 0.7643,
+      "step": 4531
+    },
+    {
+      "epoch": 0.9286885245901639,
+      "grad_norm": 6.868460178375244,
+      "learning_rate": 2.655971084717157e-07,
+      "loss": 1.096,
+      "step": 4532
+    },
+    {
+      "epoch": 0.9288934426229508,
+      "grad_norm": 6.097474098205566,
+      "learning_rate": 2.64079625867526e-07,
+      "loss": 1.0868,
+      "step": 4533
+    },
+    {
+      "epoch": 0.9290983606557377,
+      "grad_norm": 2.98313570022583,
+      "learning_rate": 2.625664327343136e-07,
+      "loss": 0.6413,
+      "step": 4534
+    },
+    {
+      "epoch": 0.9293032786885246,
+      "grad_norm": 5.57362174987793,
+      "learning_rate": 2.610575297387652e-07,
+      "loss": 0.7781,
+      "step": 4535
+    },
+    {
+      "epoch": 0.9295081967213115,
+      "grad_norm": 5.042677402496338,
+      "learning_rate": 2.59552917545679e-07,
+      "loss": 1.0898,
+      "step": 4536
+    },
+    {
+      "epoch": 0.9297131147540983,
+      "grad_norm": 4.901484489440918,
+      "learning_rate": 2.580525968179548e-07,
+      "loss": 1.2369,
+      "step": 4537
+    },
+    {
+      "epoch": 0.9299180327868852,
+      "grad_norm": 4.250878810882568,
+      "learning_rate": 2.5655656821661047e-07,
+      "loss": 1.0434,
+      "step": 4538
+    },
+    {
+      "epoch": 0.9301229508196721,
+      "grad_norm": 5.208902359008789,
+      "learning_rate": 2.5506483240076873e-07,
+      "loss": 0.7499,
+      "step": 4539
+    },
+    {
+      "epoch": 0.930327868852459,
+      "grad_norm": 5.192914009094238,
+      "learning_rate": 2.5357739002765947e-07,
+      "loss": 1.1687,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9305327868852459,
+      "grad_norm": 2.9843757152557373,
+      "learning_rate": 2.520942417526251e-07,
+      "loss": 0.7779,
+      "step": 4541
+    },
+    {
+      "epoch": 0.9307377049180328,
+      "grad_norm": 2.914138078689575,
+      "learning_rate": 2.506153882291129e-07,
+      "loss": 0.887,
+      "step": 4542
+    },
+    {
+      "epoch": 0.9309426229508196,
+      "grad_norm": 6.310726165771484,
+      "learning_rate": 2.4914083010868064e-07,
+      "loss": 1.1341,
+      "step": 4543
+    },
+    {
+      "epoch": 0.9311475409836065,
+      "grad_norm": 3.339491844177246,
+      "learning_rate": 2.4767056804098966e-07,
+      "loss": 1.1823,
+      "step": 4544
+    },
+    {
+      "epoch": 0.9313524590163934,
+      "grad_norm": 5.24036979675293,
+      "learning_rate": 2.4620460267381184e-07,
+      "loss": 1.0019,
+      "step": 4545
+    },
+    {
+      "epoch": 0.9315573770491803,
+      "grad_norm": 3.1798386573791504,
+      "learning_rate": 2.447429346530272e-07,
+      "loss": 0.9702,
+      "step": 4546
+    },
+    {
+      "epoch": 0.9317622950819672,
+      "grad_norm": 2.439377784729004,
+      "learning_rate": 2.4328556462261734e-07,
+      "loss": 0.7441,
+      "step": 4547
+    },
+    {
+      "epoch": 0.9319672131147541,
+      "grad_norm": 2.7697558403015137,
+      "learning_rate": 2.4183249322467404e-07,
+      "loss": 0.8894,
+      "step": 4548
+    },
+    {
+      "epoch": 0.9321721311475409,
+      "grad_norm": 4.803014278411865,
+      "learning_rate": 2.4038372109939533e-07,
+      "loss": 0.6057,
+      "step": 4549
+    },
+    {
+      "epoch": 0.9323770491803278,
+      "grad_norm": 4.430539131164551,
+      "learning_rate": 2.389392488850828e-07,
+      "loss": 0.8644,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9325819672131147,
+      "grad_norm": 4.692493915557861,
+      "learning_rate": 2.3749907721814626e-07,
+      "loss": 0.9995,
+      "step": 4551
+    },
+    {
+      "epoch": 0.9327868852459016,
+      "grad_norm": 2.6818017959594727,
+      "learning_rate": 2.3606320673309812e-07,
+      "loss": 1.0146,
+      "step": 4552
+    },
+    {
+      "epoch": 0.9329918032786885,
+      "grad_norm": 2.3294243812561035,
+      "learning_rate": 2.3463163806255906e-07,
+      "loss": 0.8727,
+      "step": 4553
+    },
+    {
+      "epoch": 0.9331967213114755,
+      "grad_norm": 3.45025634765625,
+      "learning_rate": 2.3320437183725008e-07,
+      "loss": 0.8664,
+      "step": 4554
+    },
+    {
+      "epoch": 0.9334016393442623,
+      "grad_norm": 3.76697039604187,
+      "learning_rate": 2.317814086859982e-07,
+      "loss": 0.8508,
+      "step": 4555
+    },
+    {
+      "epoch": 0.9336065573770492,
+      "grad_norm": 2.7648279666900635,
+      "learning_rate": 2.3036274923573742e-07,
+      "loss": 0.6285,
+      "step": 4556
+    },
+    {
+      "epoch": 0.9338114754098361,
+      "grad_norm": 2.4523303508758545,
+      "learning_rate": 2.289483941115056e-07,
+      "loss": 0.9669,
+      "step": 4557
+    },
+    {
+      "epoch": 0.934016393442623,
+      "grad_norm": 2.9447133541107178,
+      "learning_rate": 2.275383439364376e-07,
+      "loss": 0.8724,
+      "step": 4558
+    },
+    {
+      "epoch": 0.9342213114754099,
+      "grad_norm": 3.986558437347412,
+      "learning_rate": 2.2613259933177757e-07,
+      "loss": 0.8887,
+      "step": 4559
+    },
+    {
+      "epoch": 0.9344262295081968,
+      "grad_norm": 2.4745535850524902,
+      "learning_rate": 2.247311609168723e-07,
+      "loss": 0.7464,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9346311475409836,
+      "grad_norm": 2.687863826751709,
+      "learning_rate": 2.2333402930916904e-07,
+      "loss": 0.9432,
+      "step": 4561
+    },
+    {
+      "epoch": 0.9348360655737705,
+      "grad_norm": 3.9296038150787354,
+      "learning_rate": 2.2194120512421978e-07,
+      "loss": 0.9788,
+      "step": 4562
+    },
+    {
+      "epoch": 0.9350409836065574,
+      "grad_norm": 5.324974536895752,
+      "learning_rate": 2.2055268897567818e-07,
+      "loss": 0.7382,
+      "step": 4563
+    },
+    {
+      "epoch": 0.9352459016393443,
+      "grad_norm": 7.502002239227295,
+      "learning_rate": 2.1916848147529924e-07,
+      "loss": 0.5401,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9354508196721312,
+      "grad_norm": 3.2273049354553223,
+      "learning_rate": 2.1778858323293627e-07,
+      "loss": 0.9069,
+      "step": 4565
+    },
+    {
+      "epoch": 0.9356557377049181,
+      "grad_norm": 3.4945321083068848,
+      "learning_rate": 2.1641299485655188e-07,
+      "loss": 0.9801,
+      "step": 4566
+    },
+    {
+      "epoch": 0.9358606557377049,
+      "grad_norm": 3.921988010406494,
+      "learning_rate": 2.1504171695220567e-07,
+      "loss": 0.8015,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9360655737704918,
+      "grad_norm": 3.9128658771514893,
+      "learning_rate": 2.1367475012405547e-07,
+      "loss": 1.0751,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9362704918032787,
+      "grad_norm": 4.8767828941345215,
+      "learning_rate": 2.1231209497436288e-07,
+      "loss": 0.807,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9364754098360656,
+      "grad_norm": 5.62116813659668,
+      "learning_rate": 2.109537521034888e-07,
+      "loss": 0.8504,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9366803278688525,
+      "grad_norm": 3.62616229057312,
+      "learning_rate": 2.0959972210989664e-07,
+      "loss": 0.8127,
+      "step": 4571
+    },
+    {
+      "epoch": 0.9368852459016394,
+      "grad_norm": 5.948451519012451,
+      "learning_rate": 2.0825000559014595e-07,
+      "loss": 0.9368,
+      "step": 4572
+    },
+    {
+      "epoch": 0.9370901639344262,
+      "grad_norm": 7.157898902893066,
+      "learning_rate": 2.069046031388966e-07,
+      "loss": 1.1564,
+      "step": 4573
+    },
+    {
+      "epoch": 0.9372950819672131,
+      "grad_norm": 3.1072590351104736,
+      "learning_rate": 2.0556351534891107e-07,
+      "loss": 0.8548,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 5.619899749755859,
