diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4895 @@
+{
+  "best_metric": 0.9113475177304965,
+  "best_model_checkpoint": "xtreme_s_w2v2_minds14.en-US/checkpoint-660",
+  "epoch": 149.95238095238096,
+  "global_step": 750,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 2.641,
+      "step": 1
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 2.6452,
+      "step": 2
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 2.6486,
+      "step": 3
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 2.6441,
+      "step": 4
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 2.6357,
+      "step": 5
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 3.3131,
+      "step": 6
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.1e-05,
+      "loss": 2.6395,
+      "step": 7
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 2.6469,
+      "step": 8
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 2.6319,
+      "step": 9
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 2.6474,
+      "step": 10
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 3.3083,
+      "step": 11
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 2.6442,
+      "step": 12
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.9e-05,
+      "loss": 2.6387,
+      "step": 13
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.2e-05,
+      "loss": 2.6331,
+      "step": 14
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 2.6465,
+      "step": 15
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 3.3036,
+      "step": 16
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.1e-05,
+      "loss": 2.6447,
+      "step": 17
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 2.6328,
+      "step": 18
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 2.6381,
+      "step": 19
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 2.6482,
+      "step": 20
+    },
+    {
+      "epoch": 3.95,
+      "eval_accuracy": 0.07446808510638298,
+      "eval_f1": 0.02421503881357896,
+      "eval_loss": 2.642108917236328,
+      "eval_runtime": 17.4817,
+      "eval_samples_per_second": 16.131,
+      "eval_steps_per_second": 1.03,
+      "step": 20
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 3.306,
+      "step": 21
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 2.6445,
+      "step": 22
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 6.9e-05,
+      "loss": 2.6341,
+      "step": 23
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 2.637,
+      "step": 24
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.5e-05,
+      "loss": 2.6324,
+      "step": 25
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.8e-05,
+      "loss": 3.2998,
+      "step": 26
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 8.1e-05,
+      "loss": 2.6362,
+      "step": 27
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 8.4e-05,
+      "loss": 2.6354,
+      "step": 28
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 2.6384,
+      "step": 29
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 2.6309,
+      "step": 30
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 9.3e-05,
+      "loss": 3.2932,
+      "step": 31
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 2.6361,
+      "step": 32
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 9.9e-05,
+      "loss": 2.6396,
+      "step": 33
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.000102,
+      "loss": 2.6269,
+      "step": 34
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 2.6339,
+      "step": 35
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 3.2846,
+      "step": 36
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 2.6277,
+      "step": 37
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 2.6415,
+      "step": 38
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 0.000117,
+      "loss": 2.6375,
+      "step": 39
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 2.6292,
+      "step": 40
+    },
+    {
+      "epoch": 7.95,
+      "eval_accuracy": 0.08156028368794327,
+      "eval_f1": 0.010772833723653397,
+      "eval_loss": 2.635920524597168,
+      "eval_runtime": 16.2568,
+      "eval_samples_per_second": 17.347,
+      "eval_steps_per_second": 1.107,
+      "step": 40
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 3.2859,
+      "step": 41
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 2.6388,
+      "step": 42
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 0.000129,
+      "loss": 2.6264,
+      "step": 43
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 2.605,
+      "step": 44
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 0.000135,
+      "loss": 2.6417,
+      "step": 45
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 0.000138,
+      "loss": 3.2791,
+      "step": 46
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 2.622,
+      "step": 47
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 2.6282,
+      "step": 48
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 0.000147,
+      "loss": 2.6193,
+      "step": 49
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 0.00015,
+      "loss": 2.6165,
+      "step": 50
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 0.00015299999999999998,
+      "loss": 3.2538,
+      "step": 51
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 0.000156,
+      "loss": 2.6171,
+      "step": 52
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 0.000159,
+      "loss": 2.5955,
+      "step": 53
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 0.000162,
+      "loss": 2.6157,
+      "step": 54
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 0.000165,
+      "loss": 2.5878,
+      "step": 55
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 0.000168,
+      "loss": 3.2237,
+      "step": 56
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 0.000168,
+      "loss": 2.5616,
+      "step": 57
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 0.00017099999999999998,
+      "loss": 2.5695,
+      "step": 58
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 2.5583,
+      "step": 59
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 0.00017699999999999997,
+      "loss": 2.5993,
+      "step": 60
+    },
+    {
+      "epoch": 11.95,
+      "eval_accuracy": 0.0673758865248227,
+      "eval_f1": 0.01669758812615955,
+      "eval_loss": 2.630133867263794,
+      "eval_runtime": 16.2814,
+      "eval_samples_per_second": 17.32,
+      "eval_steps_per_second": 1.106,
+      "step": 60
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 0.00017699999999999997,
+      "loss": 3.2149,
+      "step": 61
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 2.5651,
+      "step": 62
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 0.00018299999999999998,
+      "loss": 2.5673,
+      "step": 63
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 0.000186,
+      "loss": 2.5464,
+      "step": 64
+    },
+    {
+      "epoch": 12.95,
+      "learning_rate": 0.000186,
+      "loss": 2.5236,
+      "step": 65
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 3.1678,
+      "step": 66
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 2.54,
+      "step": 67
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 2.5246,
+      "step": 68
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 0.000195,
+      "loss": 2.5038,
+      "step": 69
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 0.000198,
+      "loss": 2.559,
+      "step": 70
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 0.000201,
+      "loss": 3.0771,
+      "step": 71
+    },
+    {
+      "epoch": 14.38,
+      "learning_rate": 0.000204,
+      "loss": 2.5082,
+      "step": 72
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 0.00020699999999999996,
+      "loss": 2.5046,
+      "step": 73
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 2.4716,
+      "step": 74
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 2.498,
+      "step": 75
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 3.0287,
+      "step": 76
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 2.4355,
+      "step": 77
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 2.3877,
+      "step": 78
+    },
+    {
+      "epoch": 15.76,
+      "learning_rate": 0.000225,
+      "loss": 2.394,
+      "step": 79
+    },
+    {
+      "epoch": 15.95,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 2.4023,
+      "step": 80
+    },
+    {
+      "epoch": 15.95,
+      "eval_accuracy": 0.1453900709219858,
+      "eval_f1": 0.11052692674072857,
+      "eval_loss": 2.5513856410980225,
+      "eval_runtime": 16.363,
+      "eval_samples_per_second": 17.234,
+      "eval_steps_per_second": 1.1,
+      "step": 80
+    },
+    {
+      "epoch": 16.19,
+      "learning_rate": 0.00023099999999999998,
+      "loss": 2.9883,
+      "step": 81
+    },
+    {
+      "epoch": 16.38,
+      "learning_rate": 0.000234,
+      "loss": 2.312,
+      "step": 82
+    },
+    {
+      "epoch": 16.57,
+      "learning_rate": 0.000237,
+      "loss": 2.3485,
+      "step": 83
+    },
+    {
+      "epoch": 16.76,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 2.2478,
+      "step": 84
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 0.000243,
+      "loss": 2.3112,
+      "step": 85
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 2.7548,
+      "step": 86
+    },
+    {
+      "epoch": 17.38,
+      "learning_rate": 0.000249,
+      "loss": 2.0562,
+      "step": 87
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 2.0455,
+      "step": 88
+    },
+    {
+      "epoch": 17.76,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 2.1711,
+      "step": 89
+    },
+    {
+      "epoch": 17.95,
+      "learning_rate": 0.000258,
+      "loss": 2.0953,
+      "step": 90
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 0.000261,
+      "loss": 2.4917,
+      "step": 91
+    },
+    {
+      "epoch": 18.38,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 1.949,
+      "step": 92
+    },
+    {
+      "epoch": 18.57,
+      "learning_rate": 0.000267,
+      "loss": 1.7933,
+      "step": 93
+    },
+    {
+      "epoch": 18.76,
+      "learning_rate": 0.00027,
+      "loss": 1.7928,
+      "step": 94
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 0.00027299999999999997,
+      "loss": 1.8219,
+      "step": 95
+    },
+    {
+      "epoch": 19.19,
+      "learning_rate": 0.000276,
+      "loss": 2.0037,
+      "step": 96
+    },
+    {
+      "epoch": 19.38,
+      "learning_rate": 0.000279,
+      "loss": 1.7528,
+      "step": 97
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 1.4667,
+      "step": 98
+    },
+    {
+      "epoch": 19.76,
+      "learning_rate": 0.000285,
+      "loss": 1.3495,
+      "step": 99
+    },
+    {
+      "epoch": 19.95,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 1.4015,
+      "step": 100
+    },
+    {
+      "epoch": 19.95,
+      "eval_accuracy": 0.5851063829787234,
+      "eval_f1": 0.559944797763728,
+      "eval_loss": 1.6842906475067139,
+      "eval_runtime": 16.2668,
+      "eval_samples_per_second": 17.336,
+      "eval_steps_per_second": 1.107,
+      "step": 100
+    },
+    {
+      "epoch": 20.19,
+      "learning_rate": 0.00029099999999999997,
+      "loss": 1.7232,
+      "step": 101
+    },
+    {
+      "epoch": 20.38,
+      "learning_rate": 0.000294,
+      "loss": 1.225,
+      "step": 102
+    },
+    {
+      "epoch": 20.57,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 1.0986,
+      "step": 103
+    },
+    {
+      "epoch": 20.76,
+      "learning_rate": 0.0003,
+      "loss": 1.0969,
+      "step": 104
+    },
+    {
+      "epoch": 20.95,
+      "learning_rate": 0.00029953846153846155,
+      "loss": 1.0952,
+      "step": 105
+    },
+    {
+      "epoch": 21.19,
+      "learning_rate": 0.00029907692307692307,
+      "loss": 1.2957,
+      "step": 106
+    },
+    {
+      "epoch": 21.38,
+      "learning_rate": 0.0002986153846153846,
+      "loss": 0.9155,
+      "step": 107
+    },
+    {
+      "epoch": 21.57,
+      "learning_rate": 0.0002981538461538461,
+      "loss": 0.7972,
+      "step": 108
+    },
+    {
+      "epoch": 21.76,
+      "learning_rate": 0.0002976923076923077,
+      "loss": 0.8987,
+      "step": 109
+    },
+    {
+      "epoch": 21.95,
+      "learning_rate": 0.0002972307692307692,
+      "loss": 0.7826,
+      "step": 110
+    },
+    {
+      "epoch": 22.19,
+      "learning_rate": 0.00029676923076923077,
+      "loss": 0.8078,
+      "step": 111
+    },
+    {
+      "epoch": 22.38,
+      "learning_rate": 0.0002963076923076923,
+      "loss": 0.6328,
+      "step": 112
+    },
+    {
+      "epoch": 22.57,
+      "learning_rate": 0.0002958461538461538,
+      "loss": 0.6511,
+      "step": 113
+    },
+    {
+      "epoch": 22.76,
