diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6980 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "global_step": 1150,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9982608695652174e-05,
+      "loss": 0.5753,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.996521739130435e-05,
+      "loss": 0.7514,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9947826086956524e-05,
+      "loss": 0.5382,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9930434782608696e-05,
+      "loss": 0.7397,
+      "step": 4
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9913043478260872e-05,
+      "loss": 0.6228,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9895652173913044e-05,
+      "loss": 0.5894,
+      "step": 6
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.987826086956522e-05,
+      "loss": 0.7549,
+      "step": 7
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9860869565217395e-05,
+      "loss": 0.4033,
+      "step": 8
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9843478260869567e-05,
+      "loss": 0.5795,
+      "step": 9
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9826086956521742e-05,
+      "loss": 0.6001,
+      "step": 10
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9808695652173914e-05,
+      "loss": 0.7615,
+      "step": 11
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.979130434782609e-05,
+      "loss": 0.551,
+      "step": 12
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9773913043478265e-05,
+      "loss": 0.7247,
+      "step": 13
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9756521739130437e-05,
+      "loss": 0.6531,
+      "step": 14
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.973913043478261e-05,
+      "loss": 0.4637,
+      "step": 15
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9721739130434784e-05,
+      "loss": 0.5278,
+      "step": 16
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9704347826086956e-05,
+      "loss": 0.5302,
+      "step": 17
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9686956521739132e-05,
+      "loss": 0.7791,
+      "step": 18
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9669565217391307e-05,
+      "loss": 0.6974,
+      "step": 19
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.965217391304348e-05,
+      "loss": 0.5094,
+      "step": 20
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.963478260869565e-05,
+      "loss": 0.633,
+      "step": 21
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9617391304347827e-05,
+      "loss": 0.5473,
+      "step": 22
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.5769,
+      "step": 23
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9582608695652177e-05,
+      "loss": 0.5745,
+      "step": 24
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.5602,
+      "step": 25
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.954782608695652e-05,
+      "loss": 0.7122,
+      "step": 26
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9530434782608697e-05,
+      "loss": 0.7701,
+      "step": 27
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9513043478260872e-05,
+      "loss": 0.5944,
+      "step": 28
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9495652173913044e-05,
+      "loss": 0.7585,
+      "step": 29
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.947826086956522e-05,
+      "loss": 0.4709,
+      "step": 30
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9460869565217392e-05,
+      "loss": 0.5949,
+      "step": 31
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9443478260869567e-05,
+      "loss": 0.5356,
+      "step": 32
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9426086956521743e-05,
+      "loss": 0.4729,
+      "step": 33
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9408695652173915e-05,
+      "loss": 0.4826,
+      "step": 34
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9391304347826087e-05,
+      "loss": 0.5672,
+      "step": 35
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9373913043478262e-05,
+      "loss": 0.7598,
+      "step": 36
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9356521739130437e-05,
+      "loss": 0.6398,
+      "step": 37
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.933913043478261e-05,
+      "loss": 0.6299,
+      "step": 38
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9321739130434785e-05,
+      "loss": 0.6392,
+      "step": 39
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9304347826086957e-05,
+      "loss": 0.7103,
+      "step": 40
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9286956521739132e-05,
+      "loss": 0.5011,
+      "step": 41
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9269565217391308e-05,
+      "loss": 0.3659,
+      "step": 42
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.925217391304348e-05,
+      "loss": 0.4543,
+      "step": 43
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9234782608695655e-05,
+      "loss": 0.6743,
+      "step": 44
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9217391304347827e-05,
+      "loss": 0.5207,
+      "step": 45
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.6568,
+      "step": 46
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9182608695652175e-05,
+      "loss": 0.5027,
+      "step": 47
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.916521739130435e-05,
+      "loss": 0.5637,
+      "step": 48
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9147826086956522e-05,
+      "loss": 0.4606,
+      "step": 49
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 0.6994,
+      "step": 50
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.911304347826087e-05,
+      "loss": 0.5158,
+      "step": 51
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9095652173913045e-05,
+      "loss": 0.6494,
+      "step": 52
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.907826086956522e-05,
+      "loss": 0.6689,
+      "step": 53
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9060869565217392e-05,
+      "loss": 0.3677,
+      "step": 54
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9043478260869568e-05,
+      "loss": 0.7658,
+      "step": 55
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.902608695652174e-05,
+      "loss": 0.4006,
+      "step": 56
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9008695652173915e-05,
+      "loss": 0.301,
+      "step": 57
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.899130434782609e-05,
+      "loss": 0.5438,
+      "step": 58
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8973913043478262e-05,
+      "loss": 0.5882,
+      "step": 59
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8956521739130434e-05,
+      "loss": 0.5507,
+      "step": 60
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.893913043478261e-05,
+      "loss": 0.8205,
+      "step": 61
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8921739130434785e-05,
+      "loss": 0.5634,
+      "step": 62
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8904347826086957e-05,
+      "loss": 0.427,
+      "step": 63
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8886956521739133e-05,
+      "loss": 0.6946,
+      "step": 64
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8869565217391305e-05,
+      "loss": 0.4922,
+      "step": 65
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.885217391304348e-05,
+      "loss": 0.5063,
+      "step": 66
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.8834782608695656e-05,
+      "loss": 0.6757,
+      "step": 67
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8817391304347828e-05,
+      "loss": 0.5164,
+      "step": 68
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.88e-05,
+      "loss": 0.5827,
+      "step": 69
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.8782608695652175e-05,
+      "loss": 0.6119,
+      "step": 70
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.876521739130435e-05,
+      "loss": 0.6298,
+      "step": 71
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.8747826086956526e-05,
+      "loss": 0.4407,
+      "step": 72
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.8730434782608698e-05,
+      "loss": 0.4486,
+      "step": 73
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.871304347826087e-05,
+      "loss": 0.4184,
+      "step": 74
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8695652173913045e-05,
+      "loss": 0.6509,
+      "step": 75
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.867826086956522e-05,
+      "loss": 0.2587,
+      "step": 76
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8660869565217393e-05,
+      "loss": 0.5633,
+      "step": 77
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.8643478260869568e-05,
+      "loss": 0.5623,
+      "step": 78
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.862608695652174e-05,
+      "loss": 0.689,
+      "step": 79
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8608695652173912e-05,
+      "loss": 0.5268,
+      "step": 80
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8591304347826087e-05,
+      "loss": 0.4938,
+      "step": 81
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.8573913043478263e-05,
+      "loss": 0.7229,
+      "step": 82
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.8556521739130435e-05,
+      "loss": 0.439,
+      "step": 83
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.853913043478261e-05,
+      "loss": 0.5143,
+      "step": 84
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8521739130434782e-05,
+      "loss": 0.6849,
+      "step": 85
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8504347826086958e-05,
+      "loss": 0.5098,
+      "step": 86
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8486956521739133e-05,
+      "loss": 0.6135,
+      "step": 87
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8469565217391305e-05,
+      "loss": 0.6127,
+      "step": 88
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.845217391304348e-05,
+      "loss": 0.5749,
+      "step": 89
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.8434782608695653e-05,
+      "loss": 0.5202,
+      "step": 90
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8417391304347828e-05,
+      "loss": 0.5382,
+      "step": 91
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.569,
+      "step": 92
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8382608695652175e-05,
+      "loss": 0.5202,
+      "step": 93
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8365217391304347e-05,
+      "loss": 0.2982,
+      "step": 94
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8347826086956523e-05,
+      "loss": 0.6126,
+      "step": 95
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.8330434782608698e-05,
+      "loss": 0.453,
+      "step": 96
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.831304347826087e-05,
+      "loss": 0.5852,
+      "step": 97
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.8295652173913046e-05,
+      "loss": 0.8428,
+      "step": 98
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.8278260869565218e-05,
+      "loss": 0.4392,
+      "step": 99
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 0.4946,
+      "step": 100
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.824347826086957e-05,