+      "learning_rate": 2.042267428110467e-07,
+      "loss": 0.8895,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9377049180327869,
+      "grad_norm": 4.109395980834961,
+      "learning_rate": 2.0289428611426244e-07,
+      "loss": 1.0139,
+      "step": 4576
+    },
+    {
+      "epoch": 0.9379098360655738,
+      "grad_norm": 4.942025184631348,
+      "learning_rate": 2.0156614584561307e-07,
+      "loss": 0.9705,
+      "step": 4577
+    },
+    {
+      "epoch": 0.9381147540983606,
+      "grad_norm": 5.086979866027832,
+      "learning_rate": 2.002423225902561e-07,
+      "loss": 1.1356,
+      "step": 4578
+    },
+    {
+      "epoch": 0.9383196721311475,
+      "grad_norm": 2.572826862335205,
+      "learning_rate": 1.9892281693144167e-07,
+      "loss": 0.8334,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9385245901639344,
+      "grad_norm": 4.914937973022461,
+      "learning_rate": 1.9760762945051915e-07,
+      "loss": 0.6685,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9387295081967213,
+      "grad_norm": 4.053006172180176,
+      "learning_rate": 1.962967607269406e-07,
+      "loss": 0.9191,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9389344262295082,
+      "grad_norm": 2.006136178970337,
+      "learning_rate": 1.9499021133824847e-07,
+      "loss": 0.9192,
+      "step": 4582
+    },
+    {
+      "epoch": 0.9391393442622951,
+      "grad_norm": 2.453728437423706,
+      "learning_rate": 1.936879818600834e-07,
+      "loss": 0.8375,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9393442622950819,
+      "grad_norm": 2.8720626831054688,
+      "learning_rate": 1.923900728661876e-07,
+      "loss": 0.9774,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9395491803278688,
+      "grad_norm": 2.365172863006592,
+      "learning_rate": 1.910964849283936e-07,
+      "loss": 0.8222,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9397540983606557,
+      "grad_norm": 2.50349760055542,
+      "learning_rate": 1.8980721861663554e-07,
+      "loss": 0.9287,
+      "step": 4586
+    },
+    {
+      "epoch": 0.9399590163934426,
+      "grad_norm": 2.9505858421325684,
+      "learning_rate": 1.885222744989379e-07,
+      "loss": 0.8176,
+      "step": 4587
+    },
+    {
+      "epoch": 0.9401639344262295,
+      "grad_norm": 5.0626983642578125,
+      "learning_rate": 1.8724165314142672e-07,
+      "loss": 0.7306,
+      "step": 4588
+    },
+    {
+      "epoch": 0.9403688524590164,
+      "grad_norm": 3.2956182956695557,
+      "learning_rate": 1.8596535510831958e-07,
+      "loss": 0.8287,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9405737704918032,
+      "grad_norm": 6.476056098937988,
+      "learning_rate": 1.846933809619278e-07,
+      "loss": 0.9019,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9407786885245901,
+      "grad_norm": 6.896029949188232,
+      "learning_rate": 1.8342573126266417e-07,
+      "loss": 1.2192,
+      "step": 4591
+    },
+    {
+      "epoch": 0.940983606557377,
+      "grad_norm": 4.420958518981934,
+      "learning_rate": 1.8216240656903195e-07,
+      "loss": 0.9116,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9411885245901639,
+      "grad_norm": 1.9351675510406494,
+      "learning_rate": 1.809034074376259e-07,
+      "loss": 0.6634,
+      "step": 4593
+    },
+    {
+      "epoch": 0.9413934426229508,
+      "grad_norm": 6.789204120635986,
+      "learning_rate": 1.7964873442314233e-07,
+      "loss": 1.1729,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9415983606557377,
+      "grad_norm": 3.9799907207489014,
+      "learning_rate": 1.7839838807836461e-07,
+      "loss": 0.7392,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9418032786885245,
+      "grad_norm": 2.943087100982666,
+      "learning_rate": 1.771523689541732e-07,
+      "loss": 0.8316,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9420081967213115,
+      "grad_norm": 6.501157283782959,
+      "learning_rate": 1.7591067759954338e-07,
+      "loss": 1.1698,
+      "step": 4597
+    },
+    {
+      "epoch": 0.9422131147540984,
+      "grad_norm": 4.3158955574035645,
+      "learning_rate": 1.746733145615409e-07,
+      "loss": 0.695,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9424180327868853,
+      "grad_norm": 4.294624328613281,
+      "learning_rate": 1.7344028038532522e-07,
+      "loss": 0.9108,
+      "step": 4599
+    },
+    {
+      "epoch": 0.9426229508196722,
+      "grad_norm": 2.995269536972046,
+      "learning_rate": 1.7221157561414958e-07,
+      "loss": 0.6342,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9428278688524591,
+      "grad_norm": 2.674894332885742,
+      "learning_rate": 1.7098720078935872e-07,
+      "loss": 0.9429,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9430327868852459,
+      "grad_norm": 2.361056089401245,
+      "learning_rate": 1.6976715645039222e-07,
+      "loss": 0.8455,
+      "step": 4602
+    },
+    {
+      "epoch": 0.9432377049180328,
+      "grad_norm": 3.0479769706726074,
+      "learning_rate": 1.6855144313477568e-07,
+      "loss": 0.7115,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9434426229508197,
+      "grad_norm": 2.290837287902832,
+      "learning_rate": 1.6734006137813284e-07,
+      "loss": 0.8364,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9436475409836066,
+      "grad_norm": 3.3416261672973633,
+      "learning_rate": 1.6613301171417573e-07,
+      "loss": 0.9702,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9438524590163935,
+      "grad_norm": 3.6003172397613525,
+      "learning_rate": 1.649302946747111e-07,
+      "loss": 0.5889,
+      "step": 4606
+    },
+    {
+      "epoch": 0.9440573770491804,
+      "grad_norm": 9.440694808959961,
+      "learning_rate": 1.6373191078963068e-07,
+      "loss": 0.9631,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9442622950819672,
+      "grad_norm": 4.17498254776001,
+      "learning_rate": 1.6253786058692323e-07,
+      "loss": 0.9032,
+      "step": 4608
+    },
+    {
+      "epoch": 0.9444672131147541,
+      "grad_norm": 2.7480885982513428,
+      "learning_rate": 1.613481445926657e-07,
+      "loss": 0.8518,
+      "step": 4609
+    },
+    {
+      "epoch": 0.944672131147541,
+      "grad_norm": 5.16364860534668,
+      "learning_rate": 1.6016276333102542e-07,
+      "loss": 1.064,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9448770491803279,
+      "grad_norm": 5.105764865875244,
+      "learning_rate": 1.5898171732425805e-07,
+      "loss": 0.8741,
+      "step": 4611
+    },
+    {
+      "epoch": 0.9450819672131148,
+      "grad_norm": 2.2665324211120605,
+      "learning_rate": 1.5780500709271395e-07,
+      "loss": 0.7503,
+      "step": 4612
+    },
+    {
+      "epoch": 0.9452868852459017,
+      "grad_norm": 3.420356512069702,
+      "learning_rate": 1.5663263315482957e-07,
+      "loss": 1.0617,
+      "step": 4613
+    },
+    {
+      "epoch": 0.9454918032786885,
+      "grad_norm": 2.1829419136047363,
+      "learning_rate": 1.554645960271306e-07,
+      "loss": 0.6591,
+      "step": 4614
+    },
+    {
+      "epoch": 0.9456967213114754,
+      "grad_norm": 2.5126211643218994,
+      "learning_rate": 1.5430089622423317e-07,
+      "loss": 0.6224,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9459016393442623,
+      "grad_norm": 2.5718908309936523,
+      "learning_rate": 1.5314153425884492e-07,
+      "loss": 0.7686,
+      "step": 4616
+    },
+    {
+      "epoch": 0.9461065573770492,
+      "grad_norm": 1.9899239540100098,
+      "learning_rate": 1.5198651064175728e-07,
+      "loss": 0.6228,
+      "step": 4617
+    },
+    {
+      "epoch": 0.9463114754098361,
+      "grad_norm": 2.821687936782837,
+      "learning_rate": 1.508358258818532e-07,
+      "loss": 0.9696,
+      "step": 4618
+    },
+    {