+      "learning_rate": 0.0002953846153846154,
+      "loss": 0.5964,
+      "step": 114
+    },
+    {
+      "epoch": 22.95,
+      "learning_rate": 0.0002949230769230769,
+      "loss": 0.5221,
+      "step": 115
+    },
+    {
+      "epoch": 23.19,
+      "learning_rate": 0.00029446153846153847,
+      "loss": 0.4883,
+      "step": 116
+    },
+    {
+      "epoch": 23.38,
+      "learning_rate": 0.000294,
+      "loss": 0.5555,
+      "step": 117
+    },
+    {
+      "epoch": 23.57,
+      "learning_rate": 0.0002935384615384615,
+      "loss": 0.5518,
+      "step": 118
+    },
+    {
+      "epoch": 23.76,
+      "learning_rate": 0.00029307692307692303,
+      "loss": 0.3964,
+      "step": 119
+    },
+    {
+      "epoch": 23.95,
+      "learning_rate": 0.0002926153846153846,
+      "loss": 0.4379,
+      "step": 120
+    },
+    {
+      "epoch": 23.95,
+      "eval_accuracy": 0.7907801418439716,
+      "eval_f1": 0.7921343313619132,
+      "eval_loss": 0.8125914335250854,
+      "eval_runtime": 16.222,
+      "eval_samples_per_second": 17.384,
+      "eval_steps_per_second": 1.11,
+      "step": 120
+    },
+    {
+      "epoch": 24.19,
+      "learning_rate": 0.0002921538461538461,
+      "loss": 0.4538,
+      "step": 121
+    },
+    {
+      "epoch": 24.38,
+      "learning_rate": 0.0002916923076923077,
+      "loss": 0.276,
+      "step": 122
+    },
+    {
+      "epoch": 24.57,
+      "learning_rate": 0.0002912307692307692,
+      "loss": 0.2824,
+      "step": 123
+    },
+    {
+      "epoch": 24.76,
+      "learning_rate": 0.00029076923076923073,
+      "loss": 0.2573,
+      "step": 124
+    },
+    {
+      "epoch": 24.95,
+      "learning_rate": 0.0002903076923076923,
+      "loss": 0.2001,
+      "step": 125
+    },
+    {
+      "epoch": 25.19,
+      "learning_rate": 0.0002898461538461538,
+      "loss": 0.2129,
+      "step": 126
+    },
+    {
+      "epoch": 25.38,
+      "learning_rate": 0.0002893846153846154,
+      "loss": 0.2281,
+      "step": 127
+    },
+    {
+      "epoch": 25.57,
+      "learning_rate": 0.0002889230769230769,
+      "loss": 0.1805,
+      "step": 128
+    },
+    {
+      "epoch": 25.76,
+      "learning_rate": 0.00028846153846153843,
+      "loss": 0.1337,
+      "step": 129
+    },
+    {
+      "epoch": 25.95,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 0.1433,
+      "step": 130
+    },
+    {
+      "epoch": 26.19,
+      "learning_rate": 0.0002875384615384615,
+      "loss": 0.2734,
+      "step": 131
+    },
+    {
+      "epoch": 26.38,
+      "learning_rate": 0.00028707692307692305,
+      "loss": 0.1092,
+      "step": 132
+    },
+    {
+      "epoch": 26.57,
+      "learning_rate": 0.0002866153846153846,
+      "loss": 0.1105,
+      "step": 133
+    },
+    {
+      "epoch": 26.76,
+      "learning_rate": 0.00028615384615384614,
+      "loss": 0.108,
+      "step": 134
+    },
+    {
+      "epoch": 26.95,
+      "learning_rate": 0.00028569230769230766,
+      "loss": 0.1012,
+      "step": 135
+    },
+    {
+      "epoch": 27.19,
+      "learning_rate": 0.00028523076923076923,
+      "loss": 0.0906,
+      "step": 136
+    },
+    {
+      "epoch": 27.38,
+      "learning_rate": 0.00028476923076923075,
+      "loss": 0.1653,
+      "step": 137
+    },
+    {
+      "epoch": 27.57,
+      "learning_rate": 0.00028430769230769227,
+      "loss": 0.0989,
+      "step": 138
+    },
+    {
+      "epoch": 27.76,
+      "learning_rate": 0.0002838461538461538,
+      "loss": 0.1263,
+      "step": 139
+    },
+    {
+      "epoch": 27.95,
+      "learning_rate": 0.00028338461538461536,
+      "loss": 0.0642,
+      "step": 140
+    },
+    {
+      "epoch": 27.95,
+      "eval_accuracy": 0.8156028368794326,
+      "eval_f1": 0.8158391180586454,
+      "eval_loss": 0.7177959084510803,
+      "eval_runtime": 16.0095,
+      "eval_samples_per_second": 17.615,
+      "eval_steps_per_second": 1.124,
+      "step": 140
+    },
+    {
+      "epoch": 28.19,
+      "learning_rate": 0.0002829230769230769,
+      "loss": 0.1286,
+      "step": 141
+    },
+    {
+      "epoch": 28.38,
+      "learning_rate": 0.00028246153846153845,
+      "loss": 0.1433,
+      "step": 142
+    },
+    {
+      "epoch": 28.57,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 0.05,
+      "step": 143
+    },
+    {
+      "epoch": 28.76,
+      "learning_rate": 0.0002815384615384615,
+      "loss": 0.2157,
+      "step": 144
+    },
+    {
+      "epoch": 28.95,
+      "learning_rate": 0.00028107692307692306,
+      "loss": 0.04,
+      "step": 145
+    },
+    {
+      "epoch": 29.19,
+      "learning_rate": 0.0002806153846153846,
+      "loss": 0.0531,
+      "step": 146
+    },
+    {
+      "epoch": 29.38,
+      "learning_rate": 0.00028015384615384615,
+      "loss": 0.1695,
+      "step": 147
+    },
+    {
+      "epoch": 29.57,
+      "learning_rate": 0.00027969230769230767,
+      "loss": 0.0523,
+      "step": 148
+    },
+    {
+      "epoch": 29.76,
+      "learning_rate": 0.0002792307692307692,
+      "loss": 0.0822,
+      "step": 149
+    },
+    {
+      "epoch": 29.95,
+      "learning_rate": 0.0002787692307692307,
+      "loss": 0.053,
+      "step": 150
+    },
+    {
+      "epoch": 30.19,
+      "learning_rate": 0.0002783076923076923,
+      "loss": 0.0445,
+      "step": 151
+    },
+    {
+      "epoch": 30.38,
+      "learning_rate": 0.0002778461538461538,
+      "loss": 0.035,
+      "step": 152
+    },
+    {
+      "epoch": 30.57,
+      "learning_rate": 0.0002773846153846154,
+      "loss": 0.0433,
+      "step": 153
+    },
+    {
+      "epoch": 30.76,
+      "learning_rate": 0.0002769230769230769,
+      "loss": 0.0409,
+      "step": 154
+    },
+    {
+      "epoch": 30.95,
+      "learning_rate": 0.0002764615384615384,
+      "loss": 0.0334,
+      "step": 155
+    },
+    {
+      "epoch": 31.19,
+      "learning_rate": 0.000276,
+      "loss": 0.0377,
+      "step": 156
+    },
+    {
+      "epoch": 31.38,
+      "learning_rate": 0.0002755384615384615,
+      "loss": 0.0301,
+      "step": 157
+    },
+    {
+      "epoch": 31.57,
+      "learning_rate": 0.0002750769230769231,
+      "loss": 0.0337,
+      "step": 158
+    },
+    {
+      "epoch": 31.76,
+      "learning_rate": 0.0002746153846153846,
+      "loss": 0.2154,
+      "step": 159
+    },
+    {
+      "epoch": 31.95,
+      "learning_rate": 0.0002741538461538461,
+      "loss": 0.0376,
+      "step": 160
+    },
+    {
+      "epoch": 31.95,
+      "eval_accuracy": 0.8475177304964538,
+      "eval_f1": 0.847321743921314,
+      "eval_loss": 0.7286385893821716,
+      "eval_runtime": 15.9823,
+      "eval_samples_per_second": 17.645,
+      "eval_steps_per_second": 1.126,
+      "step": 160
+    },
+    {
+      "epoch": 32.19,
+      "learning_rate": 0.00027369230769230764,
+      "loss": 0.0559,
+      "step": 161
+    },
+    {
+      "epoch": 32.38,
+      "learning_rate": 0.0002732307692307692,
+      "loss": 0.025,
+      "step": 162
+    },
+    {
+      "epoch": 32.57,
+      "learning_rate": 0.00027276923076923073,
+      "loss": 0.0322,
+      "step": 163
+    },
+    {
+      "epoch": 32.76,
+      "learning_rate": 0.0002723076923076923,
+      "loss": 0.1028,
+      "step": 164
+    },
+    {
+      "epoch": 32.95,
+      "learning_rate": 0.0002718461538461538,
+      "loss": 0.025,
+      "step": 165
+    },
+    {
+      "epoch": 33.19,
+      "learning_rate": 0.00027138461538461534,
+      "loss": 0.0365,
+      "step": 166
+    },
+    {
+      "epoch": 33.38,
+      "learning_rate": 0.0002709230769230769,
+      "loss": 0.0717,
+      "step": 167
+    },
+    {
+      "epoch": 33.57,
+      "learning_rate": 0.00027046153846153843,
+      "loss": 0.2531,
+      "step": 168
+    },
+    {
+      "epoch": 33.76,
+      "learning_rate": 0.00027,
+      "loss": 0.0172,
+      "step": 169
+    },
+    {
+      "epoch": 33.95,
+      "learning_rate": 0.0002695384615384615,
+      "loss": 0.0463,
+      "step": 170
+    },
+    {
+      "epoch": 34.19,
+      "learning_rate": 0.00026907692307692304,
+      "loss": 0.0289,
+      "step": 171
+    },
+    {
+      "epoch": 34.38,
+      "learning_rate": 0.00026861538461538456,
+      "loss": 0.2272,
+      "step": 172
+    },
+    {
+      "epoch": 34.57,
+      "learning_rate": 0.00026815384615384613,
+      "loss": 0.0299,
+      "step": 173
+    },
+    {
+      "epoch": 34.76,
+      "learning_rate": 0.0002676923076923077,
+      "loss": 0.0567,
+      "step": 174
+    },
+    {
+      "epoch": 34.95,
+      "learning_rate": 0.0002672307692307692,
+      "loss": 0.0332,
+      "step": 175
+    },
+    {
+      "epoch": 35.19,
+      "learning_rate": 0.00026676923076923074,
+      "loss": 0.0433,
+      "step": 176
+    },
+    {
+      "epoch": 35.38,
+      "learning_rate": 0.00026630769230769226,
+      "loss": 0.0162,
+      "step": 177
+    },
+    {
+      "epoch": 35.57,
+      "learning_rate": 0.00026584615384615384,
+      "loss": 0.064,
+      "step": 178
+    },
+    {
+      "epoch": 35.76,
+      "learning_rate": 0.00026538461538461536,
+      "loss": 0.0583,
+      "step": 179
+    },
+    {
+      "epoch": 35.95,
+      "learning_rate": 0.00026492307692307693,
+      "loss": 0.0185,
+      "step": 180
+    },
+    {
+      "epoch": 35.95,
+      "eval_accuracy": 0.8723404255319149,
+      "eval_f1": 0.8719028565875256,
+      "eval_loss": 0.6779401898384094,
+      "eval_runtime": 15.9944,
+      "eval_samples_per_second": 17.631,
+      "eval_steps_per_second": 1.125,
+      "step": 180
+    },
+    {
+      "epoch": 36.19,
+      "learning_rate": 0.00026446153846153845,
+      "loss": 0.0242,
+      "step": 181
+    },
+    {
+      "epoch": 36.38,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 0.0203,
+      "step": 182
+    },
+    {
+      "epoch": 36.57,
+      "learning_rate": 0.00026353846153846154,
+      "loss": 0.0205,
+      "step": 183
+    },
+    {
+      "epoch": 36.76,
+      "learning_rate": 0.00026307692307692306,
+      "loss": 0.0306,
+      "step": 184
+    },
+    {
+      "epoch": 36.95,
+      "learning_rate": 0.00026261538461538463,
+      "loss": 0.0172,
+      "step": 185
+    },
+    {
+      "epoch": 37.19,
+      "learning_rate": 0.00026215384615384615,
+      "loss": 0.2075,
+      "step": 186
+    },
+    {
+      "epoch": 37.38,
+      "learning_rate": 0.00026169230769230767,
+      "loss": 0.0773,
+      "step": 187
+    },
+    {
+      "epoch": 37.57,
+      "learning_rate": 0.0002612307692307692,
+      "loss": 0.1106,
+      "step": 188
+    },
+    {
+      "epoch": 37.76,
+      "learning_rate": 0.00026076923076923076,
+      "loss": 0.0126,
+      "step": 189
+    },
+    {
+      "epoch": 37.95,
+      "learning_rate": 0.0002603076923076923,
+      "loss": 0.0392,
+      "step": 190
+    },
+    {
+      "epoch": 38.19,
+      "learning_rate": 0.00025984615384615385,
+      "loss": 0.0863,
+      "step": 191
+    },
+    {
+      "epoch": 38.38,
+      "learning_rate": 0.00025938461538461537,
+      "loss": 0.2567,
+      "step": 192
+    },
+    {
+      "epoch": 38.57,
+      "learning_rate": 0.0002589230769230769,
+      "loss": 0.1527,
+      "step": 193
+    },
+    {
+      "epoch": 38.76,
+      "learning_rate": 0.00025846153846153846,
+      "loss": 0.0199,
+      "step": 194
+    },
+    {
+      "epoch": 38.95,
+      "learning_rate": 0.000258,
+      "loss": 0.0144,
+      "step": 195
+    },
+    {
+      "epoch": 39.19,
+      "learning_rate": 0.0002575384615384615,
+      "loss": 0.0232,
+      "step": 196
+    },
+    {
+      "epoch": 39.38,
+      "learning_rate": 0.000257076923076923,
+      "loss": 0.0126,
+      "step": 197
+    },
+    {
+      "epoch": 39.57,
+      "learning_rate": 0.0002566153846153846,
+      "loss": 0.0161,
+      "step": 198
+    },
+    {
+      "epoch": 39.76,
+      "learning_rate": 0.0002561538461538461,
+      "loss": 0.2685,
+      "step": 199
+    },
+    {
+      "epoch": 39.95,
+      "learning_rate": 0.0002556923076923077,
+      "loss": 0.0752,
+      "step": 200
+    },
+    {
+      "epoch": 39.95,
+      "eval_accuracy": 0.851063829787234,
+      "eval_f1": 0.8577814171457172,
+      "eval_loss": 0.7096136212348938,
+      "eval_runtime": 16.1916,
+      "eval_samples_per_second": 17.416,
+      "eval_steps_per_second": 1.112,
+      "step": 200
+    },
+    {
+      "epoch": 40.19,
+      "learning_rate": 0.0002552307692307692,
+      "loss": 0.2436,
+      "step": 201
+    },
+    {
+      "epoch": 40.38,
+      "learning_rate": 0.0002547692307692307,
+      "loss": 0.0121,
+      "step": 202
+    },
+    {
+      "epoch": 40.57,
+      "learning_rate": 0.0002543076923076923,
+      "loss": 0.1298,
+      "step": 203
+    },
+    {
+      "epoch": 40.76,
+      "learning_rate": 0.0002538461538461538,
+      "loss": 0.0246,
+      "step": 204
+    },
+    {