+      "loss": 0.795,
+      "step": 101
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.822608695652174e-05,
+      "loss": 0.3557,
+      "step": 102
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.8208695652173916e-05,
+      "loss": 0.6206,
+      "step": 103
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.8191304347826088e-05,
+      "loss": 0.489,
+      "step": 104
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.8173913043478263e-05,
+      "loss": 0.5684,
+      "step": 105
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.815652173913044e-05,
+      "loss": 0.5602,
+      "step": 106
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.813913043478261e-05,
+      "loss": 0.5284,
+      "step": 107
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.8121739130434783e-05,
+      "loss": 0.4759,
+      "step": 108
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.8104347826086958e-05,
+      "loss": 0.4353,
+      "step": 109
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.808695652173913e-05,
+      "loss": 0.6774,
+      "step": 110
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.8069565217391306e-05,
+      "loss": 0.7922,
+      "step": 111
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.805217391304348e-05,
+      "loss": 0.5102,
+      "step": 112
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.8034782608695653e-05,
+      "loss": 0.4687,
+      "step": 113
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8017391304347825e-05,
+      "loss": 0.7305,
+      "step": 114
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8e-05,
+      "loss": 0.4995,
+      "step": 115
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.7982608695652176e-05,
+      "loss": 0.4522,
+      "step": 116
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.796521739130435e-05,
+      "loss": 0.5282,
+      "step": 117
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7947826086956523e-05,
+      "loss": 0.7347,
+      "step": 118
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.7930434782608695e-05,
+      "loss": 0.5023,
+      "step": 119
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.791304347826087e-05,
+      "loss": 0.3845,
+      "step": 120
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7895652173913046e-05,
+      "loss": 0.5378,
+      "step": 121
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7878260869565218e-05,
+      "loss": 0.3955,
+      "step": 122
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7860869565217394e-05,
+      "loss": 0.3116,
+      "step": 123
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.7843478260869566e-05,
+      "loss": 0.4914,
+      "step": 124
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.782608695652174e-05,
+      "loss": 0.5629,
+      "step": 125
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.7808695652173916e-05,
+      "loss": 0.2251,
+      "step": 126
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.779130434782609e-05,
+      "loss": 0.3285,
+      "step": 127
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.777391304347826e-05,
+      "loss": 0.3974,
+      "step": 128
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.7756521739130436e-05,
+      "loss": 0.2562,
+      "step": 129
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.773913043478261e-05,
+      "loss": 0.2804,
+      "step": 130
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.7721739130434783e-05,
+      "loss": 0.5984,
+      "step": 131
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.770434782608696e-05,
+      "loss": 0.8529,
+      "step": 132
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.768695652173913e-05,
+      "loss": 0.6385,
+      "step": 133
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7669565217391306e-05,
+      "loss": 0.3558,
+      "step": 134
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.765217391304348e-05,
+      "loss": 0.3149,
+      "step": 135
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7634782608695653e-05,
+      "loss": 0.3815,
+      "step": 136
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.761739130434783e-05,
+      "loss": 0.4676,
+      "step": 137
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.76e-05,
+      "loss": 0.6148,
+      "step": 138
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7582608695652176e-05,
+      "loss": 0.3845,
+      "step": 139
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.756521739130435e-05,
+      "loss": 0.5324,
+      "step": 140
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7547826086956524e-05,
+      "loss": 0.4193,
+      "step": 141
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7530434782608696e-05,
+      "loss": 0.6118,
+      "step": 142
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.751304347826087e-05,
+      "loss": 0.2856,
+      "step": 143
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7495652173913043e-05,
+      "loss": 0.3601,
+      "step": 144
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.747826086956522e-05,
+      "loss": 0.3205,
+      "step": 145
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7460869565217394e-05,
+      "loss": 0.3526,
+      "step": 146
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7443478260869566e-05,
+      "loss": 0.5599,
+      "step": 147
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.742608695652174e-05,
+      "loss": 0.4693,
+      "step": 148
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7408695652173913e-05,
+      "loss": 0.3326,
+      "step": 149
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.6522,
+      "step": 150
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7373913043478264e-05,
+      "loss": 0.4743,
+      "step": 151
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7356521739130436e-05,
+      "loss": 0.781,
+      "step": 152
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7339130434782608e-05,
+      "loss": 0.2015,
+      "step": 153
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7321739130434784e-05,
+      "loss": 0.1317,
+      "step": 154
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.730434782608696e-05,
+      "loss": 0.5058,
+      "step": 155
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.728695652173913e-05,
+      "loss": 0.3922,
+      "step": 156
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7269565217391306e-05,
+      "loss": 0.2571,
+      "step": 157
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.725217391304348e-05,
+      "loss": 0.1081,
+      "step": 158
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7234782608695654e-05,
+      "loss": 0.6332,
+      "step": 159
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.721739130434783e-05,
+      "loss": 0.2597,
+      "step": 160
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.72e-05,
+      "loss": 0.7022,
+      "step": 161
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7182608695652173e-05,
+      "loss": 0.4827,
+      "step": 162
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.716521739130435e-05,
+      "loss": 0.2314,
+      "step": 163
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7147826086956524e-05,
+      "loss": 0.5084,
+      "step": 164
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.71304347826087e-05,
+      "loss": 0.3332,
+      "step": 165
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.711304347826087e-05,
+      "loss": 0.4277,
+      "step": 166
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7095652173913044e-05,
+      "loss": 0.2968,
+      "step": 167
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.707826086956522e-05,
+      "loss": 0.4463,
+      "step": 168
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7060869565217394e-05,
+      "loss": 0.5358,
+      "step": 169
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7043478260869566e-05,
+      "loss": 0.5242,
+      "step": 170
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7026086956521742e-05,
+      "loss": 0.4157,
+      "step": 171
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7008695652173914e-05,
+      "loss": 0.4456,
+      "step": 172
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.6991304347826086e-05,
+      "loss": 0.3063,
+      "step": 173
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.697391304347826e-05,
+      "loss": 0.3954,
+      "step": 174
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.6956521739130437e-05,
+      "loss": 0.2935,
+      "step": 175
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.693913043478261e-05,
+      "loss": 0.4669,
+      "step": 176
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.6921739130434784e-05,
+      "loss": 0.482,
+      "step": 177
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.6904347826086956e-05,
+      "loss": 0.61,
+      "step": 178
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.688695652173913e-05,
+      "loss": 0.5151,
+      "step": 179
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.6869565217391307e-05,
+      "loss": 0.3137,
+      "step": 180
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.685217391304348e-05,
+      "loss": 0.2869,
+      "step": 181
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.6834782608695654e-05,
+      "loss": 0.2031,
+      "step": 182
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.6817391304347826e-05,
+      "loss": 0.271,
+      "step": 183
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.4179,
+      "step": 184
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.6782608695652177e-05,
+      "loss": 0.4197,
+      "step": 185
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.676521739130435e-05,
+      "loss": 0.362,
+      "step": 186
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.674782608695652e-05,
+      "loss": 0.4576,
+      "step": 187
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6730434782608697e-05,
+      "loss": 0.3602,
+      "step": 188
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6713043478260872e-05,
+      "loss": 0.8584,
+      "step": 189
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6695652173913044e-05,
+      "loss": 0.3587,
+      "step": 190
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.667826086956522e-05,
+      "loss": 0.443,
+      "step": 191
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.666086956521739e-05,
+      "loss": 0.6045,
+      "step": 192
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6643478260869567e-05,
+      "loss": 0.6547,
+      "step": 193
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6626086956521742e-05,
+      "loss": 0.4472,
+      "step": 194
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6608695652173914e-05,
+      "loss": 0.3199,
+      "step": 195
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.659130434782609e-05,
+      "loss": 0.4287,
+      "step": 196
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6573913043478262e-05,
+      "loss": 0.2837,
+      "step": 197
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6556521739130437e-05,
+      "loss": 0.2362,
+      "step": 198
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6539130434782613e-05,
+      "loss": 0.4401,
+      "step": 199
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6521739130434785e-05,