+      "epoch": 0.946516393442623,
+      "grad_norm": 3.0342941284179688,
+      "learning_rate": 1.4968948048610488e-07,
+      "loss": 0.8431,
+      "step": 4619
+    },
+    {
+      "epoch": 0.9467213114754098,
+      "grad_norm": 3.307262659072876,
+      "learning_rate": 1.4854747495956944e-07,
+      "loss": 1.026,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9469262295081967,
+      "grad_norm": 8.440423965454102,
+      "learning_rate": 1.4740980980539555e-07,
+      "loss": 0.8211,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9471311475409836,
+      "grad_norm": 3.6637074947357178,
+      "learning_rate": 1.4627648552481777e-07,
+      "loss": 0.9958,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9473360655737705,
+      "grad_norm": 6.30264949798584,
+      "learning_rate": 1.4514750261715894e-07,
+      "loss": 0.9247,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9475409836065574,
+      "grad_norm": 3.775164842605591,
+      "learning_rate": 1.440228615798256e-07,
+      "loss": 0.6836,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9477459016393442,
+      "grad_norm": 4.907685279846191,
+      "learning_rate": 1.4290256290831805e-07,
+      "loss": 0.8643,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9479508196721311,
+      "grad_norm": 3.127647638320923,
+      "learning_rate": 1.4178660709621707e-07,
+      "loss": 1.0518,
+      "step": 4626
+    },
+    {
+      "epoch": 0.948155737704918,
+      "grad_norm": 5.926332473754883,
+      "learning_rate": 1.4067499463519486e-07,
+      "loss": 0.9629,
+      "step": 4627
+    },
+    {
+      "epoch": 0.9483606557377049,
+      "grad_norm": 4.423453330993652,
+      "learning_rate": 1.3956772601500633e-07,
+      "loss": 0.7109,
+      "step": 4628
+    },
+    {
+      "epoch": 0.9485655737704918,
+      "grad_norm": 4.605277061462402,
+      "learning_rate": 1.3846480172349352e-07,
+      "loss": 0.9623,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9487704918032787,
+      "grad_norm": 3.1950745582580566,
+      "learning_rate": 1.373662222465866e-07,
+      "loss": 0.8541,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9489754098360655,
+      "grad_norm": 2.49467134475708,
+      "learning_rate": 1.3627198806829833e-07,
+      "loss": 0.9342,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9491803278688524,
+      "grad_norm": 4.040081977844238,
+      "learning_rate": 1.351820996707298e-07,
+      "loss": 1.1573,
+      "step": 4632
+    },
+    {
+      "epoch": 0.9493852459016393,
+      "grad_norm": 2.967435121536255,
+      "learning_rate": 1.340965575340658e-07,
+      "loss": 0.9804,
+      "step": 4633
+    },
+    {
+      "epoch": 0.9495901639344262,
+      "grad_norm": 3.9201622009277344,
+      "learning_rate": 1.3301536213657707e-07,
+      "loss": 0.9625,
+      "step": 4634
+    },
+    {
+      "epoch": 0.9497950819672131,
+      "grad_norm": 4.549825668334961,
+      "learning_rate": 1.31938513954617e-07,
+      "loss": 0.6441,
+      "step": 4635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.335994720458984,
+      "learning_rate": 1.308660134626283e-07,
+      "loss": 0.9326,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9502049180327868,
+      "grad_norm": 2.8594255447387695,
+      "learning_rate": 1.2979786113313408e-07,
+      "loss": 0.9588,
+      "step": 4637
+    },
+    {
+      "epoch": 0.9504098360655737,
+      "grad_norm": 3.2183258533477783,
+      "learning_rate": 1.2873405743674127e-07,
+      "loss": 0.7638,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9506147540983606,
+      "grad_norm": 4.761150360107422,
+      "learning_rate": 1.2767460284214495e-07,
+      "loss": 0.8177,
+      "step": 4639
+    },
+    {
+      "epoch": 0.9508196721311475,
+      "grad_norm": 4.081814765930176,
+      "learning_rate": 1.2661949781611948e-07,
+      "loss": 0.7972,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9510245901639345,
+      "grad_norm": 4.019655704498291,
+      "learning_rate": 1.2556874282352749e-07,
+      "loss": 0.8296,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9512295081967214,
+      "grad_norm": 6.327611446380615,
+      "learning_rate": 1.2452233832731197e-07,
+      "loss": 0.9201,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9514344262295082,
+      "grad_norm": 3.557035207748413,
+      "learning_rate": 1.234802847884997e-07,
+      "loss": 0.8926,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9516393442622951,
+      "grad_norm": 4.2079668045043945,
+      "learning_rate": 1.22442582666199e-07,
+      "loss": 0.9767,
+      "step": 4644
+    },
+    {
+      "epoch": 0.951844262295082,
+      "grad_norm": 5.119531631469727,
+      "learning_rate": 1.2140923241760637e-07,
+      "loss": 0.8324,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9520491803278689,
+      "grad_norm": 2.935218095779419,
+      "learning_rate": 1.203802344979932e-07,
+      "loss": 0.876,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9522540983606558,
+      "grad_norm": 4.890944004058838,
+      "learning_rate": 1.1935558936072123e-07,
+      "loss": 0.8849,
+      "step": 4647
+    },
+    {
+      "epoch": 0.9524590163934427,
+      "grad_norm": 4.358798980712891,
+      "learning_rate": 1.1833529745722827e-07,
+      "loss": 0.7986,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9526639344262295,
+      "grad_norm": 2.950639247894287,
+      "learning_rate": 1.1731935923703586e-07,
+      "loss": 0.5939,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9528688524590164,
+      "grad_norm": 3.908559799194336,
+      "learning_rate": 1.1630777514775038e-07,
+      "loss": 0.7097,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9530737704918033,
+      "grad_norm": 2.7039847373962402,
+      "learning_rate": 1.1530054563505644e-07,
+      "loss": 0.5788,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9532786885245902,
+      "grad_norm": 7.311082363128662,
+      "learning_rate": 1.1429767114272127e-07,
+      "loss": 1.0626,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9534836065573771,
+      "grad_norm": 4.319540500640869,
+      "learning_rate": 1.1329915211259368e-07,
+      "loss": 0.911,
+      "step": 4653
+    },
+    {
+      "epoch": 0.953688524590164,
+      "grad_norm": 2.84609317779541,
+      "learning_rate": 1.1230498898460173e-07,
+      "loss": 0.9643,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9538934426229508,
+      "grad_norm": 4.521946907043457,
+      "learning_rate": 1.113151821967573e-07,
+      "loss": 1.0468,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9540983606557377,
+      "grad_norm": 2.959608316421509,
+      "learning_rate": 1.1032973218515042e-07,
+      "loss": 0.8027,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9543032786885246,
+      "grad_norm": 4.4375715255737305,
+      "learning_rate": 1.0934863938395269e-07,
+      "loss": 0.8854,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9545081967213115,
+      "grad_norm": 2.6524710655212402,
+      "learning_rate": 1.0837190422541721e-07,
+      "loss": 1.0846,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9547131147540984,
+      "grad_norm": 2.889451503753662,
+      "learning_rate": 1.0739952713987423e-07,
+      "loss": 0.7607,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9549180327868853,
+      "grad_norm": 3.762242317199707,
+      "learning_rate": 1.0643150855573658e-07,
+      "loss": 1.0886,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9551229508196721,
+      "grad_norm": 3.442072629928589,
+      "learning_rate": 1.0546784889949536e-07,
+      "loss": 1.1102,
+      "step": 4661
+    },
+    {
+      "epoch": 0.955327868852459,
+      "grad_norm": 7.794597625732422,
+      "learning_rate": 1.0450854859572202e-07,