+      "epoch": 40.95,
+      "learning_rate": 0.0002533846153846154,
+      "loss": 0.0171,
+      "step": 205
+    },
+    {
+      "epoch": 41.19,
+      "learning_rate": 0.0002529230769230769,
+      "loss": 0.065,
+      "step": 206
+    },
+    {
+      "epoch": 41.38,
+      "learning_rate": 0.0002524615384615384,
+      "loss": 0.0257,
+      "step": 207
+    },
+    {
+      "epoch": 41.57,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 0.0189,
+      "step": 208
+    },
+    {
+      "epoch": 41.76,
+      "learning_rate": 0.0002515384615384615,
+      "loss": 0.0238,
+      "step": 209
+    },
+    {
+      "epoch": 41.95,
+      "learning_rate": 0.00025107692307692304,
+      "loss": 0.016,
+      "step": 210
+    },
+    {
+      "epoch": 42.19,
+      "learning_rate": 0.0002506153846153846,
+      "loss": 0.0154,
+      "step": 211
+    },
+    {
+      "epoch": 42.38,
+      "learning_rate": 0.00025015384615384613,
+      "loss": 0.0114,
+      "step": 212
+    },
+    {
+      "epoch": 42.57,
+      "learning_rate": 0.00024969230769230765,
+      "loss": 0.0097,
+      "step": 213
+    },
+    {
+      "epoch": 42.76,
+      "learning_rate": 0.0002492307692307692,
+      "loss": 0.0135,
+      "step": 214
+    },
+    {
+      "epoch": 42.95,
+      "learning_rate": 0.00024876923076923074,
+      "loss": 0.2339,
+      "step": 215
+    },
+    {
+      "epoch": 43.19,
+      "learning_rate": 0.0002483076923076923,
+      "loss": 0.0125,
+      "step": 216
+    },
+    {
+      "epoch": 43.38,
+      "learning_rate": 0.00024784615384615383,
+      "loss": 0.0154,
+      "step": 217
+    },
+    {
+      "epoch": 43.57,
+      "learning_rate": 0.00024738461538461535,
+      "loss": 0.0118,
+      "step": 218
+    },
+    {
+      "epoch": 43.76,
+      "learning_rate": 0.00024692307692307687,
+      "loss": 0.0101,
+      "step": 219
+    },
+    {
+      "epoch": 43.95,
+      "learning_rate": 0.00024646153846153844,
+      "loss": 0.0266,
+      "step": 220
+    },
+    {
+      "epoch": 43.95,
+      "eval_accuracy": 0.8546099290780141,
+      "eval_f1": 0.8596362724182274,
+      "eval_loss": 0.7654581069946289,
+      "eval_runtime": 16.2536,
+      "eval_samples_per_second": 17.35,
+      "eval_steps_per_second": 1.107,
+      "step": 220
+    },
+    {
+      "epoch": 44.19,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 0.0122,
+      "step": 221
+    },
+    {
+      "epoch": 44.38,
+      "learning_rate": 0.00024553846153846154,
+      "loss": 0.3488,
+      "step": 222
+    },
+    {
+      "epoch": 44.57,
+      "learning_rate": 0.00024507692307692305,
+      "loss": 0.0128,
+      "step": 223
+    },
+    {
+      "epoch": 44.76,
+      "learning_rate": 0.0002446153846153846,
+      "loss": 0.011,
+      "step": 224
+    },
+    {
+      "epoch": 44.95,
+      "learning_rate": 0.00024415384615384615,
+      "loss": 0.0116,
+      "step": 225
+    },
+    {
+      "epoch": 45.19,
+      "learning_rate": 0.00024369230769230767,
+      "loss": 0.2668,
+      "step": 226
+    },
+    {
+      "epoch": 45.38,
+      "learning_rate": 0.0002432307692307692,
+      "loss": 0.1324,
+      "step": 227
+    },
+    {
+      "epoch": 45.57,
+      "learning_rate": 0.00024276923076923073,
+      "loss": 0.0081,
+      "step": 228
+    },
+    {
+      "epoch": 45.76,
+      "learning_rate": 0.0002423076923076923,
+      "loss": 0.0084,
+      "step": 229
+    },
+    {
+      "epoch": 45.95,
+      "learning_rate": 0.00024184615384615382,
+      "loss": 0.174,
+      "step": 230
+    },
+    {
+      "epoch": 46.19,
+      "learning_rate": 0.00024138461538461537,
+      "loss": 0.0216,
+      "step": 231
+    },
+    {
+      "epoch": 46.38,
+      "learning_rate": 0.0002409230769230769,
+      "loss": 0.0159,
+      "step": 232
+    },
+    {
+      "epoch": 46.57,
+      "learning_rate": 0.00024046153846153843,
+      "loss": 0.0097,
+      "step": 233
+    },
+    {
+      "epoch": 46.76,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.0094,
+      "step": 234
+    },
+    {
+      "epoch": 46.95,
+      "learning_rate": 0.00023953846153846152,
+      "loss": 0.0319,
+      "step": 235
+    },
+    {
+      "epoch": 47.19,
+      "learning_rate": 0.00023907692307692307,
+      "loss": 0.2269,
+      "step": 236
+    },
+    {
+      "epoch": 47.38,
+      "learning_rate": 0.0002386153846153846,
+      "loss": 0.0179,
+      "step": 237
+    },
+    {
+      "epoch": 47.57,
+      "learning_rate": 0.00023815384615384614,
+      "loss": 0.0366,
+      "step": 238
+    },
+    {
+      "epoch": 47.76,
+      "learning_rate": 0.00023769230769230765,
+      "loss": 0.0112,
+      "step": 239
+    },
+    {
+      "epoch": 47.95,
+      "learning_rate": 0.0002372307692307692,
+      "loss": 0.0078,
+      "step": 240
+    },
+    {
+      "epoch": 47.95,
+      "eval_accuracy": 0.851063829787234,
+      "eval_f1": 0.8562613066492183,
+      "eval_loss": 0.7622692584991455,
+      "eval_runtime": 16.272,
+      "eval_samples_per_second": 17.33,
+      "eval_steps_per_second": 1.106,
+      "step": 240
+    },
+    {
+      "epoch": 48.19,
+      "learning_rate": 0.00023676923076923075,
+      "loss": 0.0107,
+      "step": 241
+    },
+    {
+      "epoch": 48.38,
+      "learning_rate": 0.0002363076923076923,
+      "loss": 0.0167,
+      "step": 242
+    },
+    {
+      "epoch": 48.57,
+      "learning_rate": 0.00023584615384615384,
+      "loss": 0.0088,
+      "step": 243
+    },
+    {
+      "epoch": 48.76,
+      "learning_rate": 0.00023538461538461536,
+      "loss": 0.0078,
+      "step": 244
+    },
+    {
+      "epoch": 48.95,
+      "learning_rate": 0.0002349230769230769,
+      "loss": 0.0085,
+      "step": 245
+    },
+    {
+      "epoch": 49.19,
+      "learning_rate": 0.00023446153846153842,
+      "loss": 0.0122,
+      "step": 246
+    },
+    {
+      "epoch": 49.38,
+      "learning_rate": 0.000234,
+      "loss": 0.0377,
+      "step": 247
+    },
+    {
+      "epoch": 49.57,
+      "learning_rate": 0.00023353846153846151,
+      "loss": 0.0083,
+      "step": 248
+    },
+    {
+      "epoch": 49.76,
+      "learning_rate": 0.00023307692307692306,
+      "loss": 0.0083,
+      "step": 249
+    },
+    {
+      "epoch": 49.95,
+      "learning_rate": 0.00023261538461538458,
+      "loss": 0.0073,
+      "step": 250
+    },
+    {
+      "epoch": 50.19,
+      "learning_rate": 0.00023215384615384613,
+      "loss": 0.0099,
+      "step": 251
+    },
+    {
+      "epoch": 50.38,
+      "learning_rate": 0.0002316923076923077,
+      "loss": 0.0088,
+      "step": 252
+    },
+    {
+      "epoch": 50.57,
+      "learning_rate": 0.00023123076923076922,
+      "loss": 0.0086,
+      "step": 253
+    },
+    {
+      "epoch": 50.76,
+      "learning_rate": 0.00023076923076923076,
+      "loss": 0.0625,
+      "step": 254
+    },
+    {
+      "epoch": 50.95,
+      "learning_rate": 0.00023030769230769228,
+      "loss": 0.0101,
+      "step": 255
+    },
+    {
+      "epoch": 51.19,
+      "learning_rate": 0.00022984615384615383,
+      "loss": 0.012,
+      "step": 256
+    },
+    {
+      "epoch": 51.38,
+      "learning_rate": 0.00022938461538461535,
+      "loss": 0.0687,
+      "step": 257
+    },
+    {
+      "epoch": 51.57,
+      "learning_rate": 0.00022892307692307692,
+      "loss": 0.0061,
+      "step": 258
+    },
+    {
+      "epoch": 51.76,
+      "learning_rate": 0.00022846153846153844,
+      "loss": 0.0063,
+      "step": 259
+    },
+    {
+      "epoch": 51.95,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 0.007,
+      "step": 260
+    },
+    {
+      "epoch": 51.95,
+      "eval_accuracy": 0.875886524822695,
+      "eval_f1": 0.8794189931188913,
+      "eval_loss": 0.6619934439659119,
+      "eval_runtime": 16.1754,
+      "eval_samples_per_second": 17.434,
+      "eval_steps_per_second": 1.113,
+      "step": 260
+    },
+    {
+      "epoch": 52.19,
+      "learning_rate": 0.0002275384615384615,
+      "loss": 0.04,
+      "step": 261
+    },
+    {
+      "epoch": 52.38,
+      "learning_rate": 0.00022707692307692305,
+      "loss": 0.0087,
+      "step": 262
+    },
+    {
+      "epoch": 52.57,
+      "learning_rate": 0.0002266153846153846,
+      "loss": 0.0115,
+      "step": 263
+    },
+    {
+      "epoch": 52.76,
+      "learning_rate": 0.00022615384615384614,
+      "loss": 0.0136,
+      "step": 264
+    },
+    {
+      "epoch": 52.95,
+      "learning_rate": 0.0002256923076923077,
+      "loss": 0.0334,
+      "step": 265
+    },
+    {
+      "epoch": 53.19,
+      "learning_rate": 0.0002252307692307692,
+      "loss": 0.0175,
+      "step": 266
+    },
+    {
+      "epoch": 53.38,
+      "learning_rate": 0.00022476923076923075,
+      "loss": 0.0061,
+      "step": 267
+    },
+    {
+      "epoch": 53.57,
+      "learning_rate": 0.00022430769230769227,
+      "loss": 0.0067,
+      "step": 268
+    },
+    {
+      "epoch": 53.76,
+      "learning_rate": 0.00022384615384615382,
+      "loss": 0.0067,
+      "step": 269
+    },
+    {
+      "epoch": 53.95,
+      "learning_rate": 0.00022338461538461536,
+      "loss": 0.0061,
+      "step": 270
+    },
+    {
+      "epoch": 54.19,
+      "learning_rate": 0.0002229230769230769,
+      "loss": 0.0153,
+      "step": 271
+    },
+    {
+      "epoch": 54.38,
+      "learning_rate": 0.00022246153846153846,
+      "loss": 0.0271,
+      "step": 272
+    },
+    {
+      "epoch": 54.57,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 0.0056,
+      "step": 273
+    },
+    {
+      "epoch": 54.76,
+      "learning_rate": 0.00022153846153846152,
+      "loss": 0.0067,
+      "step": 274
+    },
+    {
+      "epoch": 54.95,
+      "learning_rate": 0.00022107692307692304,
+      "loss": 0.0049,
+      "step": 275
+    },
+    {
+      "epoch": 55.19,
+      "learning_rate": 0.0002206153846153846,
+      "loss": 0.0067,
+      "step": 276
+    },
+    {
+      "epoch": 55.38,
+      "learning_rate": 0.00022015384615384613,
+      "loss": 0.0123,
+      "step": 277
+    },
+    {
+      "epoch": 55.57,
+      "learning_rate": 0.00021969230769230768,
+      "loss": 0.0051,
+      "step": 278
+    },
+    {
+      "epoch": 55.76,
+      "learning_rate": 0.0002192307692307692,
+      "loss": 0.0059,
+      "step": 279
+    },
+    {
+      "epoch": 55.95,
+      "learning_rate": 0.00021876923076923074,
+      "loss": 0.0047,
+      "step": 280
+    },
+    {
+      "epoch": 55.95,
+      "eval_accuracy": 0.900709219858156,
+      "eval_f1": 0.9044900822526744,
+      "eval_loss": 0.5935764312744141,
+      "eval_runtime": 16.1857,
+      "eval_samples_per_second": 17.423,
+      "eval_steps_per_second": 1.112,
+      "step": 280
+    },
+    {
+      "epoch": 56.19,
+      "learning_rate": 0.00021830769230769226,
+      "loss": 0.0062,
+      "step": 281
+    },
+    {
+      "epoch": 56.38,
+      "learning_rate": 0.00021784615384615383,
+      "loss": 0.0077,
+      "step": 282
+    },
+    {
+      "epoch": 56.57,
+      "learning_rate": 0.00021738461538461538,
+      "loss": 0.0056,
+      "step": 283
+    },
+    {
+      "epoch": 56.76,
+      "learning_rate": 0.0002169230769230769,
+      "loss": 0.0048,
+      "step": 284
+    },
+    {
+      "epoch": 56.95,
+      "learning_rate": 0.00021646153846153845,
+      "loss": 0.0166,
+      "step": 285
+    },
+    {
+      "epoch": 57.19,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 0.2203,
+      "step": 286
+    },
+    {
+      "epoch": 57.38,
+      "learning_rate": 0.00021553846153846154,
+      "loss": 0.0044,
+      "step": 287
+    },
+    {
+      "epoch": 57.57,
+      "learning_rate": 0.00021507692307692306,
+      "loss": 0.0057,
+      "step": 288
+    },
+    {
+      "epoch": 57.76,
+      "learning_rate": 0.0002146153846153846,
+      "loss": 0.0046,
+      "step": 289
+    },
+    {
+      "epoch": 57.95,
+      "learning_rate": 0.00021415384615384612,
+      "loss": 0.0082,
+      "step": 290
+    },
+    {
+      "epoch": 58.19,
+      "learning_rate": 0.00021369230769230767,
+      "loss": 0.007,
+      "step": 291
+    },
+    {
+      "epoch": 58.38,
+      "learning_rate": 0.00021323076923076921,
+      "loss": 0.0057,
+      "step": 292
+    },
+    {
+      "epoch": 58.57,
+      "learning_rate": 0.00021276923076923076,
+      "loss": 0.0066,
+      "step": 293
+    },
+    {
+      "epoch": 58.76,
+      "learning_rate": 0.0002123076923076923,
+      "loss": 0.108,
+      "step": 294
+    },
+    {
+      "epoch": 58.95,
+      "learning_rate": 0.00021184615384615382,
+      "loss": 0.0045,
+      "step": 295
+    },
+    {
+      "epoch": 59.19,
+      "learning_rate": 0.00021138461538461537,
+      "loss": 0.0058,
+      "step": 296
+    },
+    {
+      "epoch": 59.38,
+      "learning_rate": 0.0002109230769230769,
+      "loss": 0.0059,
+      "step": 297
+    },