+      "loss": 0.5061,
+      "step": 200
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6504347826086957e-05,
+      "loss": 0.3449,
+      "step": 201
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6486956521739132e-05,
+      "loss": 0.1536,
+      "step": 202
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6469565217391304e-05,
+      "loss": 0.4143,
+      "step": 203
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.645217391304348e-05,
+      "loss": 0.2272,
+      "step": 204
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6434782608695655e-05,
+      "loss": 0.4418,
+      "step": 205
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6417391304347827e-05,
+      "loss": 0.3822,
+      "step": 206
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.64e-05,
+      "loss": 0.2973,
+      "step": 207
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6382608695652174e-05,
+      "loss": 0.2386,
+      "step": 208
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.636521739130435e-05,
+      "loss": 0.262,
+      "step": 209
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6347826086956525e-05,
+      "loss": 0.6586,
+      "step": 210
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6330434782608697e-05,
+      "loss": 0.2293,
+      "step": 211
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.631304347826087e-05,
+      "loss": 0.245,
+      "step": 212
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6295652173913044e-05,
+      "loss": 0.4197,
+      "step": 213
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.627826086956522e-05,
+      "loss": 0.3156,
+      "step": 214
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6260869565217392e-05,
+      "loss": 0.3987,
+      "step": 215
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6243478260869567e-05,
+      "loss": 0.7833,
+      "step": 216
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.622608695652174e-05,
+      "loss": 0.403,
+      "step": 217
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6208695652173915e-05,
+      "loss": 0.4266,
+      "step": 218
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.619130434782609e-05,
+      "loss": 0.6049,
+      "step": 219
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.6173913043478262e-05,
+      "loss": 0.3202,
+      "step": 220
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.6156521739130434e-05,
+      "loss": 0.266,
+      "step": 221
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.613913043478261e-05,
+      "loss": 0.7602,
+      "step": 222
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6121739130434785e-05,
+      "loss": 0.5766,
+      "step": 223
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6104347826086957e-05,
+      "loss": 0.4624,
+      "step": 224
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.6086956521739132e-05,
+      "loss": 0.2977,
+      "step": 225
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.6069565217391304e-05,
+      "loss": 0.3935,
+      "step": 226
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.605217391304348e-05,
+      "loss": 0.7221,
+      "step": 227
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.6034782608695655e-05,
+      "loss": 0.416,
+      "step": 228
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6017391304347827e-05,
+      "loss": 0.5559,
+      "step": 229
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.1212,
+      "step": 230
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8799019607843137,
+      "eval_combined_score": 0.896741103848947,
+      "eval_f1": 0.9135802469135803,
+      "eval_loss": 0.3400943875312805,
+      "eval_runtime": 0.838,
+      "eval_samples_per_second": 486.848,
+      "eval_steps_per_second": 31.025,
+      "step": 230
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5982608695652175e-05,
+      "loss": 0.2575,
+      "step": 231
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.596521739130435e-05,
+      "loss": 0.2924,
+      "step": 232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5947826086956522e-05,
+      "loss": 0.1243,
+      "step": 233
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5930434782608698e-05,
+      "loss": 0.4676,
+      "step": 234
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.591304347826087e-05,
+      "loss": 0.4703,
+      "step": 235
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5895652173913045e-05,
+      "loss": 0.1455,
+      "step": 236
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5878260869565217e-05,
+      "loss": 0.2733,
+      "step": 237
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5860869565217392e-05,
+      "loss": 0.2674,
+      "step": 238
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5843478260869568e-05,
+      "loss": 0.2892,
+      "step": 239
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.582608695652174e-05,
+      "loss": 0.2632,
+      "step": 240
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5808695652173915e-05,
+      "loss": 0.2075,
+      "step": 241
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5791304347826087e-05,
+      "loss": 0.2362,
+      "step": 242
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5773913043478263e-05,
+      "loss": 0.4886,
+      "step": 243
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5756521739130438e-05,
+      "loss": 0.3003,
+      "step": 244
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.573913043478261e-05,
+      "loss": 0.4119,
+      "step": 245
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5721739130434782e-05,
+      "loss": 0.5518,
+      "step": 246
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5704347826086957e-05,
+      "loss": 0.0634,
+      "step": 247
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5686956521739133e-05,
+      "loss": 0.5598,
+      "step": 248
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5669565217391305e-05,
+      "loss": 0.2881,
+      "step": 249
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 0.4026,
+      "step": 250
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.5634782608695652e-05,
+      "loss": 0.7736,
+      "step": 251
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.5617391304347828e-05,
+      "loss": 0.678,
+      "step": 252
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.2728,
+      "step": 253
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.5582608695652175e-05,
+      "loss": 0.3382,
+      "step": 254
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.5565217391304347e-05,
+      "loss": 0.5084,
+      "step": 255
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.5547826086956523e-05,
+      "loss": 0.4209,
+      "step": 256
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.5530434782608698e-05,
+      "loss": 0.2575,
+      "step": 257
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.5513043478260873e-05,
+      "loss": 0.5152,
+      "step": 258
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.5495652173913045e-05,
+      "loss": 0.878,
+      "step": 259
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.5478260869565217e-05,
+      "loss": 0.391,
+      "step": 260
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.5460869565217393e-05,
+      "loss": 0.3526,
+      "step": 261
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.5443478260869568e-05,
+      "loss": 0.2194,
+      "step": 262
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.542608695652174e-05,
+      "loss": 0.2898,
+      "step": 263
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.5408695652173916e-05,
+      "loss": 0.0896,
+      "step": 264
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.5391304347826088e-05,
+      "loss": 0.9961,
+      "step": 265
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.537391304347826e-05,
+      "loss": 0.5364,
+      "step": 266
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.5356521739130435e-05,
+      "loss": 0.4723,
+      "step": 267
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.533913043478261e-05,
+      "loss": 0.4401,
+      "step": 268
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.5321739130434782e-05,
+      "loss": 0.8362,
+      "step": 269
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.5304347826086958e-05,
+      "loss": 0.6872,
+      "step": 270
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.528695652173913e-05,
+      "loss": 0.8144,
+      "step": 271
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.5269565217391305e-05,
+      "loss": 0.2772,
+      "step": 272
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.5252173913043479e-05,
+      "loss": 0.5225,
+      "step": 273
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.5234782608695653e-05,
+      "loss": 0.5575,
+      "step": 274
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.2228,
+      "step": 275
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.2939,
+      "step": 276
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.5182608695652174e-05,
+      "loss": 0.2553,
+      "step": 277
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.516521739130435e-05,
+      "loss": 0.385,
+      "step": 278
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.5147826086956523e-05,
+      "loss": 0.3448,
+      "step": 279
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.5130434782608697e-05,
+      "loss": 0.341,
+      "step": 280
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.5113043478260872e-05,
+      "loss": 0.2939,
+      "step": 281
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.5095652173913044e-05,
+      "loss": 0.3279,
+      "step": 282
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.5078260869565218e-05,
+      "loss": 0.3804,
+      "step": 283
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.5060869565217393e-05,
+      "loss": 0.3251,
+      "step": 284
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.5043478260869567e-05,
+      "loss": 0.4146,
+      "step": 285
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.5026086956521739e-05,
+      "loss": 0.2304,
+      "step": 286
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.5008695652173914e-05,
+      "loss": 0.5483,
+      "step": 287
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.4991304347826088e-05,
+      "loss": 0.3224,
+      "step": 288
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4973913043478263e-05,
+      "loss": 0.1592,
+      "step": 289
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4956521739130436e-05,
+      "loss": 0.1218,
+      "step": 290
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.493913043478261e-05,
+      "loss": 0.4082,
+      "step": 291
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4921739130434785e-05,
+      "loss": 0.4803,
+      "step": 292
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4904347826086958e-05,
+      "loss": 0.3713,
+      "step": 293
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.488695652173913e-05,
+      "loss": 0.1216,
+      "step": 294
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.4869565217391306e-05,
+      "loss": 0.3276,
+      "step": 295
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.485217391304348e-05,
+      "loss": 0.149,
+      "step": 296
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.4834782608695653e-05,