+      "loss": 1.046,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9555327868852459,
+      "grad_norm": 2.076500177383423,
+      "learning_rate": 1.0355360806706738e-07,
+      "loss": 0.814,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9557377049180328,
+      "grad_norm": 5.049490928649902,
+      "learning_rate": 1.0260302773425934e-07,
+      "loss": 1.2089,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9559426229508197,
+      "grad_norm": 2.6297507286071777,
+      "learning_rate": 1.0165680801610734e-07,
+      "loss": 0.9361,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9561475409836065,
+      "grad_norm": 3.5040817260742188,
+      "learning_rate": 1.0071494932949899e-07,
+      "loss": 0.9886,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9563524590163934,
+      "grad_norm": 2.9291086196899414,
+      "learning_rate": 9.977745208940015e-08,
+      "loss": 0.878,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9565573770491803,
+      "grad_norm": 4.0706048011779785,
+      "learning_rate": 9.884431670885486e-08,
+      "loss": 0.9901,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9567622950819672,
+      "grad_norm": 8.084559440612793,
+      "learning_rate": 9.791554359898647e-08,
+      "loss": 0.7318,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9569672131147541,
+      "grad_norm": 4.46732234954834,
+      "learning_rate": 9.699113316899544e-08,
+      "loss": 0.848,
+      "step": 4670
+    },
+    {
+      "epoch": 0.957172131147541,
+      "grad_norm": 3.963242292404175,
+      "learning_rate": 9.607108582615932e-08,
+      "loss": 0.9056,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9573770491803278,
+      "grad_norm": 5.31718111038208,
+      "learning_rate": 9.515540197583716e-08,
+      "loss": 0.7792,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9575819672131147,
+      "grad_norm": 5.615189552307129,
+      "learning_rate": 9.424408202145852e-08,
+      "loss": 0.9228,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9577868852459016,
+      "grad_norm": 4.72765588760376,
+      "learning_rate": 9.333712636453773e-08,
+      "loss": 0.9328,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9579918032786885,
+      "grad_norm": 4.9201884269714355,
+      "learning_rate": 9.243453540466185e-08,
+      "loss": 1.2397,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9581967213114754,
+      "grad_norm": 5.384622573852539,
+      "learning_rate": 9.153630953949722e-08,
+      "loss": 0.8148,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9584016393442623,
+      "grad_norm": 4.654952526092529,
+      "learning_rate": 9.064244916478504e-08,
+      "loss": 1.0563,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9586065573770491,
+      "grad_norm": 3.7940080165863037,
+      "learning_rate": 8.975295467434364e-08,
+      "loss": 1.0678,
+      "step": 4678
+    },
+    {
+      "epoch": 0.958811475409836,
+      "grad_norm": 3.912221670150757,
+      "learning_rate": 8.886782646006953e-08,
+      "loss": 0.8928,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9590163934426229,
+      "grad_norm": 10.505925178527832,
+      "learning_rate": 8.798706491193298e-08,
+      "loss": 0.8756,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9592213114754098,
+      "grad_norm": 1.8277795314788818,
+      "learning_rate": 8.71106704179836e-08,
+      "loss": 0.7098,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9594262295081967,
+      "grad_norm": 2.7744057178497314,
+      "learning_rate": 8.623864336434251e-08,
+      "loss": 0.6009,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9596311475409836,
+      "grad_norm": 3.7459352016448975,
+      "learning_rate": 8.537098413521127e-08,
+      "loss": 1.0686,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9598360655737705,
+      "grad_norm": 2.852245807647705,
+      "learning_rate": 8.4507693112863e-08,
+      "loss": 0.8958,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9600409836065574,
+      "grad_norm": 5.427455902099609,
+      "learning_rate": 8.364877067764899e-08,
+      "loss": 1.0278,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9602459016393443,
+      "grad_norm": 7.535678863525391,
+      "learning_rate": 8.279421720799541e-08,
+      "loss": 0.8863,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9604508196721312,
+      "grad_norm": 3.116666316986084,
+      "learning_rate": 8.194403308040333e-08,
+      "loss": 1.0056,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9606557377049181,
+      "grad_norm": 3.4681453704833984,
+      "learning_rate": 8.109821866944645e-08,
+      "loss": 0.7924,
+      "step": 4688
+    },
+    {
+      "epoch": 0.960860655737705,
+      "grad_norm": 3.7127208709716797,
+      "learning_rate": 8.02567743477778e-08,
+      "loss": 0.9698,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9610655737704918,
+      "grad_norm": 4.538241863250732,
+      "learning_rate": 7.941970048612191e-08,
+      "loss": 0.8598,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9612704918032787,
+      "grad_norm": 2.192434549331665,
+      "learning_rate": 7.858699745327825e-08,
+      "loss": 0.721,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9614754098360656,
+      "grad_norm": 3.853100299835205,
+      "learning_rate": 7.775866561612e-08,
+      "loss": 0.8706,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9616803278688525,
+      "grad_norm": 3.603015899658203,
+      "learning_rate": 7.693470533959746e-08,
+      "loss": 0.9118,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9618852459016394,
+      "grad_norm": 5.807964324951172,
+      "learning_rate": 7.611511698672913e-08,
+      "loss": 0.8696,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9620901639344263,
+      "grad_norm": 2.9474189281463623,
+      "learning_rate": 7.529990091861394e-08,
+      "loss": 0.9433,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9622950819672131,
+      "grad_norm": 5.455679893493652,
+      "learning_rate": 7.448905749441904e-08,
+      "loss": 1.0741,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9625,
+      "grad_norm": 3.456758499145508,
+      "learning_rate": 7.368258707139087e-08,
+      "loss": 1.0422,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9627049180327869,
+      "grad_norm": 4.173806667327881,
+      "learning_rate": 7.288049000484187e-08,
+      "loss": 0.9043,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9629098360655738,
+      "grad_norm": 3.12457275390625,
+      "learning_rate": 7.208276664816382e-08,
+      "loss": 0.8789,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9631147540983607,
+      "grad_norm": 2.5612223148345947,
+      "learning_rate": 7.128941735281891e-08,
+      "loss": 0.6765,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9633196721311476,
+      "grad_norm": 3.637901544570923,
+      "learning_rate": 7.050044246834198e-08,
+      "loss": 0.8726,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9635245901639344,
+      "grad_norm": 5.616108417510986,
+      "learning_rate": 6.971584234234164e-08,
+      "loss": 1.0791,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9637295081967213,
+      "grad_norm": 2.1211585998535156,
+      "learning_rate": 6.893561732049803e-08,
+      "loss": 0.7201,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9639344262295082,
+      "grad_norm": 3.2062880992889404,
+      "learning_rate": 6.815976774656507e-08,
+      "loss": 1.0844,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9641393442622951,
+      "grad_norm": 3.6530797481536865,
+      "learning_rate": 6.738829396236712e-08,
+      "loss": 0.7401,
+      "step": 4705
+    },
+    {
+      "epoch": 0.964344262295082,
+      "grad_norm": 5.900174140930176,
+      "learning_rate": 6.662119630780339e-08,