+    {
+      "epoch": 59.57,
+      "learning_rate": 0.00021046153846153844,
+      "loss": 0.0736,
+      "step": 298
+    },
+    {
+      "epoch": 59.76,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.005,
+      "step": 299
+    },
+    {
+      "epoch": 59.95,
+      "learning_rate": 0.00020953846153846153,
+      "loss": 0.0067,
+      "step": 300
+    },
+    {
+      "epoch": 59.95,
+      "eval_accuracy": 0.8617021276595744,
+      "eval_f1": 0.8546408961595013,
+      "eval_loss": 0.8279188275337219,
+      "eval_runtime": 16.1625,
+      "eval_samples_per_second": 17.448,
+      "eval_steps_per_second": 1.114,
+      "step": 300
+    },
+    {
+      "epoch": 60.19,
+      "learning_rate": 0.00020907692307692307,
+      "loss": 0.0091,
+      "step": 301
+    },
+    {
+      "epoch": 60.38,
+      "learning_rate": 0.0002086153846153846,
+      "loss": 0.0073,
+      "step": 302
+    },
+    {
+      "epoch": 60.57,
+      "learning_rate": 0.00020815384615384614,
+      "loss": 0.0048,
+      "step": 303
+    },
+    {
+      "epoch": 60.76,
+      "learning_rate": 0.00020769230769230766,
+      "loss": 0.0053,
+      "step": 304
+    },
+    {
+      "epoch": 60.95,
+      "learning_rate": 0.00020723076923076923,
+      "loss": 0.0284,
+      "step": 305
+    },
+    {
+      "epoch": 61.19,
+      "learning_rate": 0.00020676923076923075,
+      "loss": 0.0114,
+      "step": 306
+    },
+    {
+      "epoch": 61.38,
+      "learning_rate": 0.0002063076923076923,
+      "loss": 0.0848,
+      "step": 307
+    },
+    {
+      "epoch": 61.57,
+      "learning_rate": 0.00020584615384615381,
+      "loss": 0.0043,
+      "step": 308
+    },
+    {
+      "epoch": 61.76,
+      "learning_rate": 0.00020538461538461536,
+      "loss": 0.0061,
+      "step": 309
+    },
+    {
+      "epoch": 61.95,
+      "learning_rate": 0.00020492307692307688,
+      "loss": 0.0058,
+      "step": 310
+    },
+    {
+      "epoch": 62.19,
+      "learning_rate": 0.00020446153846153845,
+      "loss": 0.0409,
+      "step": 311
+    },
+    {
+      "epoch": 62.38,
+      "learning_rate": 0.000204,
+      "loss": 0.0546,
+      "step": 312
+    },
+    {
+      "epoch": 62.57,
+      "learning_rate": 0.00020353846153846152,
+      "loss": 0.0089,
+      "step": 313
+    },
+    {
+      "epoch": 62.76,
+      "learning_rate": 0.00020307692307692306,
+      "loss": 0.006,
+      "step": 314
+    },
+    {
+      "epoch": 62.95,
+      "learning_rate": 0.00020261538461538458,
+      "loss": 0.0047,
+      "step": 315
+    },
+    {
+      "epoch": 63.19,
+      "learning_rate": 0.00020215384615384616,
+      "loss": 0.0831,
+      "step": 316
+    },
+    {
+      "epoch": 63.38,
+      "learning_rate": 0.00020169230769230767,
+      "loss": 0.0152,
+      "step": 317
+    },
+    {
+      "epoch": 63.57,
+      "learning_rate": 0.00020123076923076922,
+      "loss": 0.0714,
+      "step": 318
+    },
+    {
+      "epoch": 63.76,
+      "learning_rate": 0.00020076923076923074,
+      "loss": 0.0039,
+      "step": 319
+    },
+    {
+      "epoch": 63.95,
+      "learning_rate": 0.00020030769230769229,
+      "loss": 0.0394,
+      "step": 320
+    },
+    {
+      "epoch": 63.95,
+      "eval_accuracy": 0.8226950354609929,
+      "eval_f1": 0.835930513062929,
+      "eval_loss": 0.8765814900398254,
+      "eval_runtime": 16.2062,
+      "eval_samples_per_second": 17.401,
+      "eval_steps_per_second": 1.111,
+      "step": 320
+    },
+    {
+      "epoch": 64.19,
+      "learning_rate": 0.00019984615384615386,
+      "loss": 0.1083,
+      "step": 321
+    },
+    {
+      "epoch": 64.38,
+      "learning_rate": 0.00019938461538461538,
+      "loss": 0.0071,
+      "step": 322
+    },
+    {
+      "epoch": 64.57,
+      "learning_rate": 0.00019892307692307692,
+      "loss": 0.0179,
+      "step": 323
+    },
+    {
+      "epoch": 64.76,
+      "learning_rate": 0.00019846153846153844,
+      "loss": 0.0044,
+      "step": 324
+    },
+    {
+      "epoch": 64.95,
+      "learning_rate": 0.000198,
+      "loss": 0.2961,
+      "step": 325
+    },
+    {
+      "epoch": 65.19,
+      "learning_rate": 0.0001975384615384615,
+      "loss": 0.0057,
+      "step": 326
+    },
+    {
+      "epoch": 65.38,
+      "learning_rate": 0.00019707692307692305,
+      "loss": 0.0048,
+      "step": 327
+    },
+    {
+      "epoch": 65.57,
+      "learning_rate": 0.0001966153846153846,
+      "loss": 0.0037,
+      "step": 328
+    },
+    {
+      "epoch": 65.76,
+      "learning_rate": 0.00019615384615384615,
+      "loss": 0.0091,
+      "step": 329
+    },
+    {
+      "epoch": 65.95,
+      "learning_rate": 0.00019569230769230766,
+      "loss": 0.1506,
+      "step": 330
+    },
+    {
+      "epoch": 66.19,
+      "learning_rate": 0.0001952307692307692,
+      "loss": 0.0054,
+      "step": 331
+    },
+    {
+      "epoch": 66.38,
+      "learning_rate": 0.00019476923076923076,
+      "loss": 0.0072,
+      "step": 332
+    },
+    {
+      "epoch": 66.57,
+      "learning_rate": 0.00019430769230769227,
+      "loss": 0.0224,
+      "step": 333
+    },
+    {
+      "epoch": 66.76,
+      "learning_rate": 0.00019384615384615385,
+      "loss": 0.0078,
+      "step": 334
+    },
+    {
+      "epoch": 66.95,
+      "learning_rate": 0.00019338461538461537,
+      "loss": 0.0062,
+      "step": 335
+    },
+    {
+      "epoch": 67.19,
+      "learning_rate": 0.0001929230769230769,
+      "loss": 0.2077,
+      "step": 336
+    },
+    {
+      "epoch": 67.38,
+      "learning_rate": 0.00019246153846153843,
+      "loss": 0.0053,
+      "step": 337
+    },
+    {
+      "epoch": 67.57,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 0.0051,
+      "step": 338
+    },
+    {
+      "epoch": 67.76,
+      "learning_rate": 0.0001915384615384615,
+      "loss": 0.0193,
+      "step": 339
+    },
+    {
+      "epoch": 67.95,
+      "learning_rate": 0.00019107692307692307,
+      "loss": 0.0051,
+      "step": 340
+    },
+    {
+      "epoch": 67.95,
+      "eval_accuracy": 0.8475177304964538,
+      "eval_f1": 0.8482733642091616,
+      "eval_loss": 0.8097020983695984,
+      "eval_runtime": 16.2074,
+      "eval_samples_per_second": 17.399,
+      "eval_steps_per_second": 1.111,
+      "step": 340
+    },
+    {
+      "epoch": 68.19,
+      "learning_rate": 0.00019061538461538462,
+      "loss": 0.0177,
+      "step": 341
+    },
+    {
+      "epoch": 68.38,
+      "learning_rate": 0.00019015384615384613,
+      "loss": 0.0058,
+      "step": 342
+    },
+    {
+      "epoch": 68.57,
+      "learning_rate": 0.00018969230769230768,
+      "loss": 0.0078,
+      "step": 343
+    },
+    {
+      "epoch": 68.76,
+      "learning_rate": 0.0001892307692307692,
+      "loss": 0.0065,
+      "step": 344
+    },
+    {
+      "epoch": 68.95,
+      "learning_rate": 0.00018876923076923077,
+      "loss": 0.3679,
+      "step": 345
+    },
+    {
+      "epoch": 69.19,
+      "learning_rate": 0.0001883076923076923,
+      "loss": 0.0137,
+      "step": 346
+    },
+    {
+      "epoch": 69.38,
+      "learning_rate": 0.00018784615384615384,
+      "loss": 0.0045,
+      "step": 347
+    },
+    {
+      "epoch": 69.57,
+      "learning_rate": 0.00018738461538461536,
+      "loss": 0.0857,
+      "step": 348
+    },
+    {
+      "epoch": 69.76,
+      "learning_rate": 0.0001869230769230769,
+      "loss": 0.0044,
+      "step": 349
+    },
+    {
+      "epoch": 69.95,
+      "learning_rate": 0.00018646153846153842,
+      "loss": 0.0074,
+      "step": 350
+    },
+    {
+      "epoch": 70.19,
+      "learning_rate": 0.000186,
+      "loss": 0.0065,
+      "step": 351
+    },
+    {
+      "epoch": 70.38,
+      "learning_rate": 0.00018553846153846154,
+      "loss": 0.0156,
+      "step": 352
+    },
+    {
+      "epoch": 70.57,
+      "learning_rate": 0.00018507692307692306,
+      "loss": 0.0204,
+      "step": 353
+    },
+    {
+      "epoch": 70.76,
+      "learning_rate": 0.0001846153846153846,
+      "loss": 0.0063,
+      "step": 354
+    },
+    {
+      "epoch": 70.95,
+      "learning_rate": 0.00018415384615384612,
+      "loss": 0.0045,
+      "step": 355
+    },
+    {
+      "epoch": 71.19,
+      "learning_rate": 0.00018369230769230767,
+      "loss": 0.0083,
+      "step": 356
+    },
+    {
+      "epoch": 71.38,
+      "learning_rate": 0.00018323076923076922,
+      "loss": 0.0054,
+      "step": 357
+    },
+    {
+      "epoch": 71.57,
+      "learning_rate": 0.00018276923076923076,
+      "loss": 0.007,
+      "step": 358
+    },
+    {
+      "epoch": 71.76,
+      "learning_rate": 0.00018230769230769228,
+      "loss": 0.0113,
+      "step": 359
+    },
+    {
+      "epoch": 71.95,
+      "learning_rate": 0.00018184615384615383,
+      "loss": 0.0095,
+      "step": 360
+    },
+    {
+      "epoch": 71.95,
+      "eval_accuracy": 0.9078014184397163,
+      "eval_f1": 0.9082660804773691,
+      "eval_loss": 0.6095272898674011,
+      "eval_runtime": 16.2322,
+      "eval_samples_per_second": 17.373,
+      "eval_steps_per_second": 1.109,
+      "step": 360
+    },
+    {
+      "epoch": 72.19,
+      "learning_rate": 0.00018138461538461537,
+      "loss": 0.0069,
+      "step": 361
+    },
+    {
+      "epoch": 72.38,
+      "learning_rate": 0.0001809230769230769,
+      "loss": 0.0047,
+      "step": 362
+    },
+    {
+      "epoch": 72.57,
+      "learning_rate": 0.00018046153846153847,
+      "loss": 0.0037,
+      "step": 363
+    },
+    {
+      "epoch": 72.76,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.0041,
+      "step": 364
+    },
+    {
+      "epoch": 72.95,
+      "learning_rate": 0.00017953846153846153,
+      "loss": 0.0035,
+      "step": 365
+    },
+    {
+      "epoch": 73.19,
+      "learning_rate": 0.00017907692307692305,
+      "loss": 0.0043,
+      "step": 366
+    },
+    {
+      "epoch": 73.38,
+      "learning_rate": 0.0001786153846153846,
+      "loss": 0.0033,
+      "step": 367
+    },
+    {
+      "epoch": 73.57,
+      "learning_rate": 0.00017815384615384611,
+      "loss": 0.0037,
+      "step": 368
+    },
+    {
+      "epoch": 73.76,
+      "learning_rate": 0.0001776923076923077,
+      "loss": 0.0044,
+      "step": 369
+    },
+    {
+      "epoch": 73.95,
+      "learning_rate": 0.00017723076923076923,
+      "loss": 0.0035,
+      "step": 370
+    },
+    {
+      "epoch": 74.19,
+      "learning_rate": 0.00017676923076923075,
+      "loss": 0.0037,
+      "step": 371
+    },
+    {
+      "epoch": 74.38,
+      "learning_rate": 0.0001763076923076923,
+      "loss": 0.0079,
+      "step": 372
+    },
+    {
+      "epoch": 74.57,
+      "learning_rate": 0.00017584615384615382,
+      "loss": 0.0051,
+      "step": 373
+    },
+    {
+      "epoch": 74.76,
+      "learning_rate": 0.0001753846153846154,
+      "loss": 0.0033,
+      "step": 374
+    },
+    {
+      "epoch": 74.95,
+      "learning_rate": 0.0001749230769230769,
+      "loss": 0.0034,
+      "step": 375
+    },
+    {
+      "epoch": 75.19,
+      "learning_rate": 0.00017446153846153846,
+      "loss": 0.0038,
+      "step": 376
+    },
+    {
+      "epoch": 75.38,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 0.0029,
+      "step": 377
+    },
+    {
+      "epoch": 75.57,
+      "learning_rate": 0.00017353846153846152,
+      "loss": 0.0036,
+      "step": 378
+    },
+    {
+      "epoch": 75.76,
+      "learning_rate": 0.00017307692307692304,
+      "loss": 0.006,
+      "step": 379
+    },
+    {
+      "epoch": 75.95,
+      "learning_rate": 0.0001726153846153846,
+      "loss": 0.0026,
+      "step": 380
+    },
+    {
+      "epoch": 75.95,
+      "eval_accuracy": 0.8865248226950354,
+      "eval_f1": 0.88890426954986,
+      "eval_loss": 0.5286417007446289,
+      "eval_runtime": 16.2032,
+      "eval_samples_per_second": 17.404,
+      "eval_steps_per_second": 1.111,
+      "step": 380
+    },
+    {
+      "epoch": 76.19,
+      "learning_rate": 0.00017215384615384616,
+      "loss": 0.0036,
+      "step": 381
+    },
+    {
+      "epoch": 76.38,
+      "learning_rate": 0.00017169230769230768,
+      "loss": 0.0026,
+      "step": 382
+    },
+    {
+      "epoch": 76.57,
+      "learning_rate": 0.00017123076923076922,
+      "loss": 0.0026,
+      "step": 383
+    },
+    {
+      "epoch": 76.76,
+      "learning_rate": 0.00017076923076923074,
+      "loss": 0.021,
+      "step": 384
+    },
+    {
+      "epoch": 76.95,
+      "learning_rate": 0.0001703076923076923,
+      "loss": 0.0027,
+      "step": 385
+    },
+    {
+      "epoch": 77.19,
+      "learning_rate": 0.00016984615384615383,
+      "loss": 0.0032,
+      "step": 386
+    },
+    {
+      "epoch": 77.38,
+      "learning_rate": 0.00016938461538461538,
+      "loss": 0.0026,
+      "step": 387
+    },
+    {
+      "epoch": 77.57,
+      "learning_rate": 0.0001689230769230769,