+      "loss": 0.2914,
+      "step": 297
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4817391304347829e-05,
+      "loss": 0.2604,
+      "step": 298
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.48e-05,
+      "loss": 0.2767,
+      "step": 299
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 0.5331,
+      "step": 300
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.476521739130435e-05,
+      "loss": 0.1141,
+      "step": 301
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.4747826086956523e-05,
+      "loss": 0.1176,
+      "step": 302
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.4730434782608695e-05,
+      "loss": 0.5076,
+      "step": 303
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.4713043478260871e-05,
+      "loss": 0.2047,
+      "step": 304
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.4695652173913045e-05,
+      "loss": 0.6033,
+      "step": 305
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.467826086956522e-05,
+      "loss": 0.2011,
+      "step": 306
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.4660869565217392e-05,
+      "loss": 0.1223,
+      "step": 307
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4643478260869566e-05,
+      "loss": 0.0917,
+      "step": 308
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.4626086956521741e-05,
+      "loss": 0.1207,
+      "step": 309
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4608695652173915e-05,
+      "loss": 0.2436,
+      "step": 310
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.4591304347826087e-05,
+      "loss": 0.5129,
+      "step": 311
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4573913043478262e-05,
+      "loss": 0.1075,
+      "step": 312
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.4556521739130436e-05,
+      "loss": 0.3466,
+      "step": 313
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.453913043478261e-05,
+      "loss": 0.7119,
+      "step": 314
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.4521739130434785e-05,
+      "loss": 0.1999,
+      "step": 315
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.4504347826086957e-05,
+      "loss": 0.3681,
+      "step": 316
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.448695652173913e-05,
+      "loss": 0.4795,
+      "step": 317
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.4469565217391306e-05,
+      "loss": 0.2033,
+      "step": 318
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.4452173913043478e-05,
+      "loss": 0.1562,
+      "step": 319
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.4434782608695654e-05,
+      "loss": 0.5578,
+      "step": 320
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.4417391304347827e-05,
+      "loss": 0.275,
+      "step": 321
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.4992,
+      "step": 322
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.4382608695652176e-05,
+      "loss": 0.4364,
+      "step": 323
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.4365217391304348e-05,
+      "loss": 0.2421,
+      "step": 324
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.4347826086956522e-05,
+      "loss": 0.3475,
+      "step": 325
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.4330434782608698e-05,
+      "loss": 0.1773,
+      "step": 326
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.4313043478260871e-05,
+      "loss": 0.144,
+      "step": 327
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.4295652173913043e-05,
+      "loss": 0.5584,
+      "step": 328
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.4278260869565219e-05,
+      "loss": 0.0916,
+      "step": 329
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.4260869565217392e-05,
+      "loss": 0.5589,
+      "step": 330
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.4243478260869566e-05,
+      "loss": 0.4504,
+      "step": 331
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.4226086956521742e-05,
+      "loss": 0.4808,
+      "step": 332
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.4208695652173914e-05,
+      "loss": 0.4521,
+      "step": 333
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.4191304347826087e-05,
+      "loss": 0.55,
+      "step": 334
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.4173913043478263e-05,
+      "loss": 0.7677,
+      "step": 335
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.4156521739130435e-05,
+      "loss": 0.3933,
+      "step": 336
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.413913043478261e-05,
+      "loss": 0.2556,
+      "step": 337
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.4121739130434784e-05,
+      "loss": 0.2292,
+      "step": 338
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.4104347826086958e-05,
+      "loss": 0.4732,
+      "step": 339
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.4086956521739133e-05,
+      "loss": 0.1577,
+      "step": 340
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.4069565217391305e-05,
+      "loss": 0.3132,
+      "step": 341
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.4052173913043479e-05,
+      "loss": 0.2144,
+      "step": 342
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.4034782608695654e-05,
+      "loss": 0.2625,
+      "step": 343
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.4017391304347828e-05,
+      "loss": 0.3131,
+      "step": 344
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.4e-05,
+      "loss": 0.313,
+      "step": 345
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3982608695652175e-05,
+      "loss": 0.262,
+      "step": 346
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3965217391304349e-05,
+      "loss": 0.5384,
+      "step": 347
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3947826086956523e-05,
+      "loss": 0.1984,
+      "step": 348
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3930434782608698e-05,
+      "loss": 0.5003,
+      "step": 349
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 0.1027,
+      "step": 350
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3895652173913045e-05,
+      "loss": 0.3822,
+      "step": 351
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.387826086956522e-05,
+      "loss": 0.2108,
+      "step": 352
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3860869565217391e-05,
+      "loss": 0.114,
+      "step": 353
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.3843478260869567e-05,
+      "loss": 0.1998,
+      "step": 354
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.382608695652174e-05,
+      "loss": 0.3795,
+      "step": 355
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3808695652173914e-05,
+      "loss": 0.2995,
+      "step": 356
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.379130434782609e-05,
+      "loss": 0.5332,
+      "step": 357
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3773913043478261e-05,
+      "loss": 0.6547,
+      "step": 358
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3756521739130435e-05,
+      "loss": 0.4123,
+      "step": 359
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.373913043478261e-05,
+      "loss": 0.1137,
+      "step": 360
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3721739130434784e-05,
+      "loss": 0.0974,
+      "step": 361
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3704347826086956e-05,
+      "loss": 0.2198,
+      "step": 362
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3686956521739132e-05,
+      "loss": 0.2858,
+      "step": 363
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3669565217391305e-05,
+      "loss": 0.1757,
+      "step": 364
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3652173913043479e-05,
+      "loss": 0.2423,
+      "step": 365
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3634782608695653e-05,
+      "loss": 0.5822,
+      "step": 366
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3617391304347827e-05,
+      "loss": 0.0828,
+      "step": 367
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.3966,
+      "step": 368
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3582608695652176e-05,
+      "loss": 0.2143,
+      "step": 369
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3565217391304348e-05,
+      "loss": 0.1393,
+      "step": 370
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3547826086956523e-05,
+      "loss": 0.4234,
+      "step": 371
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3530434782608697e-05,
+      "loss": 0.3362,
+      "step": 372
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.351304347826087e-05,
+      "loss": 0.1123,
+      "step": 373
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3495652173913046e-05,
+      "loss": 0.3911,
+      "step": 374
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3478260869565218e-05,
+      "loss": 0.0732,
+      "step": 375
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3460869565217392e-05,
+      "loss": 0.1994,
+      "step": 376
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3443478260869567e-05,
+      "loss": 0.7325,
+      "step": 377
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.342608695652174e-05,
+      "loss": 0.7171,
+      "step": 378
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3408695652173913e-05,
+      "loss": 0.2008,
+      "step": 379
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3391304347826088e-05,
+      "loss": 0.4034,
+      "step": 380
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3373913043478262e-05,
+      "loss": 0.4566,
+      "step": 381
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3356521739130437e-05,
+      "loss": 0.2695,
+      "step": 382
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.333913043478261e-05,
+      "loss": 0.2316,
+      "step": 383
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3321739130434783e-05,
+      "loss": 0.5321,
+      "step": 384
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3304347826086958e-05,
+      "loss": 0.1439,
+      "step": 385
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3286956521739132e-05,
+      "loss": 0.5678,
+      "step": 386
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3269565217391304e-05,
+      "loss": 0.1575,
+      "step": 387
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.325217391304348e-05,
+      "loss": 0.4681,
+      "step": 388
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3234782608695653e-05,
+      "loss": 0.1743,
+      "step": 389
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3217391304347827e-05,
+      "loss": 0.2348,
+      "step": 390
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.2463,
+      "step": 391
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3182608695652174e-05,
+      "loss": 0.1224,
+      "step": 392
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.3165217391304348e-05,
+      "loss": 0.2687,
+      "step": 393
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.3147826086956524e-05,
+      "loss": 0.3428,
+      "step": 394
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3130434782608697e-05,
+      "loss": 0.3908,
+      "step": 395
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.311304347826087e-05,
+      "loss": 0.6197,
+      "step": 396