+      "loss": 1.021,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9645491803278688,
+      "grad_norm": 3.6601500511169434,
+      "learning_rate": 6.585847512084132e-08,
+      "loss": 0.8587,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9647540983606557,
+      "grad_norm": 2.6138594150543213,
+      "learning_rate": 6.510013073752208e-08,
+      "loss": 0.7777,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9649590163934426,
+      "grad_norm": 4.302557468414307,
+      "learning_rate": 6.434616349195955e-08,
+      "loss": 1.1946,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9651639344262295,
+      "grad_norm": 5.743391990661621,
+      "learning_rate": 6.359657371633798e-08,
+      "loss": 1.2459,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9653688524590164,
+      "grad_norm": 2.850720167160034,
+      "learning_rate": 6.285136174091322e-08,
+      "loss": 0.8041,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9655737704918033,
+      "grad_norm": 2.6544601917266846,
+      "learning_rate": 6.211052789401151e-08,
+      "loss": 0.9477,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9657786885245901,
+      "grad_norm": 2.0855460166931152,
+      "learning_rate": 6.137407250203064e-08,
+      "loss": 0.9251,
+      "step": 4713
+    },
+    {
+      "epoch": 0.965983606557377,
+      "grad_norm": 4.60994815826416,
+      "learning_rate": 6.064199588944108e-08,
+      "loss": 0.9182,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9661885245901639,
+      "grad_norm": 3.6065452098846436,
+      "learning_rate": 5.991429837878149e-08,
+      "loss": 0.8996,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9663934426229508,
+      "grad_norm": 8.280268669128418,
+      "learning_rate": 5.919098029066317e-08,
+      "loss": 0.8511,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9665983606557377,
+      "grad_norm": 8.249752044677734,
+      "learning_rate": 5.847204194376566e-08,
+      "loss": 0.883,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9668032786885246,
+      "grad_norm": 2.925647258758545,
+      "learning_rate": 5.775748365484224e-08,
+      "loss": 0.9206,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9670081967213114,
+      "grad_norm": 3.7475597858428955,
+      "learning_rate": 5.704730573871109e-08,
+      "loss": 0.7482,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9672131147540983,
+      "grad_norm": 6.2693634033203125,
+      "learning_rate": 5.634150850826747e-08,
+      "loss": 0.7899,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9674180327868852,
+      "grad_norm": 2.714207649230957,
+      "learning_rate": 5.564009227447042e-08,
+      "loss": 0.8214,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9676229508196721,
+      "grad_norm": 4.748673915863037,
+      "learning_rate": 5.4943057346352744e-08,
+      "loss": 0.9247,
+      "step": 4722
+    },
+    {
+      "epoch": 0.967827868852459,
+      "grad_norm": 3.5408010482788086,
+      "learning_rate": 5.425040403101434e-08,
+      "loss": 0.7305,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9680327868852459,
+      "grad_norm": 6.815944671630859,
+      "learning_rate": 5.356213263362664e-08,
+      "loss": 0.5199,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9682377049180327,
+      "grad_norm": 2.366973400115967,
+      "learning_rate": 5.287824345742931e-08,
+      "loss": 0.9776,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9684426229508196,
+      "grad_norm": 3.777381420135498,
+      "learning_rate": 5.219873680373133e-08,
+      "loss": 0.9692,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9686475409836065,
+      "grad_norm": 3.1239967346191406,
+      "learning_rate": 5.152361297191211e-08,
+      "loss": 0.854,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9688524590163935,
+      "grad_norm": 5.084742546081543,
+      "learning_rate": 5.085287225941704e-08,
+      "loss": 1.1262,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9690573770491804,
+      "grad_norm": 2.53395414352417,
+      "learning_rate": 5.018651496176419e-08,
+      "loss": 1.1156,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9692622950819673,
+      "grad_norm": 2.233175039291382,
+      "learning_rate": 4.9524541372537595e-08,
+      "loss": 0.8232,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9694672131147541,
+      "grad_norm": 3.2812530994415283,
+      "learning_rate": 4.886695178339173e-08,
+      "loss": 0.8496,
+      "step": 4731
+    },
+    {
+      "epoch": 0.969672131147541,
+      "grad_norm": 4.248039245605469,
+      "learning_rate": 4.82137464840482e-08,
+      "loss": 0.9258,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9698770491803279,
+      "grad_norm": 2.7110140323638916,
+      "learning_rate": 4.756492576229788e-08,
+      "loss": 0.6423,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9700819672131148,
+      "grad_norm": 3.909982204437256,
+      "learning_rate": 4.6920489903997666e-08,
+      "loss": 0.9199,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9702868852459017,
+      "grad_norm": 2.23545241355896,
+      "learning_rate": 4.628043919307712e-08,
+      "loss": 0.8274,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9704918032786886,
+      "grad_norm": 3.9910366535186768,
+      "learning_rate": 4.564477391152955e-08,
+      "loss": 0.8303,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9706967213114754,
+      "grad_norm": 3.70741605758667,
+      "learning_rate": 4.5013494339417596e-08,
+      "loss": 0.7495,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9709016393442623,
+      "grad_norm": 4.228809356689453,
+      "learning_rate": 4.4386600754872114e-08,
+      "loss": 1.2008,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9711065573770492,
+      "grad_norm": 5.735398769378662,
+      "learning_rate": 4.3764093434091045e-08,
+      "loss": 0.802,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9713114754098361,
+      "grad_norm": 3.1980271339416504,
+      "learning_rate": 4.3145972651339464e-08,
+      "loss": 0.812,
+      "step": 4740
+    },
+    {
+      "epoch": 0.971516393442623,
+      "grad_norm": 2.424659013748169,
+      "learning_rate": 4.253223867894951e-08,
+      "loss": 0.6726,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9717213114754099,
+      "grad_norm": 3.6123745441436768,
+      "learning_rate": 4.192289178732378e-08,
+      "loss": 0.8759,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9719262295081967,
+      "grad_norm": 5.6173553466796875,
+      "learning_rate": 4.131793224492753e-08,
+      "loss": 0.6408,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9721311475409836,
+      "grad_norm": 4.283719062805176,
+      "learning_rate": 4.071736031829532e-08,
+      "loss": 1.0061,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9723360655737705,
+      "grad_norm": 3.085566520690918,
+      "learning_rate": 4.012117627202883e-08,
+      "loss": 0.8929,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9725409836065574,
+      "grad_norm": 2.8841662406921387,
+      "learning_rate": 3.952938036879572e-08,
+      "loss": 0.699,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9727459016393443,
+      "grad_norm": 3.5454108715057373,
+      "learning_rate": 3.894197286933077e-08,
+      "loss": 1.052,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9729508196721312,
+      "grad_norm": 2.176743507385254,
+      "learning_rate": 3.835895403243472e-08,
+      "loss": 0.8043,
+      "step": 4748
+    },
+    {
+      "epoch": 0.973155737704918,
+      "grad_norm": 2.6868834495544434,
+      "learning_rate": 3.7780324114975454e-08,
+      "loss": 0.6374,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9733606557377049,
+      "grad_norm": 2.2345731258392334,