+      "loss": 0.0025,
+      "step": 388
+    },
+    {
+      "epoch": 77.76,
+      "learning_rate": 0.00016846153846153844,
+      "loss": 0.0025,
+      "step": 389
+    },
+    {
+      "epoch": 77.95,
+      "learning_rate": 0.000168,
+      "loss": 0.0025,
+      "step": 390
+    },
+    {
+      "epoch": 78.19,
+      "learning_rate": 0.0001675384615384615,
+      "loss": 0.0033,
+      "step": 391
+    },
+    {
+      "epoch": 78.38,
+      "learning_rate": 0.00016707692307692308,
+      "loss": 0.0023,
+      "step": 392
+    },
+    {
+      "epoch": 78.57,
+      "learning_rate": 0.0001666153846153846,
+      "loss": 0.0028,
+      "step": 393
+    },
+    {
+      "epoch": 78.76,
+      "learning_rate": 0.00016615384615384615,
+      "loss": 0.0028,
+      "step": 394
+    },
+    {
+      "epoch": 78.95,
+      "learning_rate": 0.00016569230769230767,
+      "loss": 0.0023,
+      "step": 395
+    },
+    {
+      "epoch": 79.19,
+      "learning_rate": 0.0001652307692307692,
+      "loss": 0.0033,
+      "step": 396
+    },
+    {
+      "epoch": 79.38,
+      "learning_rate": 0.00016476923076923073,
+      "loss": 0.0024,
+      "step": 397
+    },
+    {
+      "epoch": 79.57,
+      "learning_rate": 0.0001643076923076923,
+      "loss": 0.0026,
+      "step": 398
+    },
+    {
+      "epoch": 79.76,
+      "learning_rate": 0.00016384615384615382,
+      "loss": 0.0022,
+      "step": 399
+    },
+    {
+      "epoch": 79.95,
+      "learning_rate": 0.00016338461538461537,
+      "loss": 0.0023,
+      "step": 400
+    },
+    {
+      "epoch": 79.95,
+      "eval_accuracy": 0.8936170212765957,
+      "eval_f1": 0.8926330265114482,
+      "eval_loss": 0.7218273878097534,
+      "eval_runtime": 16.1704,
+      "eval_samples_per_second": 17.439,
+      "eval_steps_per_second": 1.113,
+      "step": 400
+    },
+    {
+      "epoch": 80.19,
+      "learning_rate": 0.00016292307692307692,
+      "loss": 0.003,
+      "step": 401
+    },
+    {
+      "epoch": 80.38,
+      "learning_rate": 0.00016246153846153843,
+      "loss": 0.0027,
+      "step": 402
+    },
+    {
+      "epoch": 80.57,
+      "learning_rate": 0.000162,
+      "loss": 0.0039,
+      "step": 403
+    },
+    {
+      "epoch": 80.76,
+      "learning_rate": 0.00016153846153846153,
+      "loss": 0.0022,
+      "step": 404
+    },
+    {
+      "epoch": 80.95,
+      "learning_rate": 0.00016107692307692307,
+      "loss": 0.0029,
+      "step": 405
+    },
+    {
+      "epoch": 81.19,
+      "learning_rate": 0.0001606153846153846,
+      "loss": 0.0031,
+      "step": 406
+    },
+    {
+      "epoch": 81.38,
+      "learning_rate": 0.00016015384615384614,
+      "loss": 0.0024,
+      "step": 407
+    },
+    {
+      "epoch": 81.57,
+      "learning_rate": 0.00015969230769230766,
+      "loss": 0.0024,
+      "step": 408
+    },
+    {
+      "epoch": 81.76,
+      "learning_rate": 0.00015923076923076923,
+      "loss": 0.0045,
+      "step": 409
+    },
+    {
+      "epoch": 81.95,
+      "learning_rate": 0.00015876923076923078,
+      "loss": 0.0027,
+      "step": 410
+    },
+    {
+      "epoch": 82.19,
+      "learning_rate": 0.0001583076923076923,
+      "loss": 0.0026,
+      "step": 411
+    },
+    {
+      "epoch": 82.38,
+      "learning_rate": 0.00015784615384615384,
+      "loss": 0.0023,
+      "step": 412
+    },
+    {
+      "epoch": 82.57,
+      "learning_rate": 0.00015738461538461536,
+      "loss": 0.002,
+      "step": 413
+    },
+    {
+      "epoch": 82.76,
+      "learning_rate": 0.0001569230769230769,
+      "loss": 0.002,
+      "step": 414
+    },
+    {
+      "epoch": 82.95,
+      "learning_rate": 0.00015646153846153845,
+      "loss": 0.0021,
+      "step": 415
+    },
+    {
+      "epoch": 83.19,
+      "learning_rate": 0.000156,
+      "loss": 0.0027,
+      "step": 416
+    },
+    {
+      "epoch": 83.38,
+      "learning_rate": 0.00015553846153846152,
+      "loss": 0.002,
+      "step": 417
+    },
+    {
+      "epoch": 83.57,
+      "learning_rate": 0.00015507692307692306,
+      "loss": 0.0019,
+      "step": 418
+    },
+    {
+      "epoch": 83.76,
+      "learning_rate": 0.00015461538461538458,
+      "loss": 0.0021,
+      "step": 419
+    },
+    {
+      "epoch": 83.95,
+      "learning_rate": 0.00015415384615384613,
+      "loss": 0.0023,
+      "step": 420
+    },
+    {
+      "epoch": 83.95,
+      "eval_accuracy": 0.8971631205673759,
+      "eval_f1": 0.8997205755375566,
+      "eval_loss": 0.6551127433776855,
+      "eval_runtime": 16.2269,
+      "eval_samples_per_second": 17.379,
+      "eval_steps_per_second": 1.109,
+      "step": 420
+    },
+    {
+      "epoch": 84.19,
+      "learning_rate": 0.0001536923076923077,
+      "loss": 0.0025,
+      "step": 421
+    },
+    {
+      "epoch": 84.38,
+      "learning_rate": 0.00015323076923076922,
+      "loss": 0.0022,
+      "step": 422
+    },
+    {
+      "epoch": 84.57,
+      "learning_rate": 0.00015276923076923077,
+      "loss": 0.0019,
+      "step": 423
+    },
+    {
+      "epoch": 84.76,
+      "learning_rate": 0.00015230769230769228,
+      "loss": 0.0023,
+      "step": 424
+    },
+    {
+      "epoch": 84.95,
+      "learning_rate": 0.00015184615384615383,
+      "loss": 0.0098,
+      "step": 425
+    },
+    {
+      "epoch": 85.19,
+      "learning_rate": 0.00015138461538461535,
+      "loss": 0.0025,
+      "step": 426
+    },
+    {
+      "epoch": 85.38,
+      "learning_rate": 0.00015092307692307692,
+      "loss": 0.0023,
+      "step": 427
+    },
+    {
+      "epoch": 85.57,
+      "learning_rate": 0.00015046153846153844,
+      "loss": 0.0035,
+      "step": 428
+    },
+    {
+      "epoch": 85.76,
+      "learning_rate": 0.00015,
+      "loss": 0.0021,
+      "step": 429
+    },
+    {
+      "epoch": 85.95,
+      "learning_rate": 0.00014953846153846153,
+      "loss": 0.0017,
+      "step": 430
+    },
+    {
+      "epoch": 86.19,
+      "learning_rate": 0.00014907692307692305,
+      "loss": 0.0026,
+      "step": 431
+    },
+    {
+      "epoch": 86.38,
+      "learning_rate": 0.0001486153846153846,
+      "loss": 0.0021,
+      "step": 432
+    },
+    {
+      "epoch": 86.57,
+      "learning_rate": 0.00014815384615384614,
+      "loss": 0.002,
+      "step": 433
+    },
+    {
+      "epoch": 86.76,
+      "learning_rate": 0.0001476923076923077,
+      "loss": 0.014,
+      "step": 434
+    },
+    {
+      "epoch": 86.95,
+      "learning_rate": 0.00014723076923076924,
+      "loss": 0.002,
+      "step": 435
+    },
+    {
+      "epoch": 87.19,
+      "learning_rate": 0.00014676923076923075,
+      "loss": 0.0038,
+      "step": 436
+    },
+    {
+      "epoch": 87.38,
+      "learning_rate": 0.0001463076923076923,
+      "loss": 0.002,
+      "step": 437
+    },
+    {
+      "epoch": 87.57,
+      "learning_rate": 0.00014584615384615385,
+      "loss": 0.0028,
+      "step": 438
+    },
+    {
+      "epoch": 87.76,
+      "learning_rate": 0.00014538461538461537,
+      "loss": 0.0028,
+      "step": 439
+    },
+    {
+      "epoch": 87.95,
+      "learning_rate": 0.0001449230769230769,
+      "loss": 0.0027,
+      "step": 440
+    },
+    {
+      "epoch": 87.95,
+      "eval_accuracy": 0.8794326241134752,
+      "eval_f1": 0.8847583266741855,
+      "eval_loss": 0.6664106249809265,
+      "eval_runtime": 16.2215,
+      "eval_samples_per_second": 17.384,
+      "eval_steps_per_second": 1.11,
+      "step": 440
+    },
+    {
+      "epoch": 88.19,
+      "learning_rate": 0.00014446153846153846,
+      "loss": 0.0048,
+      "step": 441
+    },
+    {
+      "epoch": 88.38,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 0.0129,
+      "step": 442
+    },
+    {
+      "epoch": 88.57,
+      "learning_rate": 0.00014353846153846152,
+      "loss": 0.0036,
+      "step": 443
+    },
+    {
+      "epoch": 88.76,
+      "learning_rate": 0.00014307692307692307,
+      "loss": 0.002,
+      "step": 444
+    },
+    {
+      "epoch": 88.95,
+      "learning_rate": 0.00014261538461538461,
+      "loss": 0.002,
+      "step": 445
+    },
+    {
+      "epoch": 89.19,
+      "learning_rate": 0.00014215384615384613,
+      "loss": 0.0025,
+      "step": 446
+    },
+    {
+      "epoch": 89.38,
+      "learning_rate": 0.00014169230769230768,
+      "loss": 0.0022,
+      "step": 447
+    },
+    {
+      "epoch": 89.57,
+      "learning_rate": 0.00014123076923076923,
+      "loss": 0.0027,
+      "step": 448
+    },
+    {
+      "epoch": 89.76,
+      "learning_rate": 0.00014076923076923074,
+      "loss": 0.0022,
+      "step": 449
+    },
+    {
+      "epoch": 89.95,
+      "learning_rate": 0.0001403076923076923,
+      "loss": 0.0019,
+      "step": 450
+    },
+    {
+      "epoch": 90.19,
+      "learning_rate": 0.00013984615384615384,
+      "loss": 0.0022,
+      "step": 451
+    },
+    {
+      "epoch": 90.38,
+      "learning_rate": 0.00013938461538461536,
+      "loss": 0.0023,
+      "step": 452
+    },
+    {
+      "epoch": 90.57,
+      "learning_rate": 0.0001389230769230769,
+      "loss": 0.0019,
+      "step": 453
+    },
+    {
+      "epoch": 90.76,
+      "learning_rate": 0.00013846153846153845,
+      "loss": 0.0018,
+      "step": 454
+    },
+    {
+      "epoch": 90.95,
+      "learning_rate": 0.000138,
+      "loss": 0.0065,
+      "step": 455
+    },
+    {
+      "epoch": 91.19,
+      "learning_rate": 0.00013753846153846154,
+      "loss": 0.2232,
+      "step": 456
+    },
+    {
+      "epoch": 91.38,
+      "learning_rate": 0.00013707692307692306,
+      "loss": 0.0017,
+      "step": 457
+    },
+    {
+      "epoch": 91.57,
+      "learning_rate": 0.0001366153846153846,
+      "loss": 0.0018,
+      "step": 458
+    },
+    {
+      "epoch": 91.76,
+      "learning_rate": 0.00013615384615384615,
+      "loss": 0.0017,
+      "step": 459
+    },
+    {
+      "epoch": 91.95,
+      "learning_rate": 0.00013569230769230767,
+      "loss": 0.0019,
+      "step": 460
+    },
+    {
+      "epoch": 91.95,
+      "eval_accuracy": 0.9042553191489362,
+      "eval_f1": 0.9031716855807541,
+      "eval_loss": 0.5344293713569641,
+      "eval_runtime": 16.3497,
+      "eval_samples_per_second": 17.248,
+      "eval_steps_per_second": 1.101,
+      "step": 460
+    },
+    {
+      "epoch": 92.19,
+      "learning_rate": 0.00013523076923076922,
+      "loss": 0.0024,
+      "step": 461
+    },
+    {
+      "epoch": 92.38,
+      "learning_rate": 0.00013476923076923076,
+      "loss": 0.0022,
+      "step": 462
+    },
+    {
+      "epoch": 92.57,
+      "learning_rate": 0.00013430769230769228,
+      "loss": 0.0019,
+      "step": 463
+    },
+    {
+      "epoch": 92.76,
+      "learning_rate": 0.00013384615384615385,
+      "loss": 0.002,
+      "step": 464
+    },
+    {
+      "epoch": 92.95,
+      "learning_rate": 0.00013338461538461537,
+      "loss": 0.0022,
+      "step": 465
+    },
+    {
+      "epoch": 93.19,
+      "learning_rate": 0.00013292307692307692,
+      "loss": 0.0027,
+      "step": 466
+    },
+    {
+      "epoch": 93.38,
+      "learning_rate": 0.00013246153846153846,
+      "loss": 0.0023,
+      "step": 467
+    },
+    {
+      "epoch": 93.57,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 0.002,
+      "step": 468
+    },
+    {
+      "epoch": 93.76,
+      "learning_rate": 0.00013153846153846153,
+      "loss": 0.0022,
+      "step": 469
+    },
+    {
+      "epoch": 93.95,
+      "learning_rate": 0.00013107692307692308,
+      "loss": 0.0021,
+      "step": 470
+    },
+    {
+      "epoch": 94.19,
+      "learning_rate": 0.0001306153846153846,
+      "loss": 0.0026,
+      "step": 471
+    },
+    {
+      "epoch": 94.38,
+      "learning_rate": 0.00013015384615384614,
+      "loss": 0.0024,
+      "step": 472
+    },
+    {
+      "epoch": 94.57,
+      "learning_rate": 0.00012969230769230769,
+      "loss": 0.0024,
+      "step": 473
+    },
+    {
+      "epoch": 94.76,
+      "learning_rate": 0.00012923076923076923,
+      "loss": 0.002,
+      "step": 474
+    },
+    {
+      "epoch": 94.95,
+      "learning_rate": 0.00012876923076923075,
+      "loss": 0.0021,
+      "step": 475
+    },
+    {
+      "epoch": 95.19,
+      "learning_rate": 0.0001283076923076923,
+      "loss": 0.0022,
+      "step": 476
+    },
+    {
+      "epoch": 95.38,
+      "learning_rate": 0.00012784615384615384,
+      "loss": 0.0018,
+      "step": 477
+    },
+    {
+      "epoch": 95.57,
+      "learning_rate": 0.00012738461538461536,
+      "loss": 0.0023,
+      "step": 478
+    },
+    {
+      "epoch": 95.76,
+      "learning_rate": 0.0001269230769230769,
+      "loss": 0.0021,
+      "step": 479
+    },
+    {
+      "epoch": 95.95,
+      "learning_rate": 0.00012646153846153845,
+      "loss": 0.002,
+      "step": 480
+    },
+    {
+      "epoch": 95.95,
+      "eval_accuracy": 0.900709219858156,