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.3095652173913045e-05,
+      "loss": 0.492,
+      "step": 397
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.3078260869565218e-05,
+      "loss": 0.3314,
+      "step": 398
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.3060869565217394e-05,
+      "loss": 0.2878,
+      "step": 399
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.2051,
+      "step": 400
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.302608695652174e-05,
+      "loss": 0.1469,
+      "step": 401
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.3008695652173915e-05,
+      "loss": 0.1172,
+      "step": 402
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.2991304347826089e-05,
+      "loss": 0.2836,
+      "step": 403
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.297391304347826e-05,
+      "loss": 0.1137,
+      "step": 404
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2956521739130436e-05,
+      "loss": 0.1234,
+      "step": 405
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.293913043478261e-05,
+      "loss": 0.1156,
+      "step": 406
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2921739130434783e-05,
+      "loss": 0.2665,
+      "step": 407
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2904347826086959e-05,
+      "loss": 0.4269,
+      "step": 408
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.2886956521739131e-05,
+      "loss": 0.2971,
+      "step": 409
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.2869565217391305e-05,
+      "loss": 0.7088,
+      "step": 410
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.285217391304348e-05,
+      "loss": 0.4363,
+      "step": 411
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2834782608695654e-05,
+      "loss": 0.2082,
+      "step": 412
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2817391304347827e-05,
+      "loss": 0.5057,
+      "step": 413
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.2442,
+      "step": 414
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2782608695652175e-05,
+      "loss": 0.2221,
+      "step": 415
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.276521739130435e-05,
+      "loss": 0.7711,
+      "step": 416
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2747826086956522e-05,
+      "loss": 0.3222,
+      "step": 417
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2730434782608696e-05,
+      "loss": 0.4159,
+      "step": 418
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2713043478260871e-05,
+      "loss": 0.292,
+      "step": 419
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2695652173913045e-05,
+      "loss": 0.1779,
+      "step": 420
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2678260869565217e-05,
+      "loss": 0.3354,
+      "step": 421
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2660869565217392e-05,
+      "loss": 0.1743,
+      "step": 422
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2643478260869566e-05,
+      "loss": 0.3699,
+      "step": 423
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.262608695652174e-05,
+      "loss": 0.2609,
+      "step": 424
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2608695652173915e-05,
+      "loss": 0.4131,
+      "step": 425
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2591304347826087e-05,
+      "loss": 0.4248,
+      "step": 426
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2573913043478261e-05,
+      "loss": 0.3262,
+      "step": 427
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2556521739130436e-05,
+      "loss": 0.2009,
+      "step": 428
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2539130434782608e-05,
+      "loss": 0.1777,
+      "step": 429
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2521739130434784e-05,
+      "loss": 0.1225,
+      "step": 430
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2504347826086958e-05,
+      "loss": 0.4334,
+      "step": 431
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2486956521739131e-05,
+      "loss": 0.0985,
+      "step": 432
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2469565217391307e-05,
+      "loss": 0.5148,
+      "step": 433
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2452173913043479e-05,
+      "loss": 0.0474,
+      "step": 434
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2434782608695652e-05,
+      "loss": 0.3445,
+      "step": 435
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2417391304347828e-05,
+      "loss": 0.4612,
+      "step": 436
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 0.5018,
+      "step": 437
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2382608695652174e-05,
+      "loss": 0.3886,
+      "step": 438
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2365217391304349e-05,
+      "loss": 0.571,
+      "step": 439
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2347826086956523e-05,
+      "loss": 0.2972,
+      "step": 440
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2330434782608696e-05,
+      "loss": 0.3076,
+      "step": 441
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2313043478260872e-05,
+      "loss": 0.2234,
+      "step": 442
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2295652173913044e-05,
+      "loss": 0.2137,
+      "step": 443
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.227826086956522e-05,
+      "loss": 0.1354,
+      "step": 444
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2260869565217393e-05,
+      "loss": 0.2953,
+      "step": 445
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.2243478260869565e-05,
+      "loss": 0.277,
+      "step": 446
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.222608695652174e-05,
+      "loss": 0.2675,
+      "step": 447
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2208695652173914e-05,
+      "loss": 0.3333,
+      "step": 448
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2191304347826088e-05,
+      "loss": 0.1028,
+      "step": 449
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 0.2308,
+      "step": 450
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2156521739130435e-05,
+      "loss": 0.3156,
+      "step": 451
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2139130434782609e-05,
+      "loss": 0.6043,
+      "step": 452
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2121739130434784e-05,
+      "loss": 0.4437,
+      "step": 453
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2104347826086958e-05,
+      "loss": 0.1033,
+      "step": 454
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.208695652173913e-05,
+      "loss": 0.4178,
+      "step": 455
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2069565217391305e-05,
+      "loss": 0.2507,
+      "step": 456
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.205217391304348e-05,
+      "loss": 0.0593,
+      "step": 457
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2034782608695653e-05,
+      "loss": 0.1936,
+      "step": 458
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2017391304347827e-05,
+      "loss": 0.3568,
+      "step": 459
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2e-05,
+      "loss": 0.0347,
+      "step": 460
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8676470588235294,
+      "eval_combined_score": 0.8867852018856324,
+      "eval_f1": 0.9059233449477352,
+      "eval_loss": 0.3085273206233978,
+      "eval_runtime": 0.8484,
+      "eval_samples_per_second": 480.883,
+      "eval_steps_per_second": 30.644,
+      "step": 460
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.1982608695652176e-05,
+      "loss": 0.3521,
+      "step": 461
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.196521739130435e-05,
+      "loss": 0.1755,
+      "step": 462
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.1947826086956521e-05,
+      "loss": 0.1185,
+      "step": 463
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.1930434782608697e-05,
+      "loss": 0.3095,
+      "step": 464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.191304347826087e-05,
+      "loss": 0.0747,
+      "step": 465
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.1895652173913044e-05,
+      "loss": 0.2647,
+      "step": 466
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.187826086956522e-05,
+      "loss": 0.3068,
+      "step": 467
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.1860869565217392e-05,
+      "loss": 0.0687,
+      "step": 468
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.1843478260869565e-05,
+      "loss": 0.4003,
+      "step": 469
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.182608695652174e-05,
+      "loss": 0.4611,
+      "step": 470
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.1808695652173915e-05,
+      "loss": 0.066,
+      "step": 471
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.1791304347826087e-05,
+      "loss": 0.0616,
+      "step": 472
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.1773913043478262e-05,
+      "loss": 0.2787,
+      "step": 473
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.1756521739130436e-05,
+      "loss": 0.33,
+      "step": 474
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.1739130434782611e-05,
+      "loss": 0.5318,
+      "step": 475
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.1721739130434783e-05,
+      "loss": 0.0259,
+      "step": 476
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.1704347826086957e-05,
+      "loss": 0.3305,
+      "step": 477
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.1686956521739132e-05,
+      "loss": 0.2038,
+      "step": 478
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.1669565217391306e-05,
+      "loss": 0.9218,
+      "step": 479
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.1652173913043478e-05,
+      "loss": 0.2972,
+      "step": 480
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.1634782608695653e-05,
+      "loss": 0.1548,
+      "step": 481
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.1617391304347827e-05,
+      "loss": 0.4487,
+      "step": 482
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.16e-05,
+      "loss": 0.4787,
+      "step": 483
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.1582608695652176e-05,
+      "loss": 0.4108,
+      "step": 484
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.1565217391304348e-05,
+      "loss": 0.3003,
+      "step": 485
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.1547826086956522e-05,
+      "loss": 0.414,
+      "step": 486
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.1530434782608697e-05,
+      "loss": 0.141,
+      "step": 487
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.1513043478260871e-05,
+      "loss": 0.4491,
+      "step": 488
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.1495652173913043e-05,
+      "loss": 0.0346,
+      "step": 489
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.1478260869565218e-05,
+      "loss": 0.1743,
+      "step": 490
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.1460869565217392e-05,
+      "loss": 0.0551,
+      "step": 491
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.1443478260869568e-05,
+      "loss": 0.0885,
+      "step": 492
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.142608695652174e-05,