+      "learning_rate": 3.720608337188569e-08,
+      "loss": 0.7917,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9735655737704918,
+      "grad_norm": 5.357283115386963,
+      "learning_rate": 3.663623205616862e-08,
+      "loss": 0.9672,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9737704918032787,
+      "grad_norm": 3.169673442840576,
+      "learning_rate": 3.6070770418886736e-08,
+      "loss": 0.8066,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9739754098360656,
+      "grad_norm": 4.361995220184326,
+      "learning_rate": 3.55096987091752e-08,
+      "loss": 0.9033,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9741803278688524,
+      "grad_norm": 4.5290021896362305,
+      "learning_rate": 3.495301717423072e-08,
+      "loss": 0.7722,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9743852459016393,
+      "grad_norm": 4.325479507446289,
+      "learning_rate": 3.4400726059317105e-08,
+      "loss": 0.9941,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9745901639344262,
+      "grad_norm": 6.439953327178955,
+      "learning_rate": 3.385282560776415e-08,
+      "loss": 0.9778,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9747950819672131,
+      "grad_norm": 3.8114326000213623,
+      "learning_rate": 3.3309316060966545e-08,
+      "loss": 1.0433,
+      "step": 4757
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 3.3814871311187744,
+      "learning_rate": 3.277019765838607e-08,
+      "loss": 0.7989,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9752049180327869,
+      "grad_norm": 2.6158828735351562,
+      "learning_rate": 3.2235470637546064e-08,
+      "loss": 0.9184,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9754098360655737,
+      "grad_norm": 2.7317402362823486,
+      "learning_rate": 3.1705135234040284e-08,
+      "loss": 0.7014,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9756147540983606,
+      "grad_norm": 3.2760722637176514,
+      "learning_rate": 3.1179191681524055e-08,
+      "loss": 0.9424,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9758196721311475,
+      "grad_norm": 6.761127471923828,
+      "learning_rate": 3.065764021171758e-08,
+      "loss": 0.6882,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9760245901639344,
+      "grad_norm": 3.803311586380005,
+      "learning_rate": 3.0140481054408144e-08,
+      "loss": 1.0551,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9762295081967213,
+      "grad_norm": 2.6562976837158203,
+      "learning_rate": 2.962771443744572e-08,
+      "loss": 0.9337,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9764344262295082,
+      "grad_norm": 2.6537277698516846,
+      "learning_rate": 2.9119340586748478e-08,
+      "loss": 0.9656,
+      "step": 4765
+    },
+    {
+      "epoch": 0.976639344262295,
+      "grad_norm": 1.6345895528793335,
+      "learning_rate": 2.8615359726293923e-08,
+      "loss": 0.6269,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9768442622950819,
+      "grad_norm": 2.4821059703826904,
+      "learning_rate": 2.8115772078127767e-08,
+      "loss": 0.8131,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9770491803278688,
+      "grad_norm": 2.7893409729003906,
+      "learning_rate": 2.7620577862360608e-08,
+      "loss": 0.9572,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9772540983606557,
+      "grad_norm": 7.8701252937316895,
+      "learning_rate": 2.7129777297163484e-08,
+      "loss": 0.9665,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9774590163934426,
+      "grad_norm": 4.2750678062438965,
+      "learning_rate": 2.6643370598777862e-08,
+      "loss": 0.6844,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9776639344262295,
+      "grad_norm": 5.5780134201049805,
+      "learning_rate": 2.6161357981502323e-08,
+      "loss": 0.9477,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9778688524590164,
+      "grad_norm": 4.658289909362793,
+      "learning_rate": 2.5683739657704765e-08,
+      "loss": 1.118,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9780737704918033,
+      "grad_norm": 2.8530642986297607,
+      "learning_rate": 2.521051583781575e-08,
+      "loss": 0.8566,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9782786885245902,
+      "grad_norm": 2.8746492862701416,
+      "learning_rate": 2.4741686730327398e-08,
+      "loss": 0.7952,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9784836065573771,
+      "grad_norm": 3.1908652782440186,
+      "learning_rate": 2.4277252541797803e-08,
+      "loss": 1.0693,
+      "step": 4775
+    },
+    {
+      "epoch": 0.978688524590164,
+      "grad_norm": 2.6757678985595703,
+      "learning_rate": 2.3817213476848844e-08,
+      "loss": 0.7378,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9788934426229509,
+      "grad_norm": 3.542494535446167,
+      "learning_rate": 2.3361569738166168e-08,
+      "loss": 0.9489,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9790983606557377,
+      "grad_norm": 3.456127643585205,
+      "learning_rate": 2.291032152649697e-08,
+      "loss": 0.8631,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9793032786885246,
+      "grad_norm": 3.6162590980529785,
+      "learning_rate": 2.246346904065333e-08,
+      "loss": 0.7781,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9795081967213115,
+      "grad_norm": 1.8222012519836426,
+      "learning_rate": 2.2021012477509983e-08,
+      "loss": 0.5317,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9797131147540984,
+      "grad_norm": 5.526164531707764,
+      "learning_rate": 2.1582952032006556e-08,
+      "loss": 0.8967,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9799180327868853,
+      "grad_norm": 2.453303098678589,
+      "learning_rate": 2.11492878971431e-08,
+      "loss": 0.8231,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9801229508196722,
+      "grad_norm": 3.251835346221924,
+      "learning_rate": 2.0720020263985676e-08,
+      "loss": 0.9301,
+      "step": 4783
+    },
+    {
+      "epoch": 0.980327868852459,
+      "grad_norm": 3.0773613452911377,
+      "learning_rate": 2.0295149321660767e-08,
+      "loss": 0.9001,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9805327868852459,
+      "grad_norm": 3.146667242050171,
+      "learning_rate": 1.9874675257359754e-08,
+      "loss": 0.8403,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9807377049180328,
+      "grad_norm": 4.625775337219238,
+      "learning_rate": 1.945859825633556e-08,
+      "loss": 0.7867,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9809426229508197,
+      "grad_norm": 2.7848377227783203,
+      "learning_rate": 1.9046918501902655e-08,
+      "loss": 0.8345,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9811475409836066,
+      "grad_norm": 3.1589877605438232,
+      "learning_rate": 1.8639636175442622e-08,
+      "loss": 1.0147,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9813524590163935,
+      "grad_norm": 2.781991720199585,
+      "learning_rate": 1.8236751456394143e-08,
+      "loss": 0.9684,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9815573770491803,
+      "grad_norm": 3.8914635181427,
+      "learning_rate": 1.7838264522263004e-08,
+      "loss": 1.0998,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9817622950819672,
+      "grad_norm": 2.300490617752075,
+      "learning_rate": 1.744417554861433e-08,
+      "loss": 0.6995,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9819672131147541,
+      "grad_norm": 2.1424341201782227,
+      "learning_rate": 1.705448470907589e-08,
+      "loss": 0.8042,
+      "step": 4792
+    },
+    {
+      "epoch": 0.982172131147541,
+      "grad_norm": 7.690613269805908,
+      "learning_rate": 1.6669192175339245e-08,
+      "loss": 0.909,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9823770491803279,