+      "eval_f1": 0.8982682643855303,
+      "eval_loss": 0.5862882137298584,
+      "eval_runtime": 16.2593,
+      "eval_samples_per_second": 17.344,
+      "eval_steps_per_second": 1.107,
+      "step": 480
+    },
+    {
+      "epoch": 96.19,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 0.0024,
+      "step": 481
+    },
+    {
+      "epoch": 96.38,
+      "learning_rate": 0.00012553846153846152,
+      "loss": 0.0018,
+      "step": 482
+    },
+    {
+      "epoch": 96.57,
+      "learning_rate": 0.00012507692307692306,
+      "loss": 0.0021,
+      "step": 483
+    },
+    {
+      "epoch": 96.76,
+      "learning_rate": 0.0001246153846153846,
+      "loss": 0.0017,
+      "step": 484
+    },
+    {
+      "epoch": 96.95,
+      "learning_rate": 0.00012415384615384616,
+      "loss": 0.0017,
+      "step": 485
+    },
+    {
+      "epoch": 97.19,
+      "learning_rate": 0.00012369230769230768,
+      "loss": 0.0022,
+      "step": 486
+    },
+    {
+      "epoch": 97.38,
+      "learning_rate": 0.00012323076923076922,
+      "loss": 0.0016,
+      "step": 487
+    },
+    {
+      "epoch": 97.57,
+      "learning_rate": 0.00012276923076923077,
+      "loss": 0.0037,
+      "step": 488
+    },
+    {
+      "epoch": 97.76,
+      "learning_rate": 0.0001223076923076923,
+      "loss": 0.0018,
+      "step": 489
+    },
+    {
+      "epoch": 97.95,
+      "learning_rate": 0.00012184615384615383,
+      "loss": 0.0017,
+      "step": 490
+    },
+    {
+      "epoch": 98.19,
+      "learning_rate": 0.00012138461538461537,
+      "loss": 0.0021,
+      "step": 491
+    },
+    {
+      "epoch": 98.38,
+      "learning_rate": 0.00012092307692307691,
+      "loss": 0.0016,
+      "step": 492
+    },
+    {
+      "epoch": 98.57,
+      "learning_rate": 0.00012046153846153844,
+      "loss": 0.0017,
+      "step": 493
+    },
+    {
+      "epoch": 98.76,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.0019,
+      "step": 494
+    },
+    {
+      "epoch": 98.95,
+      "learning_rate": 0.00011953846153846154,
+      "loss": 0.0016,
+      "step": 495
+    },
+    {
+      "epoch": 99.19,
+      "learning_rate": 0.00011907692307692307,
+      "loss": 0.0021,
+      "step": 496
+    },
+    {
+      "epoch": 99.38,
+      "learning_rate": 0.0001186153846153846,
+      "loss": 0.0015,
+      "step": 497
+    },
+    {
+      "epoch": 99.57,
+      "learning_rate": 0.00011815384615384615,
+      "loss": 0.0016,
+      "step": 498
+    },
+    {
+      "epoch": 99.76,
+      "learning_rate": 0.00011769230769230768,
+      "loss": 0.0016,
+      "step": 499
+    },
+    {
+      "epoch": 99.95,
+      "learning_rate": 0.00011723076923076921,
+      "loss": 0.0015,
+      "step": 500
+    },
+    {
+      "epoch": 99.95,
+      "eval_accuracy": 0.9042553191489362,
+      "eval_f1": 0.9047123942914032,
+      "eval_loss": 0.5714772343635559,
+      "eval_runtime": 16.3018,
+      "eval_samples_per_second": 17.299,
+      "eval_steps_per_second": 1.104,
+      "step": 500
+    },
+    {
+      "epoch": 100.19,
+      "learning_rate": 0.00011676923076923076,
+      "loss": 0.0021,
+      "step": 501
+    },
+    {
+      "epoch": 100.38,
+      "learning_rate": 0.00011630769230769229,
+      "loss": 0.0017,
+      "step": 502
+    },
+    {
+      "epoch": 100.57,
+      "learning_rate": 0.00011584615384615385,
+      "loss": 0.0017,
+      "step": 503
+    },
+    {
+      "epoch": 100.76,
+      "learning_rate": 0.00011538461538461538,
+      "loss": 0.0015,
+      "step": 504
+    },
+    {
+      "epoch": 100.95,
+      "learning_rate": 0.00011492307692307691,
+      "loss": 0.0017,
+      "step": 505
+    },
+    {
+      "epoch": 101.19,
+      "learning_rate": 0.00011446153846153846,
+      "loss": 0.0018,
+      "step": 506
+    },
+    {
+      "epoch": 101.38,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 0.0016,
+      "step": 507
+    },
+    {
+      "epoch": 101.57,
+      "learning_rate": 0.00011353846153846153,
+      "loss": 0.0018,
+      "step": 508
+    },
+    {
+      "epoch": 101.76,
+      "learning_rate": 0.00011307692307692307,
+      "loss": 0.0016,
+      "step": 509
+    },
+    {
+      "epoch": 101.95,
+      "learning_rate": 0.0001126153846153846,
+      "loss": 0.0025,
+      "step": 510
+    },
+    {
+      "epoch": 102.19,
+      "learning_rate": 0.00011215384615384614,
+      "loss": 0.0018,
+      "step": 511
+    },
+    {
+      "epoch": 102.38,
+      "learning_rate": 0.00011169230769230768,
+      "loss": 0.0015,
+      "step": 512
+    },
+    {
+      "epoch": 102.57,
+      "learning_rate": 0.00011123076923076923,
+      "loss": 0.0081,
+      "step": 513
+    },
+    {
+      "epoch": 102.76,
+      "learning_rate": 0.00011076923076923076,
+      "loss": 0.0014,
+      "step": 514
+    },
+    {
+      "epoch": 102.95,
+      "learning_rate": 0.0001103076923076923,
+      "loss": 0.0013,
+      "step": 515
+    },
+    {
+      "epoch": 103.19,
+      "learning_rate": 0.00010984615384615384,
+      "loss": 0.0018,
+      "step": 516
+    },
+    {
+      "epoch": 103.38,
+      "learning_rate": 0.00010938461538461537,
+      "loss": 0.0015,
+      "step": 517
+    },
+    {
+      "epoch": 103.57,
+      "learning_rate": 0.00010892307692307692,
+      "loss": 0.0015,
+      "step": 518
+    },
+    {
+      "epoch": 103.76,
+      "learning_rate": 0.00010846153846153845,
+      "loss": 0.0016,
+      "step": 519
+    },
+    {
+      "epoch": 103.95,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 0.0016,
+      "step": 520
+    },
+    {
+      "epoch": 103.95,
+      "eval_accuracy": 0.8936170212765957,
+      "eval_f1": 0.8955507516991726,
+      "eval_loss": 0.5614662766456604,
+      "eval_runtime": 16.1752,
+      "eval_samples_per_second": 17.434,
+      "eval_steps_per_second": 1.113,
+      "step": 520
+    },
+    {
+      "epoch": 104.19,
+      "learning_rate": 0.00010753846153846153,
+      "loss": 0.0023,
+      "step": 521
+    },
+    {
+      "epoch": 104.38,
+      "learning_rate": 0.00010707692307692306,
+      "loss": 0.0015,
+      "step": 522
+    },
+    {
+      "epoch": 104.57,
+      "learning_rate": 0.00010661538461538461,
+      "loss": 0.0018,
+      "step": 523
+    },
+    {
+      "epoch": 104.76,
+      "learning_rate": 0.00010615384615384615,
+      "loss": 0.0017,
+      "step": 524
+    },
+    {
+      "epoch": 104.95,
+      "learning_rate": 0.00010569230769230769,
+      "loss": 0.0015,
+      "step": 525
+    },
+    {
+      "epoch": 105.19,
+      "learning_rate": 0.00010523076923076922,
+      "loss": 0.0028,
+      "step": 526
+    },
+    {
+      "epoch": 105.38,
+      "learning_rate": 0.00010476923076923076,
+      "loss": 0.0025,
+      "step": 527
+    },
+    {
+      "epoch": 105.57,
+      "learning_rate": 0.0001043076923076923,
+      "loss": 0.0146,
+      "step": 528
+    },
+    {
+      "epoch": 105.76,
+      "learning_rate": 0.00010384615384615383,
+      "loss": 0.0014,
+      "step": 529
+    },
+    {
+      "epoch": 105.95,
+      "learning_rate": 0.00010338461538461537,
+      "loss": 0.0015,
+      "step": 530
+    },
+    {
+      "epoch": 106.19,
+      "learning_rate": 0.00010292307692307691,
+      "loss": 0.0017,
+      "step": 531
+    },
+    {
+      "epoch": 106.38,
+      "learning_rate": 0.00010246153846153844,
+      "loss": 0.0041,
+      "step": 532
+    },
+    {
+      "epoch": 106.57,
+      "learning_rate": 0.000102,
+      "loss": 0.0014,
+      "step": 533
+    },
+    {
+      "epoch": 106.76,
+      "learning_rate": 0.00010153846153846153,
+      "loss": 0.0014,
+      "step": 534
+    },
+    {
+      "epoch": 106.95,
+      "learning_rate": 0.00010107692307692308,
+      "loss": 0.0015,
+      "step": 535
+    },
+    {
+      "epoch": 107.19,
+      "learning_rate": 0.00010061538461538461,
+      "loss": 0.002,
+      "step": 536
+    },
+    {
+      "epoch": 107.38,
+      "learning_rate": 0.00010015384615384614,
+      "loss": 0.0089,
+      "step": 537
+    },
+    {
+      "epoch": 107.57,
+      "learning_rate": 9.969230769230769e-05,
+      "loss": 0.0014,
+      "step": 538
+    },
+    {
+      "epoch": 107.76,
+      "learning_rate": 9.923076923076922e-05,
+      "loss": 0.0015,
+      "step": 539
+    },
+    {
+      "epoch": 107.95,
+      "learning_rate": 9.876923076923075e-05,
+      "loss": 0.0014,
+      "step": 540
+    },
+    {
+      "epoch": 107.95,
+      "eval_accuracy": 0.8936170212765957,
+      "eval_f1": 0.8965114077133559,
+      "eval_loss": 0.6353234052658081,
+      "eval_runtime": 16.1498,
+      "eval_samples_per_second": 17.462,
+      "eval_steps_per_second": 1.115,
+      "step": 540
+    },
+    {
+      "epoch": 108.19,
+      "learning_rate": 9.83076923076923e-05,
+      "loss": 0.0018,
+      "step": 541
+    },
+    {
+      "epoch": 108.38,
+      "learning_rate": 9.784615384615383e-05,
+      "loss": 0.0015,
+      "step": 542
+    },
+    {
+      "epoch": 108.57,
+      "learning_rate": 9.738461538461538e-05,
+      "loss": 0.0015,
+      "step": 543
+    },
+    {
+      "epoch": 108.76,
+      "learning_rate": 9.692307692307692e-05,
+      "loss": 0.0014,
+      "step": 544
+    },
+    {
+      "epoch": 108.95,
+      "learning_rate": 9.646153846153846e-05,
+      "loss": 0.0014,
+      "step": 545
+    },
+    {
+      "epoch": 109.19,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 0.0017,
+      "step": 546
+    },
+    {
+      "epoch": 109.38,
+      "learning_rate": 9.553846153846153e-05,
+      "loss": 0.0015,
+      "step": 547
+    },
+    {
+      "epoch": 109.57,
+      "learning_rate": 9.507692307692307e-05,
+      "loss": 0.0014,
+      "step": 548
+    },
+    {
+      "epoch": 109.76,
+      "learning_rate": 9.46153846153846e-05,
+      "loss": 0.0017,
+      "step": 549
+    },
+    {
+      "epoch": 109.95,
+      "learning_rate": 9.415384615384615e-05,
+      "loss": 0.0014,
+      "step": 550
+    },
+    {
+      "epoch": 110.19,
+      "learning_rate": 9.369230769230768e-05,
+      "loss": 0.0018,
+      "step": 551
+    },
+    {
+      "epoch": 110.38,
+      "learning_rate": 9.323076923076921e-05,
+      "loss": 0.0017,
+      "step": 552
+    },
+    {
+      "epoch": 110.57,
+      "learning_rate": 9.276923076923077e-05,
+      "loss": 0.2056,
+      "step": 553
+    },
+    {
+      "epoch": 110.76,
+      "learning_rate": 9.23076923076923e-05,
+      "loss": 0.0015,
+      "step": 554
+    },
+    {
+      "epoch": 110.95,
+      "learning_rate": 9.184615384615384e-05,
+      "loss": 0.0013,
+      "step": 555
+    },
+    {
+      "epoch": 111.19,
+      "learning_rate": 9.138461538461538e-05,
+      "loss": 0.0018,
+      "step": 556
+    },
+    {
+      "epoch": 111.38,
+      "learning_rate": 9.092307692307691e-05,
+      "loss": 0.0014,
+      "step": 557
+    },
+    {
+      "epoch": 111.57,
+      "learning_rate": 9.046153846153845e-05,
+      "loss": 0.0013,
+      "step": 558
+    },
+    {
+      "epoch": 111.76,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 0.0013,
+      "step": 559
+    },
+    {
+      "epoch": 111.95,
+      "learning_rate": 8.953846153846152e-05,
+      "loss": 0.0014,
+      "step": 560
+    },
+    {
+      "epoch": 111.95,
+      "eval_accuracy": 0.900709219858156,
+      "eval_f1": 0.9040711854187397,
+      "eval_loss": 0.5592705011367798,
+      "eval_runtime": 16.2023,
+      "eval_samples_per_second": 17.405,
+      "eval_steps_per_second": 1.111,
+      "step": 560
+    },
+    {
+      "epoch": 112.19,
+      "learning_rate": 8.907692307692306e-05,
+      "loss": 0.0019,
+      "step": 561
+    },
+    {
+      "epoch": 112.38,
+      "learning_rate": 8.861538461538462e-05,
+      "loss": 0.0014,
+      "step": 562
+    },
+    {
+      "epoch": 112.57,
+      "learning_rate": 8.815384615384615e-05,
+      "loss": 0.0016,
+      "step": 563
+    },
+    {
+      "epoch": 112.76,
+      "learning_rate": 8.76923076923077e-05,
+      "loss": 0.0144,
+      "step": 564
+    },
+    {
+      "epoch": 112.95,
+      "learning_rate": 8.723076923076923e-05,
+      "loss": 0.0021,
+      "step": 565
+    },
+    {
+      "epoch": 113.19,
+      "learning_rate": 8.676923076923076e-05,
+      "loss": 0.0017,
+      "step": 566
+    },
+    {
+      "epoch": 113.38,
+      "learning_rate": 8.63076923076923e-05,
+      "loss": 0.0012,
+      "step": 567
+    },
+    {
+      "epoch": 113.57,
+      "learning_rate": 8.584615384615384e-05,
+      "loss": 0.0014,
+      "step": 568
+    },
+    {
+      "epoch": 113.76,
+      "learning_rate": 8.538461538461537e-05,
+      "loss": 0.0014,
+      "step": 569
+    },
+    {
+      "epoch": 113.95,
+      "learning_rate": 8.492307692307692e-05,
+      "loss": 0.0013,
+      "step": 570
+    },
+    {
+      "epoch": 114.19,
+      "learning_rate": 8.446153846153845e-05,