+      "loss": 0.0693,
+      "step": 493
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.1408695652173913e-05,
+      "loss": 0.4944,
+      "step": 494
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.1391304347826089e-05,
+      "loss": 0.7993,
+      "step": 495
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.1373913043478262e-05,
+      "loss": 0.0707,
+      "step": 496
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.1356521739130434e-05,
+      "loss": 0.1095,
+      "step": 497
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.133913043478261e-05,
+      "loss": 0.8768,
+      "step": 498
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.1321739130434784e-05,
+      "loss": 0.1798,
+      "step": 499
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 0.05,
+      "step": 500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.1286956521739133e-05,
+      "loss": 0.7645,
+      "step": 501
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.1269565217391305e-05,
+      "loss": 0.2485,
+      "step": 502
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.1252173913043478e-05,
+      "loss": 0.4699,
+      "step": 503
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.1234782608695654e-05,
+      "loss": 0.2427,
+      "step": 504
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.1217391304347827e-05,
+      "loss": 0.1491,
+      "step": 505
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.0925,
+      "step": 506
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.1182608695652175e-05,
+      "loss": 0.0854,
+      "step": 507
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.1165217391304349e-05,
+      "loss": 0.3939,
+      "step": 508
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.1147826086956524e-05,
+      "loss": 0.1946,
+      "step": 509
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.1130434782608696e-05,
+      "loss": 0.3126,
+      "step": 510
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.111304347826087e-05,
+      "loss": 0.3047,
+      "step": 511
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1095652173913045e-05,
+      "loss": 0.2353,
+      "step": 512
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1078260869565219e-05,
+      "loss": 0.029,
+      "step": 513
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1060869565217391e-05,
+      "loss": 0.1769,
+      "step": 514
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.1043478260869566e-05,
+      "loss": 0.1565,
+      "step": 515
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.102608695652174e-05,
+      "loss": 0.0528,
+      "step": 516
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.1008695652173914e-05,
+      "loss": 0.4256,
+      "step": 517
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.0991304347826089e-05,
+      "loss": 0.4634,
+      "step": 518
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0973913043478261e-05,
+      "loss": 0.1755,
+      "step": 519
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0956521739130435e-05,
+      "loss": 0.5839,
+      "step": 520
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.093913043478261e-05,
+      "loss": 0.0416,
+      "step": 521
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.0921739130434782e-05,
+      "loss": 0.0632,
+      "step": 522
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.0904347826086958e-05,
+      "loss": 0.0734,
+      "step": 523
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.0886956521739131e-05,
+      "loss": 0.3189,
+      "step": 524
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 0.1013,
+      "step": 525
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.085217391304348e-05,
+      "loss": 0.1645,
+      "step": 526
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.0834782608695653e-05,
+      "loss": 0.3379,
+      "step": 527
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.0817391304347826e-05,
+      "loss": 0.2974,
+      "step": 528
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.2642,
+      "step": 529
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.0782608695652175e-05,
+      "loss": 0.188,
+      "step": 530
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.0765217391304347e-05,
+      "loss": 0.0797,
+      "step": 531
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.0747826086956523e-05,
+      "loss": 0.177,
+      "step": 532
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.0730434782608696e-05,
+      "loss": 0.05,
+      "step": 533
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.071304347826087e-05,
+      "loss": 0.154,
+      "step": 534
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0695652173913046e-05,
+      "loss": 0.241,
+      "step": 535
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0678260869565218e-05,
+      "loss": 0.4628,
+      "step": 536
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0660869565217393e-05,
+      "loss": 0.4893,
+      "step": 537
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.0643478260869567e-05,
+      "loss": 0.3185,
+      "step": 538
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.0626086956521739e-05,
+      "loss": 0.0708,
+      "step": 539
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0608695652173914e-05,
+      "loss": 0.0885,
+      "step": 540
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0591304347826088e-05,
+      "loss": 0.0732,
+      "step": 541
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0573913043478262e-05,
+      "loss": 0.0234,
+      "step": 542
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0556521739130437e-05,
+      "loss": 0.0231,
+      "step": 543
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0539130434782609e-05,
+      "loss": 0.0256,
+      "step": 544
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0521739130434783e-05,
+      "loss": 0.0252,
+      "step": 545
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0504347826086958e-05,
+      "loss": 0.6101,
+      "step": 546
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0486956521739132e-05,
+      "loss": 0.1469,
+      "step": 547
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0469565217391304e-05,
+      "loss": 0.1335,
+      "step": 548
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.045217391304348e-05,
+      "loss": 0.0244,
+      "step": 549
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 0.3112,
+      "step": 550
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0417391304347827e-05,
+      "loss": 0.0121,
+      "step": 551
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.04e-05,
+      "loss": 0.2818,
+      "step": 552
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0382608695652174e-05,
+      "loss": 0.188,
+      "step": 553
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.036521739130435e-05,
+      "loss": 0.3004,
+      "step": 554
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.0347826086956523e-05,
+      "loss": 0.3766,
+      "step": 555
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0330434782608695e-05,
+      "loss": 0.5037,
+      "step": 556
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.031304347826087e-05,
+      "loss": 0.0739,
+      "step": 557
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.0295652173913044e-05,
+      "loss": 0.1796,
+      "step": 558
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.0278260869565218e-05,
+      "loss": 0.0348,
+      "step": 559
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.0260869565217393e-05,
+      "loss": 0.6226,
+      "step": 560
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.0243478260869565e-05,
+      "loss": 0.0116,
+      "step": 561
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.022608695652174e-05,
+      "loss": 0.0451,
+      "step": 562
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.0208695652173915e-05,
+      "loss": 0.098,
+      "step": 563
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.0191304347826088e-05,
+      "loss": 0.3547,
+      "step": 564
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.017391304347826e-05,
+      "loss": 0.1367,
+      "step": 565
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.0156521739130436e-05,
+      "loss": 0.0483,
+      "step": 566
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.013913043478261e-05,
+      "loss": 0.1649,
+      "step": 567
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.0121739130434785e-05,
+      "loss": 0.0136,
+      "step": 568
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.0104347826086957e-05,
+      "loss": 0.2803,
+      "step": 569
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.008695652173913e-05,
+      "loss": 0.0053,
+      "step": 570
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.0069565217391306e-05,
+      "loss": 0.3131,
+      "step": 571
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.005217391304348e-05,
+      "loss": 0.0358,
+      "step": 572
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.0034782608695652e-05,
+      "loss": 0.8274,
+      "step": 573
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.0017391304347827e-05,
+      "loss": 0.3065,
+      "step": 574
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3316,
+      "step": 575
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 9.982608695652175e-06,
+      "loss": 0.7679,
+      "step": 576
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 9.965217391304348e-06,
+      "loss": 0.0127,
+      "step": 577
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 9.947826086956522e-06,
+      "loss": 0.0435,
+      "step": 578
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 9.930434782608697e-06,
+      "loss": 0.216,
+      "step": 579
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 9.913043478260871e-06,
+      "loss": 0.5712,
+      "step": 580
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 9.895652173913045e-06,
+      "loss": 0.0361,
+      "step": 581
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 9.878260869565218e-06,
+      "loss": 0.0117,
+      "step": 582
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 9.860869565217392e-06,
+      "loss": 0.0163,
+      "step": 583
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 9.843478260869566e-06,
+      "loss": 0.3179,
+      "step": 584
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 9.82608695652174e-06,
+      "loss": 0.3295,
+      "step": 585
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 9.808695652173913e-06,
+      "loss": 0.3849,
+      "step": 586
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 9.791304347826089e-06,
+      "loss": 0.1785,
+      "step": 587
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 9.77391304347826e-06,
+      "loss": 0.2606,
+      "step": 588
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 9.756521739130436e-06,
+      "loss": 0.324,
+      "step": 589
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.73913043478261e-06,
+      "loss": 0.6198,
+      "step": 590
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.721739130434784e-06,
+      "loss": 0.1119,
+      "step": 591
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.704347826086957e-06,
+      "loss": 0.279,
+      "step": 592
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.686956521739131e-06,