+      "grad_norm": 3.3450422286987305,
+      "learning_rate": 1.6288298117157487e-08,
+      "loss": 0.9413,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9825819672131147,
+      "grad_norm": 4.2803778648376465,
+      "learning_rate": 1.591180270234638e-08,
+      "loss": 0.7507,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9827868852459016,
+      "grad_norm": 6.236266136169434,
+      "learning_rate": 1.5539706096781016e-08,
+      "loss": 0.9224,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9829918032786885,
+      "grad_norm": 2.5962448120117188,
+      "learning_rate": 1.5172008464402477e-08,
+      "loss": 0.7476,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9831967213114754,
+      "grad_norm": 5.268637180328369,
+      "learning_rate": 1.4808709967210067e-08,
+      "loss": 0.8486,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9834016393442623,
+      "grad_norm": 2.848156452178955,
+      "learning_rate": 1.444981076526797e-08,
+      "loss": 1.0385,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 4.494401454925537,
+      "learning_rate": 1.4095311016699698e-08,
+      "loss": 1.0128,
+      "step": 4800
+    },
+    {
+      "epoch": 0.983811475409836,
+      "grad_norm": 2.7215962409973145,
+      "learning_rate": 1.3745210877691428e-08,
+      "loss": 0.7003,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9840163934426229,
+      "grad_norm": 3.556650400161743,
+      "learning_rate": 1.3399510502491996e-08,
+      "loss": 1.032,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9842213114754098,
+      "grad_norm": 4.953823089599609,
+      "learning_rate": 1.305821004340957e-08,
+      "loss": 0.9073,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9844262295081967,
+      "grad_norm": 3.7985877990722656,
+      "learning_rate": 1.2721309650816082e-08,
+      "loss": 0.7725,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9846311475409836,
+      "grad_norm": 2.3749752044677734,
+      "learning_rate": 1.2388809473142805e-08,
+      "loss": 0.88,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9848360655737705,
+      "grad_norm": 2.3270392417907715,
+      "learning_rate": 1.2060709656883662e-08,
+      "loss": 0.7974,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9850409836065573,
+      "grad_norm": 3.475003719329834,
+      "learning_rate": 1.1737010346593025e-08,
+      "loss": 0.9223,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9852459016393442,
+      "grad_norm": 4.243857383728027,
+      "learning_rate": 1.1417711684887922e-08,
+      "loss": 0.8701,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9854508196721311,
+      "grad_norm": 2.637901544570923,
+      "learning_rate": 1.1102813812445823e-08,
+      "loss": 0.769,
+      "step": 4809
+    },
+    {
+      "epoch": 0.985655737704918,
+      "grad_norm": 5.288962364196777,
+      "learning_rate": 1.0792316868003527e-08,
+      "loss": 0.8228,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9858606557377049,
+      "grad_norm": 4.917966365814209,
+      "learning_rate": 1.0486220988361607e-08,
+      "loss": 0.9548,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9860655737704918,
+      "grad_norm": 3.4266860485076904,
+      "learning_rate": 1.0184526308379961e-08,
+      "loss": 0.7157,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9862704918032786,
+      "grad_norm": 4.5732011795043945,
+      "learning_rate": 9.887232960980042e-09,
+      "loss": 0.962,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9864754098360655,
+      "grad_norm": 3.226440191268921,
+      "learning_rate": 9.594341077143743e-09,
+      "loss": 0.8485,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9866803278688525,
+      "grad_norm": 7.331930637359619,
+      "learning_rate": 9.305850785914505e-09,
+      "loss": 0.7735,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9868852459016394,
+      "grad_norm": 4.497122764587402,
+      "learning_rate": 9.021762214395102e-09,
+      "loss": 1.2078,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9870901639344263,
+      "grad_norm": 2.340007781982422,
+      "learning_rate": 8.742075487749858e-09,
+      "loss": 0.5385,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9872950819672132,
+      "grad_norm": 2.663802146911621,
+      "learning_rate": 8.466790729203534e-09,
+      "loss": 0.8244,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9875,
+      "grad_norm": 4.254405975341797,
+      "learning_rate": 8.195908060042445e-09,
+      "loss": 1.2101,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9877049180327869,
+      "grad_norm": 2.953650712966919,
+      "learning_rate": 7.929427599613348e-09,
+      "loss": 0.8615,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9879098360655738,
+      "grad_norm": 2.53422474861145,
+      "learning_rate": 7.667349465320106e-09,
+      "loss": 0.7609,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9881147540983607,
+      "grad_norm": 6.219637870788574,
+      "learning_rate": 7.409673772632575e-09,
+      "loss": 0.6827,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9883196721311476,
+      "grad_norm": 3.104104518890381,
+      "learning_rate": 7.1564006350755e-09,
+      "loss": 0.6815,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9885245901639345,
+      "grad_norm": 2.444453716278076,
+      "learning_rate": 6.907530164238508e-09,
+      "loss": 0.7702,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9887295081967213,
+      "grad_norm": 2.3098790645599365,
+      "learning_rate": 6.6630624697683336e-09,
+      "loss": 0.8782,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9889344262295082,
+      "grad_norm": 2.6998300552368164,
+      "learning_rate": 6.422997659373265e-09,
+      "loss": 0.8293,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9891393442622951,
+      "grad_norm": 2.5616202354431152,
+      "learning_rate": 6.1873358388209185e-09,
+      "loss": 0.7233,
+      "step": 4827
+    },
+    {
+      "epoch": 0.989344262295082,
+      "grad_norm": 5.779038906097412,
+      "learning_rate": 5.956077111940462e-09,
+      "loss": 0.9436,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9895491803278689,
+      "grad_norm": 2.7276320457458496,
+      "learning_rate": 5.729221580620392e-09,
+      "loss": 0.8644,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9897540983606558,
+      "grad_norm": 8.714122772216797,
+      "learning_rate": 5.506769344808538e-09,
+      "loss": 1.2226,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9899590163934426,
+      "grad_norm": 2.842207193374634,
+      "learning_rate": 5.288720502513167e-09,
+      "loss": 0.8701,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9901639344262295,
+      "grad_norm": 2.646277666091919,
+      "learning_rate": 5.075075149804099e-09,
+      "loss": 0.4371,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9903688524590164,
+      "grad_norm": 3.015545606613159,
+      "learning_rate": 4.86583338080826e-09,
+      "loss": 0.929,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9905737704918033,
+      "grad_norm": 4.54213285446167,
+      "learning_rate": 4.660995287714132e-09,
+      "loss": 1.0154,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9907786885245902,
+      "grad_norm": 5.506108283996582,
+      "learning_rate": 4.460560960769522e-09,
+      "loss": 0.8116,
+      "step": 4835
+    },
+    {
+      "epoch": 0.990983606557377,
+      "grad_norm": 5.621287822723389,
+      "learning_rate": 4.264530488283791e-09,
+      "loss": 0.908,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9911885245901639,
+      "grad_norm": 3.8919970989227295,
+      "learning_rate": 4.072903956621188e-09,
+      "loss": 1.0572,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9913934426229508,