+      "loss": 0.0018,
+      "step": 571
+    },
+    {
+      "epoch": 114.38,
+      "learning_rate": 8.4e-05,
+      "loss": 0.0016,
+      "step": 572
+    },
+    {
+      "epoch": 114.57,
+      "learning_rate": 8.353846153846154e-05,
+      "loss": 0.0014,
+      "step": 573
+    },
+    {
+      "epoch": 114.76,
+      "learning_rate": 8.307692307692307e-05,
+      "loss": 0.0019,
+      "step": 574
+    },
+    {
+      "epoch": 114.95,
+      "learning_rate": 8.26153846153846e-05,
+      "loss": 0.0016,
+      "step": 575
+    },
+    {
+      "epoch": 115.19,
+      "learning_rate": 8.215384615384615e-05,
+      "loss": 0.0027,
+      "step": 576
+    },
+    {
+      "epoch": 115.38,
+      "learning_rate": 8.169230769230768e-05,
+      "loss": 0.0025,
+      "step": 577
+    },
+    {
+      "epoch": 115.57,
+      "learning_rate": 8.123076923076922e-05,
+      "loss": 0.0015,
+      "step": 578
+    },
+    {
+      "epoch": 115.76,
+      "learning_rate": 8.076923076923076e-05,
+      "loss": 0.0638,
+      "step": 579
+    },
+    {
+      "epoch": 115.95,
+      "learning_rate": 8.03076923076923e-05,
+      "loss": 0.0013,
+      "step": 580
+    },
+    {
+      "epoch": 115.95,
+      "eval_accuracy": 0.8936170212765957,
+      "eval_f1": 0.8976616729827315,
+      "eval_loss": 0.6041242480278015,
+      "eval_runtime": 16.2096,
+      "eval_samples_per_second": 17.397,
+      "eval_steps_per_second": 1.11,
+      "step": 580
+    },
+    {
+      "epoch": 116.19,
+      "learning_rate": 7.984615384615383e-05,
+      "loss": 0.0019,
+      "step": 581
+    },
+    {
+      "epoch": 116.38,
+      "learning_rate": 7.938461538461539e-05,
+      "loss": 0.0013,
+      "step": 582
+    },
+    {
+      "epoch": 116.57,
+      "learning_rate": 7.892307692307692e-05,
+      "loss": 0.0015,
+      "step": 583
+    },
+    {
+      "epoch": 116.76,
+      "learning_rate": 7.846153846153845e-05,
+      "loss": 0.2165,
+      "step": 584
+    },
+    {
+      "epoch": 116.95,
+      "learning_rate": 7.8e-05,
+      "loss": 0.0012,
+      "step": 585
+    },
+    {
+      "epoch": 117.19,
+      "learning_rate": 7.753846153846153e-05,
+      "loss": 0.0398,
+      "step": 586
+    },
+    {
+      "epoch": 117.38,
+      "learning_rate": 7.707692307692306e-05,
+      "loss": 0.0012,
+      "step": 587
+    },
+    {
+      "epoch": 117.57,
+      "learning_rate": 7.661538461538461e-05,
+      "loss": 0.0018,
+      "step": 588
+    },
+    {
+      "epoch": 117.76,
+      "learning_rate": 7.615384615384614e-05,
+      "loss": 0.0014,
+      "step": 589
+    },
+    {
+      "epoch": 117.95,
+      "learning_rate": 7.569230769230767e-05,
+      "loss": 0.0013,
+      "step": 590
+    },
+    {
+      "epoch": 118.19,
+      "learning_rate": 7.523076923076922e-05,
+      "loss": 0.0018,
+      "step": 591
+    },
+    {
+      "epoch": 118.38,
+      "learning_rate": 7.476923076923077e-05,
+      "loss": 0.0012,
+      "step": 592
+    },
+    {
+      "epoch": 118.57,
+      "learning_rate": 7.43076923076923e-05,
+      "loss": 0.0013,
+      "step": 593
+    },
+    {
+      "epoch": 118.76,
+      "learning_rate": 7.384615384615384e-05,
+      "loss": 0.0014,
+      "step": 594
+    },
+    {
+      "epoch": 118.95,
+      "learning_rate": 7.338461538461538e-05,
+      "loss": 0.0014,
+      "step": 595
+    },
+    {
+      "epoch": 119.19,
+      "learning_rate": 7.292307692307692e-05,
+      "loss": 0.002,
+      "step": 596
+    },
+    {
+      "epoch": 119.38,
+      "learning_rate": 7.246153846153846e-05,
+      "loss": 0.0013,
+      "step": 597
+    },
+    {
+      "epoch": 119.57,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 0.3931,
+      "step": 598
+    },
+    {
+      "epoch": 119.76,
+      "learning_rate": 7.153846153846153e-05,
+      "loss": 0.0013,
+      "step": 599
+    },
+    {
+      "epoch": 119.95,
+      "learning_rate": 7.107692307692307e-05,
+      "loss": 0.0013,
+      "step": 600
+    },
+    {
+      "epoch": 119.95,
+      "eval_accuracy": 0.900709219858156,
+      "eval_f1": 0.9026478709738444,
+      "eval_loss": 0.5793901681900024,
+      "eval_runtime": 16.1917,
+      "eval_samples_per_second": 17.416,
+      "eval_steps_per_second": 1.112,
+      "step": 600
+    },
+    {
+      "epoch": 120.19,
+      "learning_rate": 7.061538461538461e-05,
+      "loss": 0.0016,
+      "step": 601
+    },
+    {
+      "epoch": 120.38,
+      "learning_rate": 7.015384615384615e-05,
+      "loss": 0.0013,
+      "step": 602
+    },
+    {
+      "epoch": 120.57,
+      "learning_rate": 6.969230769230768e-05,
+      "loss": 0.002,
+      "step": 603
+    },
+    {
+      "epoch": 120.76,
+      "learning_rate": 6.923076923076922e-05,
+      "loss": 0.0013,
+      "step": 604
+    },
+    {
+      "epoch": 120.95,
+      "learning_rate": 6.876923076923077e-05,
+      "loss": 0.0013,
+      "step": 605
+    },
+    {
+      "epoch": 121.19,
+      "learning_rate": 6.83076923076923e-05,
+      "loss": 0.005,
+      "step": 606
+    },
+    {
+      "epoch": 121.38,
+      "learning_rate": 6.784615384615383e-05,
+      "loss": 0.0014,
+      "step": 607
+    },
+    {
+      "epoch": 121.57,
+      "learning_rate": 6.738461538461538e-05,
+      "loss": 0.0013,
+      "step": 608
+    },
+    {
+      "epoch": 121.76,
+      "learning_rate": 6.692307692307693e-05,
+      "loss": 0.0012,
+      "step": 609
+    },
+    {
+      "epoch": 121.95,
+      "learning_rate": 6.646153846153846e-05,
+      "loss": 0.0013,
+      "step": 610
+    },
+    {
+      "epoch": 122.19,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 0.0017,
+      "step": 611
+    },
+    {
+      "epoch": 122.38,
+      "learning_rate": 6.553846153846154e-05,
+      "loss": 0.4012,
+      "step": 612
+    },
+    {
+      "epoch": 122.57,
+      "learning_rate": 6.507692307692307e-05,
+      "loss": 0.0012,
+      "step": 613
+    },
+    {
+      "epoch": 122.76,
+      "learning_rate": 6.461538461538462e-05,
+      "loss": 0.0016,
+      "step": 614
+    },
+    {
+      "epoch": 122.95,
+      "learning_rate": 6.415384615384615e-05,
+      "loss": 0.0394,
+      "step": 615
+    },
+    {
+      "epoch": 123.19,
+      "learning_rate": 6.369230769230768e-05,
+      "loss": 0.0047,
+      "step": 616
+    },
+    {
+      "epoch": 123.38,
+      "learning_rate": 6.323076923076923e-05,
+      "loss": 0.0013,
+      "step": 617
+    },
+    {
+      "epoch": 123.57,
+      "learning_rate": 6.276923076923076e-05,
+      "loss": 0.0019,
+      "step": 618
+    },
+    {
+      "epoch": 123.76,
+      "learning_rate": 6.23076923076923e-05,
+      "loss": 0.0013,
+      "step": 619
+    },
+    {
+      "epoch": 123.95,
+      "learning_rate": 6.184615384615384e-05,
+      "loss": 0.0012,
+      "step": 620
+    },
+    {
+      "epoch": 123.95,
+      "eval_accuracy": 0.8971631205673759,
+      "eval_f1": 0.9003317549557757,
+      "eval_loss": 0.6857544779777527,
+      "eval_runtime": 16.223,
+      "eval_samples_per_second": 17.383,
+      "eval_steps_per_second": 1.11,
+      "step": 620
+    },
+    {
+      "epoch": 124.19,
+      "learning_rate": 6.138461538461538e-05,
+      "loss": 0.0024,
+      "step": 621
+    },
+    {
+      "epoch": 124.38,
+      "learning_rate": 6.0923076923076916e-05,
+      "loss": 0.0014,
+      "step": 622
+    },
+    {
+      "epoch": 124.57,
+      "learning_rate": 6.0461538461538456e-05,
+      "loss": 0.0015,
+      "step": 623
+    },
+    {
+      "epoch": 124.76,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.0012,
+      "step": 624
+    },
+    {
+      "epoch": 124.95,
+      "learning_rate": 5.9538461538461534e-05,
+      "loss": 0.0013,
+      "step": 625
+    },
+    {
+      "epoch": 125.19,
+      "learning_rate": 5.907692307692307e-05,
+      "loss": 0.0016,
+      "step": 626
+    },
+    {
+      "epoch": 125.38,
+      "learning_rate": 5.8615384615384606e-05,
+      "loss": 0.0012,
+      "step": 627
+    },
+    {
+      "epoch": 125.57,
+      "learning_rate": 5.8153846153846145e-05,
+      "loss": 0.0014,
+      "step": 628
+    },
+    {
+      "epoch": 125.76,
+      "learning_rate": 5.769230769230769e-05,
+      "loss": 0.0014,
+      "step": 629
+    },
+    {
+      "epoch": 125.95,
+      "learning_rate": 5.723076923076923e-05,
+      "loss": 0.0013,
+      "step": 630
+    },
+    {
+      "epoch": 126.19,
+      "learning_rate": 5.676923076923076e-05,
+      "loss": 0.002,
+      "step": 631
+    },
+    {
+      "epoch": 126.38,
+      "learning_rate": 5.63076923076923e-05,
+      "loss": 0.0013,
+      "step": 632
+    },
+    {
+      "epoch": 126.57,
+      "learning_rate": 5.584615384615384e-05,
+      "loss": 0.0013,
+      "step": 633
+    },
+    {
+      "epoch": 126.76,
+      "learning_rate": 5.538461538461538e-05,
+      "loss": 0.0013,
+      "step": 634
+    },
+    {
+      "epoch": 126.95,
+      "learning_rate": 5.492307692307692e-05,
+      "loss": 0.0013,
+      "step": 635
+    },
+    {
+      "epoch": 127.19,
+      "learning_rate": 5.446153846153846e-05,
+      "loss": 0.0017,
+      "step": 636
+    },
+    {
+      "epoch": 127.38,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 0.0013,
+      "step": 637
+    },
+    {
+      "epoch": 127.57,
+      "learning_rate": 5.353846153846153e-05,
+      "loss": 0.0709,
+      "step": 638
+    },
+    {
+      "epoch": 127.76,
+      "learning_rate": 5.3076923076923076e-05,
+      "loss": 0.0014,
+      "step": 639
+    },
+    {
+      "epoch": 127.95,
+      "learning_rate": 5.261538461538461e-05,
+      "loss": 0.0013,
+      "step": 640
+    },
+    {
+      "epoch": 127.95,
+      "eval_accuracy": 0.8971631205673759,
+      "eval_f1": 0.9002257118158272,
+      "eval_loss": 0.6730301380157471,
+      "eval_runtime": 16.1062,
+      "eval_samples_per_second": 17.509,
+      "eval_steps_per_second": 1.118,
+      "step": 640
+    },
+    {
+      "epoch": 128.19,
+      "learning_rate": 5.215384615384615e-05,
+      "loss": 0.0015,
+      "step": 641
+    },
+    {
+      "epoch": 128.38,
+      "learning_rate": 5.169230769230769e-05,
+      "loss": 0.0014,
+      "step": 642
+    },
+    {
+      "epoch": 128.57,
+      "learning_rate": 5.123076923076922e-05,
+      "loss": 0.0013,
+      "step": 643
+    },
+    {
+      "epoch": 128.76,
+      "learning_rate": 5.0769230769230766e-05,
+      "loss": 0.0012,
+      "step": 644
+    },
+    {
+      "epoch": 128.95,
+      "learning_rate": 5.0307692307692305e-05,
+      "loss": 0.0012,
+      "step": 645
+    },
+    {
+      "epoch": 129.19,
+      "learning_rate": 4.9846153846153844e-05,
+      "loss": 0.002,
+      "step": 646
+    },
+    {
+      "epoch": 129.38,
+      "learning_rate": 4.938461538461538e-05,
+      "loss": 0.0012,
+      "step": 647
+    },
+    {
+      "epoch": 129.57,
+      "learning_rate": 4.8923076923076916e-05,
+      "loss": 0.0013,
+      "step": 648
+    },
+    {
+      "epoch": 129.76,
+      "learning_rate": 4.846153846153846e-05,
+      "loss": 0.0012,
+      "step": 649
+    },
+    {
+      "epoch": 129.95,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 0.0013,
+      "step": 650
+    },
+    {
+      "epoch": 130.19,
+      "learning_rate": 4.7538461538461534e-05,
+      "loss": 0.0015,
+      "step": 651
+    },
+    {
+      "epoch": 130.38,
+      "learning_rate": 4.707692307692307e-05,
+      "loss": 0.0012,
+      "step": 652
+    },
+    {
+      "epoch": 130.57,
+      "learning_rate": 4.6615384615384605e-05,
+      "loss": 0.0014,
+      "step": 653
+    },
+    {
+      "epoch": 130.76,
+      "learning_rate": 4.615384615384615e-05,
+      "loss": 0.0013,
+      "step": 654
+    },
+    {
+      "epoch": 130.95,
+      "learning_rate": 4.569230769230769e-05,
+      "loss": 0.0012,
+      "step": 655
+    },
+    {
+      "epoch": 131.19,
+      "learning_rate": 4.523076923076922e-05,
+      "loss": 0.0015,
+      "step": 656
+    },
+    {
+      "epoch": 131.38,
+      "learning_rate": 4.476923076923076e-05,
+      "loss": 0.0012,
+      "step": 657
+    },
+    {
+      "epoch": 131.57,
+      "learning_rate": 4.430769230769231e-05,
+      "loss": 0.0012,
+      "step": 658
+    },
+    {
+      "epoch": 131.76,
+      "learning_rate": 4.384615384615385e-05,
+      "loss": 0.0012,
+      "step": 659
+    },
+    {
+      "epoch": 131.95,
+      "learning_rate": 4.338461538461538e-05,
+      "loss": 0.0013,
+      "step": 660
+    },
+    {
+      "epoch": 131.95,
+      "eval_accuracy": 0.9113475177304965,
+      "eval_f1": 0.9145506874101489,
+      "eval_loss": 0.5706735849380493,
+      "eval_runtime": 16.0962,
+      "eval_samples_per_second": 17.52,
+      "eval_steps_per_second": 1.118,
+      "step": 660
+    },
+    {
+      "epoch": 132.19,
+      "learning_rate": 4.292307692307692e-05,
+      "loss": 0.0016,