+      "loss": 0.1919,
+      "step": 593
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.669565217391305e-06,
+      "loss": 0.5522,
+      "step": 594
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.652173913043478e-06,
+      "loss": 0.2225,
+      "step": 595
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.634782608695654e-06,
+      "loss": 0.2221,
+      "step": 596
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.617391304347828e-06,
+      "loss": 0.559,
+      "step": 597
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.5006,
+      "step": 598
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.582608695652175e-06,
+      "loss": 0.3474,
+      "step": 599
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.6286,
+      "step": 600
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.547826086956522e-06,
+      "loss": 0.0823,
+      "step": 601
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.530434782608696e-06,
+      "loss": 0.1592,
+      "step": 602
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.51304347826087e-06,
+      "loss": 0.0278,
+      "step": 603
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.495652173913045e-06,
+      "loss": 0.2917,
+      "step": 604
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.478260869565217e-06,
+      "loss": 0.4633,
+      "step": 605
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.460869565217393e-06,
+      "loss": 0.1219,
+      "step": 606
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.443478260869566e-06,
+      "loss": 0.3367,
+      "step": 607
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.42608695652174e-06,
+      "loss": 0.3698,
+      "step": 608
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 9.408695652173914e-06,
+      "loss": 0.2691,
+      "step": 609
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 9.391304347826087e-06,
+      "loss": 0.3997,
+      "step": 610
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 9.373913043478263e-06,
+      "loss": 0.4115,
+      "step": 611
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 9.356521739130435e-06,
+      "loss": 0.2619,
+      "step": 612
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 9.33913043478261e-06,
+      "loss": 0.1874,
+      "step": 613
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 9.321739130434784e-06,
+      "loss": 0.4613,
+      "step": 614
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 9.304347826086956e-06,
+      "loss": 0.3104,
+      "step": 615
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 9.286956521739131e-06,
+      "loss": 0.0398,
+      "step": 616
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 9.269565217391305e-06,
+      "loss": 0.0947,
+      "step": 617
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 9.252173913043479e-06,
+      "loss": 0.1311,
+      "step": 618
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 9.234782608695653e-06,
+      "loss": 0.0821,
+      "step": 619
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.217391304347826e-06,
+      "loss": 0.1061,
+      "step": 620
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.2746,
+      "step": 621
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.182608695652174e-06,
+      "loss": 0.1568,
+      "step": 622
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.165217391304349e-06,
+      "loss": 0.6806,
+      "step": 623
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.147826086956523e-06,
+      "loss": 0.2819,
+      "step": 624
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.1007,
+      "step": 625
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.11304347826087e-06,
+      "loss": 0.3024,
+      "step": 626
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 9.095652173913044e-06,
+      "loss": 0.1763,
+      "step": 627
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 9.07826086956522e-06,
+      "loss": 0.0471,
+      "step": 628
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 9.060869565217391e-06,
+      "loss": 0.1092,
+      "step": 629
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 9.043478260869565e-06,
+      "loss": 0.0747,
+      "step": 630
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 9.02608695652174e-06,
+      "loss": 0.4771,
+      "step": 631
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 9.008695652173913e-06,
+      "loss": 0.3313,
+      "step": 632
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.991304347826088e-06,
+      "loss": 0.0784,
+      "step": 633
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.973913043478262e-06,
+      "loss": 0.1701,
+      "step": 634
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.956521739130435e-06,
+      "loss": 0.0578,
+      "step": 635
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.939130434782609e-06,
+      "loss": 0.3333,
+      "step": 636
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.921739130434783e-06,
+      "loss": 0.0231,
+      "step": 637
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.904347826086958e-06,
+      "loss": 0.2633,
+      "step": 638
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.88695652173913e-06,
+      "loss": 0.1508,
+      "step": 639
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.869565217391306e-06,
+      "loss": 0.3064,
+      "step": 640
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.85217391304348e-06,
+      "loss": 0.3101,
+      "step": 641
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.834782608695653e-06,
+      "loss": 0.0312,
+      "step": 642
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.817391304347827e-06,
+      "loss": 0.024,
+      "step": 643
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.8e-06,
+      "loss": 0.0259,
+      "step": 644
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.782608695652174e-06,
+      "loss": 0.0927,
+      "step": 645
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.765217391304348e-06,
+      "loss": 0.7142,
+      "step": 646
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.747826086956522e-06,
+      "loss": 0.2146,
+      "step": 647
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.730434782608697e-06,
+      "loss": 0.5336,
+      "step": 648
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.71304347826087e-06,
+      "loss": 0.1424,
+      "step": 649
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.3703,
+      "step": 650
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.678260869565218e-06,
+      "loss": 0.0266,
+      "step": 651
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.660869565217392e-06,
+      "loss": 0.3533,
+      "step": 652
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 8.643478260869566e-06,
+      "loss": 0.6236,
+      "step": 653
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 8.62608695652174e-06,
+      "loss": 0.1666,
+      "step": 654
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.608695652173915e-06,
+      "loss": 0.3283,
+      "step": 655
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.591304347826087e-06,
+      "loss": 0.0161,
+      "step": 656
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.573913043478262e-06,
+      "loss": 0.1311,
+      "step": 657
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.556521739130436e-06,
+      "loss": 0.2977,
+      "step": 658
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.53913043478261e-06,
+      "loss": 0.3599,
+      "step": 659
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.521739130434783e-06,
+      "loss": 0.4188,
+      "step": 660
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.504347826086957e-06,
+      "loss": 0.0419,
+      "step": 661
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.48695652173913e-06,
+      "loss": 0.7119,
+      "step": 662
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.469565217391304e-06,
+      "loss": 0.0113,
+      "step": 663
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.452173913043478e-06,
+      "loss": 0.3197,
+      "step": 664
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 8.434782608695653e-06,
+      "loss": 0.1822,
+      "step": 665
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.417391304347827e-06,
+      "loss": 0.0761,
+      "step": 666
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.2202,
+      "step": 667
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.382608695652175e-06,
+      "loss": 0.0883,
+      "step": 668
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.365217391304348e-06,
+      "loss": 0.4138,
+      "step": 669
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 8.347826086956522e-06,
+      "loss": 0.0377,
+      "step": 670
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.330434782608696e-06,
+      "loss": 0.0398,
+      "step": 671
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.313043478260871e-06,
+      "loss": 0.1546,
+      "step": 672
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.295652173913045e-06,
+      "loss": 0.6689,
+      "step": 673
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.278260869565219e-06,
+      "loss": 0.0663,
+      "step": 674
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 0.179,
+      "step": 675
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 8.243478260869566e-06,
+      "loss": 0.226,
+      "step": 676
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 8.22608695652174e-06,
+      "loss": 0.2603,
+      "step": 677
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.208695652173913e-06,
+      "loss": 0.0918,
+      "step": 678
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.191304347826087e-06,
+      "loss": 0.1255,
+      "step": 679
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.173913043478263e-06,
+      "loss": 0.0142,
+      "step": 680
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.156521739130435e-06,
+      "loss": 0.0249,
+      "step": 681
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.13913043478261e-06,
+      "loss": 0.0388,
+      "step": 682
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.121739130434784e-06,
+      "loss": 0.086,
+      "step": 683
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.104347826086957e-06,
+      "loss": 0.3278,
+      "step": 684
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 8.086956521739131e-06,
+      "loss": 0.0522,
+      "step": 685
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 8.069565217391305e-06,
+      "loss": 0.1906,
+      "step": 686
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.052173913043479e-06,
+      "loss": 0.3307,
+      "step": 687
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.034782608695652e-06,
+      "loss": 0.1262,
+      "step": 688
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.017391304347828e-06,
+      "loss": 0.0651,
+      "step": 689
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.0495,
+      "step": 690
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8848039215686274,
+      "eval_combined_score": 0.9011014335435404,
+      "eval_f1": 0.9173989455184535,
+      "eval_loss": 0.35520851612091064,
+      "eval_runtime": 0.7851,
+      "eval_samples_per_second": 519.668,
+      "eval_steps_per_second": 33.116,
+      "step": 690