+      "grad_norm": 2.308781623840332,
+      "learning_rate": 3.885681450211954e-09,
+      "loss": 0.9571,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9915983606557377,
+      "grad_norm": 2.8791568279266357,
+      "learning_rate": 3.702863051541217e-09,
+      "loss": 0.9508,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9918032786885246,
+      "grad_norm": 3.523772954940796,
+      "learning_rate": 3.524448841156769e-09,
+      "loss": 0.7811,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9920081967213115,
+      "grad_norm": 3.263617992401123,
+      "learning_rate": 3.3504388976635104e-09,
+      "loss": 0.7103,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9922131147540983,
+      "grad_norm": 2.1949055194854736,
+      "learning_rate": 3.1808332977278923e-09,
+      "loss": 0.7986,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9924180327868852,
+      "grad_norm": 2.009014368057251,
+      "learning_rate": 3.0156321160734747e-09,
+      "loss": 0.6969,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9926229508196721,
+      "grad_norm": 2.987015962600708,
+      "learning_rate": 2.8548354254875897e-09,
+      "loss": 0.7656,
+      "step": 4844
+    },
+    {
+      "epoch": 0.992827868852459,
+      "grad_norm": 4.107148170471191,
+      "learning_rate": 2.6984432968124585e-09,
+      "loss": 0.7248,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9930327868852459,
+      "grad_norm": 6.07833194732666,
+      "learning_rate": 2.5464557989518524e-09,
+      "loss": 0.9024,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9932377049180328,
+      "grad_norm": 2.364089012145996,
+      "learning_rate": 2.398872998868873e-09,
+      "loss": 0.8997,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9934426229508196,
+      "grad_norm": 4.347482204437256,
+      "learning_rate": 2.2556949615870625e-09,
+      "loss": 1.0678,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9936475409836065,
+      "grad_norm": 5.096033096313477,
+      "learning_rate": 2.116921750185963e-09,
+      "loss": 0.9942,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9938524590163934,
+      "grad_norm": 2.6435739994049072,
+      "learning_rate": 1.982553425807776e-09,
+      "loss": 0.6645,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9940573770491803,
+      "grad_norm": 2.918893814086914,
+      "learning_rate": 1.8525900476529246e-09,
+      "loss": 0.9321,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9942622950819672,
+      "grad_norm": 6.757293701171875,
+      "learning_rate": 1.7270316729800507e-09,
+      "loss": 0.918,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9944672131147541,
+      "grad_norm": 2.336374282836914,
+      "learning_rate": 1.605878357108237e-09,
+      "loss": 0.8826,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9946721311475409,
+      "grad_norm": 2.692412853240967,
+      "learning_rate": 1.4891301534158964e-09,
+      "loss": 0.7343,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9948770491803278,
+      "grad_norm": 2.57889986038208,
+      "learning_rate": 1.3767871133396616e-09,
+      "loss": 0.9257,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9950819672131147,
+      "grad_norm": 2.9720041751861572,
+      "learning_rate": 1.2688492863766056e-09,
+      "loss": 0.8857,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9952868852459016,
+      "grad_norm": 10.444392204284668,
+      "learning_rate": 1.1653167200820215e-09,
+      "loss": 0.9992,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9954918032786885,
+      "grad_norm": 4.127003192901611,
+      "learning_rate": 1.0661894600694223e-09,
+      "loss": 1.0129,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9956967213114755,
+      "grad_norm": 3.8895480632781982,
+      "learning_rate": 9.714675500127613e-10,
+      "loss": 0.8695,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9959016393442623,
+      "grad_norm": 3.037191390991211,
+      "learning_rate": 8.811510316464322e-10,
+      "loss": 0.9351,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9961065573770492,
+      "grad_norm": 2.766434907913208,
+      "learning_rate": 7.952399447608284e-10,
+      "loss": 0.9524,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9963114754098361,
+      "grad_norm": 3.3028993606567383,
+      "learning_rate": 7.137343272067832e-10,
+      "loss": 1.0243,
+      "step": 4862
+    },
+    {
+      "epoch": 0.996516393442623,
+      "grad_norm": 3.4023854732513428,
+      "learning_rate": 6.366342148944604e-10,
+      "loss": 0.9381,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9967213114754099,
+      "grad_norm": 3.6413888931274414,
+      "learning_rate": 5.639396417922438e-10,
+      "loss": 0.7622,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9969262295081968,
+      "grad_norm": 3.0447328090667725,
+      "learning_rate": 4.956506399300676e-10,
+      "loss": 1.0399,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9971311475409836,
+      "grad_norm": 2.1884753704071045,
+      "learning_rate": 4.3176723939275524e-10,
+      "loss": 0.6419,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9973360655737705,
+      "grad_norm": 2.958228588104248,
+      "learning_rate": 3.722894683266809e-10,
+      "loss": 0.8349,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9975409836065574,
+      "grad_norm": 4.849521160125732,
+      "learning_rate": 3.1721735293643865e-10,
+      "loss": 0.7422,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9977459016393443,
+      "grad_norm": 3.4696919918060303,
+      "learning_rate": 2.6655091748706287e-10,
+      "loss": 1.2334,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9979508196721312,
+      "grad_norm": 3.3043882846832275,
+      "learning_rate": 2.2029018429958747e-10,
+      "loss": 0.9479,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9981557377049181,
+      "grad_norm": 3.434805393218994,
+      "learning_rate": 1.78435173756597e-10,
+      "loss": 0.9082,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9983606557377049,
+      "grad_norm": 2.563643217086792,
+      "learning_rate": 1.409859042988959e-10,
+      "loss": 0.7882,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9985655737704918,
+      "grad_norm": 2.6919634342193604,
+      "learning_rate": 1.0794239242661875e-10,
+      "loss": 0.9789,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9987704918032787,
+      "grad_norm": 3.4986417293548584,
+      "learning_rate": 7.93046526958996e-11,
+      "loss": 0.804,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9989754098360656,
+      "grad_norm": 2.0371313095092773,
+      "learning_rate": 5.5072697725533366e-11,
+      "loss": 0.5837,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9991803278688525,
+      "grad_norm": 3.0517611503601074,
+      "learning_rate": 3.5246538190314425e-11,
+      "loss": 0.7591,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9993852459016394,
+      "grad_norm": 4.439868450164795,
+      "learning_rate": 1.9826182827697992e-11,
+      "loss": 0.9439,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9995901639344262,
+      "grad_norm": 4.358758926391602,
+      "learning_rate": 8.811638430028524e-12,
+      "loss": 0.763,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9997950819672131,
+      "grad_norm": 4.495006084442139,
+      "learning_rate": 2.2029098500908617e-12,
+      "loss": 1.1564,
+      "step": 4879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.070992946624756,
+      "learning_rate": 0.0,
+      "loss": 0.9604,
+      "step": 4880
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4880,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.913704743522953e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}