+      "step": 661
+    },
+    {
+      "epoch": 132.38,
+      "learning_rate": 4.246153846153846e-05,
+      "loss": 0.0012,
+      "step": 662
+    },
+    {
+      "epoch": 132.57,
+      "learning_rate": 4.2e-05,
+      "loss": 0.0012,
+      "step": 663
+    },
+    {
+      "epoch": 132.76,
+      "learning_rate": 4.153846153846154e-05,
+      "loss": 0.0013,
+      "step": 664
+    },
+    {
+      "epoch": 132.95,
+      "learning_rate": 4.1076923076923076e-05,
+      "loss": 0.0012,
+      "step": 665
+    },
+    {
+      "epoch": 133.19,
+      "learning_rate": 4.061538461538461e-05,
+      "loss": 0.0014,
+      "step": 666
+    },
+    {
+      "epoch": 133.38,
+      "learning_rate": 4.015384615384615e-05,
+      "loss": 0.0083,
+      "step": 667
+    },
+    {
+      "epoch": 133.57,
+      "learning_rate": 3.9692307692307694e-05,
+      "loss": 0.0013,
+      "step": 668
+    },
+    {
+      "epoch": 133.76,
+      "learning_rate": 3.9230769230769226e-05,
+      "loss": 0.0012,
+      "step": 669
+    },
+    {
+      "epoch": 133.95,
+      "learning_rate": 3.8769230769230766e-05,
+      "loss": 0.0012,
+      "step": 670
+    },
+    {
+      "epoch": 134.19,
+      "learning_rate": 3.8307692307692305e-05,
+      "loss": 0.0015,
+      "step": 671
+    },
+    {
+      "epoch": 134.38,
+      "learning_rate": 3.784615384615384e-05,
+      "loss": 0.0011,
+      "step": 672
+    },
+    {
+      "epoch": 134.57,
+      "learning_rate": 3.738461538461538e-05,
+      "loss": 0.0012,
+      "step": 673
+    },
+    {
+      "epoch": 134.76,
+      "learning_rate": 3.692307692307692e-05,
+      "loss": 0.0012,
+      "step": 674
+    },
+    {
+      "epoch": 134.95,
+      "learning_rate": 3.646153846153846e-05,
+      "loss": 0.0013,
+      "step": 675
+    },
+    {
+      "epoch": 135.19,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 0.0015,
+      "step": 676
+    },
+    {
+      "epoch": 135.38,
+      "learning_rate": 3.553846153846153e-05,
+      "loss": 0.0012,
+      "step": 677
+    },
+    {
+      "epoch": 135.57,
+      "learning_rate": 3.507692307692307e-05,
+      "loss": 0.0062,
+      "step": 678
+    },
+    {
+      "epoch": 135.76,
+      "learning_rate": 3.461538461538461e-05,
+      "loss": 0.0011,
+      "step": 679
+    },
+    {
+      "epoch": 135.95,
+      "learning_rate": 3.415384615384615e-05,
+      "loss": 0.0012,
+      "step": 680
+    },
+    {
+      "epoch": 135.95,
+      "eval_accuracy": 0.9113475177304965,
+      "eval_f1": 0.9153456298930708,
+      "eval_loss": 0.560399055480957,
+      "eval_runtime": 16.2091,
+      "eval_samples_per_second": 17.398,
+      "eval_steps_per_second": 1.11,
+      "step": 680
+    },
+    {
+      "epoch": 136.19,
+      "learning_rate": 3.369230769230769e-05,
+      "loss": 0.0015,
+      "step": 681
+    },
+    {
+      "epoch": 136.38,
+      "learning_rate": 3.323076923076923e-05,
+      "loss": 0.0013,
+      "step": 682
+    },
+    {
+      "epoch": 136.57,
+      "learning_rate": 3.276923076923077e-05,
+      "loss": 0.0013,
+      "step": 683
+    },
+    {
+      "epoch": 136.76,
+      "learning_rate": 3.230769230769231e-05,
+      "loss": 0.0012,
+      "step": 684
+    },
+    {
+      "epoch": 136.95,
+      "learning_rate": 3.184615384615384e-05,
+      "loss": 0.0013,
+      "step": 685
+    },
+    {
+      "epoch": 137.19,
+      "learning_rate": 3.138461538461538e-05,
+      "loss": 0.0014,
+      "step": 686
+    },
+    {
+      "epoch": 137.38,
+      "learning_rate": 3.092307692307692e-05,
+      "loss": 0.0012,
+      "step": 687
+    },
+    {
+      "epoch": 137.57,
+      "learning_rate": 3.0461538461538458e-05,
+      "loss": 0.0014,
+      "step": 688
+    },
+    {
+      "epoch": 137.76,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 0.0012,
+      "step": 689
+    },
+    {
+      "epoch": 137.95,
+      "learning_rate": 2.9538461538461537e-05,
+      "loss": 0.0012,
+      "step": 690
+    },
+    {
+      "epoch": 138.19,
+      "learning_rate": 2.9076923076923072e-05,
+      "loss": 0.0014,
+      "step": 691
+    },
+    {
+      "epoch": 138.38,
+      "learning_rate": 2.8615384615384615e-05,
+      "loss": 0.0012,
+      "step": 692
+    },
+    {
+      "epoch": 138.57,
+      "learning_rate": 2.815384615384615e-05,
+      "loss": 0.0023,
+      "step": 693
+    },
+    {
+      "epoch": 138.76,
+      "learning_rate": 2.769230769230769e-05,
+      "loss": 0.0012,
+      "step": 694
+    },
+    {
+      "epoch": 138.95,
+      "learning_rate": 2.723076923076923e-05,
+      "loss": 0.0012,
+      "step": 695
+    },
+    {
+      "epoch": 139.19,
+      "learning_rate": 2.6769230769230765e-05,
+      "loss": 0.0017,
+      "step": 696
+    },
+    {
+      "epoch": 139.38,
+      "learning_rate": 2.6307692307692304e-05,
+      "loss": 0.0012,
+      "step": 697
+    },
+    {
+      "epoch": 139.57,
+      "learning_rate": 2.5846153846153844e-05,
+      "loss": 0.0012,
+      "step": 698
+    },
+    {
+      "epoch": 139.76,
+      "learning_rate": 2.5384615384615383e-05,
+      "loss": 0.0011,
+      "step": 699
+    },
+    {
+      "epoch": 139.95,
+      "learning_rate": 2.4923076923076922e-05,
+      "loss": 0.0019,
+      "step": 700
+    },
+    {
+      "epoch": 139.95,
+      "eval_accuracy": 0.9078014184397163,
+      "eval_f1": 0.9114341773753465,
+      "eval_loss": 0.5468190312385559,
+      "eval_runtime": 16.0855,
+      "eval_samples_per_second": 17.531,
+      "eval_steps_per_second": 1.119,
+      "step": 700
+    },
+    {
+      "epoch": 140.19,
+      "learning_rate": 2.4461538461538458e-05,
+      "loss": 0.0015,
+      "step": 701
+    },
+    {
+      "epoch": 140.38,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 0.0011,
+      "step": 702
+    },
+    {
+      "epoch": 140.57,
+      "learning_rate": 2.3538461538461536e-05,
+      "loss": 0.0012,
+      "step": 703
+    },
+    {
+      "epoch": 140.76,
+      "learning_rate": 2.3076923076923076e-05,
+      "loss": 0.0012,
+      "step": 704
+    },
+    {
+      "epoch": 140.95,
+      "learning_rate": 2.261538461538461e-05,
+      "loss": 0.0012,
+      "step": 705
+    },
+    {
+      "epoch": 141.19,
+      "learning_rate": 2.2153846153846154e-05,
+      "loss": 0.0014,
+      "step": 706
+    },
+    {
+      "epoch": 141.38,
+      "learning_rate": 2.169230769230769e-05,
+      "loss": 0.0012,
+      "step": 707
+    },
+    {
+      "epoch": 141.57,
+      "learning_rate": 2.123076923076923e-05,
+      "loss": 0.0012,
+      "step": 708
+    },
+    {
+      "epoch": 141.76,
+      "learning_rate": 2.076923076923077e-05,
+      "loss": 0.0012,
+      "step": 709
+    },
+    {
+      "epoch": 141.95,
+      "learning_rate": 2.0307692307692304e-05,
+      "loss": 0.0012,
+      "step": 710
+    },
+    {
+      "epoch": 142.19,
+      "learning_rate": 1.9846153846153847e-05,
+      "loss": 0.2135,
+      "step": 711
+    },
+    {
+      "epoch": 142.38,
+      "learning_rate": 1.9384615384615383e-05,
+      "loss": 0.0012,
+      "step": 712
+    },
+    {
+      "epoch": 142.57,
+      "learning_rate": 1.892307692307692e-05,
+      "loss": 0.0014,
+      "step": 713
+    },
+    {
+      "epoch": 142.76,
+      "learning_rate": 1.846153846153846e-05,
+      "loss": 0.0012,
+      "step": 714
+    },
+    {
+      "epoch": 142.95,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 0.0012,
+      "step": 715
+    },
+    {
+      "epoch": 143.19,
+      "learning_rate": 1.7538461538461536e-05,
+      "loss": 0.0014,
+      "step": 716
+    },
+    {
+      "epoch": 143.38,
+      "learning_rate": 1.7076923076923076e-05,
+      "loss": 0.0011,
+      "step": 717
+    },
+    {
+      "epoch": 143.57,
+      "learning_rate": 1.6615384615384615e-05,
+      "loss": 0.0012,
+      "step": 718
+    },
+    {
+      "epoch": 143.76,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 0.0012,
+      "step": 719
+    },
+    {
+      "epoch": 143.95,
+      "learning_rate": 1.569230769230769e-05,
+      "loss": 0.0015,
+      "step": 720
+    },
+    {
+      "epoch": 143.95,
+      "eval_accuracy": 0.9113475177304965,
+      "eval_f1": 0.9143793727695214,
+      "eval_loss": 0.5361447334289551,
+      "eval_runtime": 16.0749,
+      "eval_samples_per_second": 17.543,
+      "eval_steps_per_second": 1.12,
+      "step": 720
+    },
+    {
+      "epoch": 144.19,
+      "learning_rate": 1.5230769230769229e-05,
+      "loss": 0.0016,
+      "step": 721
+    },
+    {
+      "epoch": 144.38,
+      "learning_rate": 1.4769230769230768e-05,
+      "loss": 0.0012,
+      "step": 722
+    },
+    {
+      "epoch": 144.57,
+      "learning_rate": 1.4307692307692308e-05,
+      "loss": 0.0011,
+      "step": 723
+    },
+    {
+      "epoch": 144.76,
+      "learning_rate": 1.3846153846153845e-05,
+      "loss": 0.0012,
+      "step": 724
+    },
+    {
+      "epoch": 144.95,
+      "learning_rate": 1.3384615384615383e-05,
+      "loss": 0.0012,
+      "step": 725
+    },
+    {
+      "epoch": 145.19,
+      "learning_rate": 1.2923076923076922e-05,
+      "loss": 0.0015,
+      "step": 726
+    },
+    {
+      "epoch": 145.38,
+      "learning_rate": 1.2461538461538461e-05,
+      "loss": 0.0012,
+      "step": 727
+    },
+    {
+      "epoch": 145.57,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 0.0013,
+      "step": 728
+    },
+    {
+      "epoch": 145.76,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 0.0011,
+      "step": 729
+    },
+    {
+      "epoch": 145.95,
+      "learning_rate": 1.1076923076923077e-05,
+      "loss": 0.0047,
+      "step": 730
+    },
+    {
+      "epoch": 146.19,
+      "learning_rate": 1.0615384615384615e-05,
+      "loss": 0.0016,
+      "step": 731
+    },
+    {
+      "epoch": 146.38,
+      "learning_rate": 1.0153846153846152e-05,
+      "loss": 0.0012,
+      "step": 732
+    },
+    {
+      "epoch": 146.57,
+      "learning_rate": 9.692307692307691e-06,
+      "loss": 0.0011,
+      "step": 733
+    },
+    {
+      "epoch": 146.76,
+      "learning_rate": 9.23076923076923e-06,
+      "loss": 0.0012,
+      "step": 734
+    },
+    {
+      "epoch": 146.95,
+      "learning_rate": 8.769230769230768e-06,
+      "loss": 0.0012,
+      "step": 735
+    },
+    {
+      "epoch": 147.19,
+      "learning_rate": 8.307692307692307e-06,
+      "loss": 0.0015,
+      "step": 736
+    },
+    {
+      "epoch": 147.38,
+      "learning_rate": 7.846153846153845e-06,
+      "loss": 0.0013,
+      "step": 737
+    },
+    {
+      "epoch": 147.57,
+      "learning_rate": 7.384615384615384e-06,
+      "loss": 0.0012,
+      "step": 738
+    },
+    {
+      "epoch": 147.76,
+      "learning_rate": 6.9230769230769225e-06,
+      "loss": 0.0012,
+      "step": 739
+    },
+    {
+      "epoch": 147.95,
+      "learning_rate": 6.461538461538461e-06,
+      "loss": 0.0012,
+      "step": 740
+    },
+    {
+      "epoch": 147.95,
+      "eval_accuracy": 0.9113475177304965,
+      "eval_f1": 0.9143793727695214,
+      "eval_loss": 0.5337392091751099,
+      "eval_runtime": 16.1902,
+      "eval_samples_per_second": 17.418,
+      "eval_steps_per_second": 1.112,
+      "step": 740
+    },
+    {
+      "epoch": 148.19,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 0.0016,
+      "step": 741
+    },
+    {
+      "epoch": 148.38,
+      "learning_rate": 5.5384615384615385e-06,
+      "loss": 0.0012,
+      "step": 742
+    },
+    {
+      "epoch": 148.57,
+      "learning_rate": 5.076923076923076e-06,
+      "loss": 0.0011,
+      "step": 743
+    },
+    {
+      "epoch": 148.76,
+      "learning_rate": 4.615384615384615e-06,
+      "loss": 0.0011,
+      "step": 744
+    },
+    {
+      "epoch": 148.95,
+      "learning_rate": 4.153846153846154e-06,
+      "loss": 0.0012,
+      "step": 745
+    },
+    {
+      "epoch": 149.19,
+      "learning_rate": 3.692307692307692e-06,
+      "loss": 0.0016,
+      "step": 746
+    },
+    {
+      "epoch": 149.38,
+      "learning_rate": 3.2307692307692305e-06,
+      "loss": 0.0012,
+      "step": 747
+    },
+    {
+      "epoch": 149.57,
+      "learning_rate": 2.7692307692307693e-06,
+      "loss": 0.0013,
+      "step": 748
+    },
+    {
+      "epoch": 149.76,
+      "learning_rate": 2.3076923076923077e-06,
+      "loss": 0.0011,
+      "step": 749
+    },
+    {
+      "epoch": 149.95,
+      "learning_rate": 1.846153846153846e-06,
+      "loss": 0.0011,
+      "step": 750
+    },
+    {
+      "epoch": 149.95,
+      "step": 750,
+      "total_flos": 6.656769751940858e+18,
+      "train_loss": 0.38806585280100503,
+      "train_runtime": 2433.2545,
+      "train_samples_per_second": 10.295,
+      "train_steps_per_second": 0.308
+    }
+  ],
+  "max_steps": 750,
+  "num_train_epochs": 150,
+  "total_flos": 6.656769751940858e+18,
+  "trial_name": null,
+  "trial_params": null
+}