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.982608695652175e-06,
+      "loss": 0.0735,
+      "step": 691
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 7.965217391304349e-06,
+      "loss": 0.0293,
+      "step": 692
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 7.947826086956522e-06,
+      "loss": 0.417,
+      "step": 693
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 7.930434782608696e-06,
+      "loss": 0.056,
+      "step": 694
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 7.91304347826087e-06,
+      "loss": 0.0181,
+      "step": 695
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 7.895652173913044e-06,
+      "loss": 0.0273,
+      "step": 696
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 7.878260869565219e-06,
+      "loss": 0.0357,
+      "step": 697
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 7.860869565217391e-06,
+      "loss": 0.0412,
+      "step": 698
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 7.843478260869566e-06,
+      "loss": 0.0303,
+      "step": 699
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.2836,
+      "step": 700
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 7.808695652173914e-06,
+      "loss": 0.04,
+      "step": 701
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 7.791304347826088e-06,
+      "loss": 0.0543,
+      "step": 702
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 7.773913043478261e-06,
+      "loss": 0.1054,
+      "step": 703
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 7.756521739130437e-06,
+      "loss": 0.3375,
+      "step": 704
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.739130434782609e-06,
+      "loss": 0.3433,
+      "step": 705
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.721739130434784e-06,
+      "loss": 0.2622,
+      "step": 706
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.704347826086958e-06,
+      "loss": 0.0086,
+      "step": 707
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 7.68695652173913e-06,
+      "loss": 0.0078,
+      "step": 708
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 7.669565217391305e-06,
+      "loss": 0.0452,
+      "step": 709
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 7.652173913043479e-06,
+      "loss": 0.0546,
+      "step": 710
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 7.634782608695653e-06,
+      "loss": 0.0276,
+      "step": 711
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 7.617391304347826e-06,
+      "loss": 0.0205,
+      "step": 712
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.1078,
+      "step": 713
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 7.582608695652175e-06,
+      "loss": 0.1543,
+      "step": 714
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 7.565217391304348e-06,
+      "loss": 0.3508,
+      "step": 715
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 7.547826086956522e-06,
+      "loss": 0.0058,
+      "step": 716
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 7.530434782608697e-06,
+      "loss": 0.0056,
+      "step": 717
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 7.5130434782608695e-06,
+      "loss": 0.0063,
+      "step": 718
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 7.495652173913044e-06,
+      "loss": 0.1083,
+      "step": 719
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 7.478260869565218e-06,
+      "loss": 0.0892,
+      "step": 720
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 7.460869565217392e-06,
+      "loss": 0.0056,
+      "step": 721
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 7.443478260869565e-06,
+      "loss": 0.1768,
+      "step": 722
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 7.42608695652174e-06,
+      "loss": 0.0087,
+      "step": 723
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 7.408695652173914e-06,
+      "loss": 0.7567,
+      "step": 724
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 0.0652,
+      "step": 725
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.373913043478262e-06,
+      "loss": 0.1594,
+      "step": 726
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 7.3565217391304354e-06,
+      "loss": 0.0048,
+      "step": 727
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.33913043478261e-06,
+      "loss": 0.1259,
+      "step": 728
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.321739130434783e-06,
+      "loss": 0.0036,
+      "step": 729
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.304347826086957e-06,
+      "loss": 0.1217,
+      "step": 730
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.286956521739131e-06,
+      "loss": 0.0666,
+      "step": 731
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.269565217391305e-06,
+      "loss": 0.2018,
+      "step": 732
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.2521739130434785e-06,
+      "loss": 0.2622,
+      "step": 733
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 7.234782608695653e-06,
+      "loss": 0.3308,
+      "step": 734
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 7.217391304347827e-06,
+      "loss": 0.4958,
+      "step": 735
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.004,
+      "step": 736
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 7.182608695652174e-06,
+      "loss": 0.692,
+      "step": 737
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 7.165217391304349e-06,
+      "loss": 0.3041,
+      "step": 738
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 7.147826086956522e-06,
+      "loss": 0.3666,
+      "step": 739
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 7.130434782608696e-06,
+      "loss": 0.7629,
+      "step": 740
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 7.113043478260871e-06,
+      "loss": 0.1392,
+      "step": 741
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 7.095652173913044e-06,
+      "loss": 0.3131,
+      "step": 742
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 7.078260869565217e-06,
+      "loss": 0.0964,
+      "step": 743
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 7.060869565217392e-06,
+      "loss": 0.0036,
+      "step": 744
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 7.0434782608695665e-06,
+      "loss": 0.0362,
+      "step": 745
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 7.026086956521739e-06,
+      "loss": 0.0088,
+      "step": 746
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 7.008695652173914e-06,
+      "loss": 0.3983,
+      "step": 747
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 6.991304347826088e-06,
+      "loss": 0.0062,
+      "step": 748
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 6.973913043478261e-06,
+      "loss": 0.0094,
+      "step": 749
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.0055,
+      "step": 750
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 6.93913043478261e-06,
+      "loss": 0.6365,
+      "step": 751
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 6.921739130434783e-06,
+      "loss": 0.1401,
+      "step": 752
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 6.904347826086957e-06,
+      "loss": 0.0085,
+      "step": 753
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 6.886956521739131e-06,
+      "loss": 0.8913,
+      "step": 754
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 6.869565217391305e-06,
+      "loss": 0.373,
+      "step": 755
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 6.852173913043478e-06,
+      "loss": 0.3847,
+      "step": 756
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 6.834782608695653e-06,
+      "loss": 0.102,
+      "step": 757
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 6.817391304347826e-06,
+      "loss": 0.0043,
+      "step": 758
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.0034,
+      "step": 759
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 6.782608695652174e-06,
+      "loss": 0.0271,
+      "step": 760
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 6.765217391304348e-06,
+      "loss": 0.0083,
+      "step": 761
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 6.747826086956523e-06,
+      "loss": 0.2851,
+      "step": 762
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 6.730434782608696e-06,
+      "loss": 0.0044,
+      "step": 763
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 6.71304347826087e-06,
+      "loss": 0.2568,
+      "step": 764
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 6.695652173913044e-06,
+      "loss": 0.0077,
+      "step": 765
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 6.678260869565219e-06,
+      "loss": 0.1357,
+      "step": 766
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 6.6608695652173915e-06,
+      "loss": 0.5939,
+      "step": 767
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 6.643478260869566e-06,
+      "loss": 0.5382,
+      "step": 768
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 6.62608695652174e-06,
+      "loss": 0.0284,
+      "step": 769
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 6.6086956521739135e-06,
+      "loss": 0.0241,
+      "step": 770
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 6.591304347826087e-06,
+      "loss": 0.2738,
+      "step": 771
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 6.573913043478262e-06,
+      "loss": 0.0095,
+      "step": 772
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 6.556521739130435e-06,
+      "loss": 0.615,
+      "step": 773
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 6.539130434782609e-06,
+      "loss": 0.006,
+      "step": 774
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.2177,
+      "step": 775
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 6.5043478260869574e-06,
+      "loss": 0.1905,
+      "step": 776
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 6.48695652173913e-06,
+      "loss": 0.1562,
+      "step": 777
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 6.469565217391305e-06,
+      "loss": 0.0029,
+      "step": 778
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 6.4521739130434794e-06,
+      "loss": 0.0063,
+      "step": 779
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 6.434782608695652e-06,
+      "loss": 0.0041,
+      "step": 780
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 6.417391304347827e-06,
+      "loss": 0.5212,
+      "step": 781
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.3126,
+      "step": 782
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 6.382608695652175e-06,
+      "loss": 0.3126,
+      "step": 783
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 6.365217391304348e-06,
+      "loss": 0.0092,
+      "step": 784
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 6.3478260869565225e-06,
+      "loss": 0.3373,
+      "step": 785
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 6.330434782608696e-06,
+      "loss": 0.4636,
+      "step": 786
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 6.31304347826087e-06,
+      "loss": 0.004,
+      "step": 787
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 6.295652173913044e-06,
+      "loss": 0.0846,
+      "step": 788
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 6.278260869565218e-06,
+      "loss": 0.0035,
+      "step": 789
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 6.260869565217392e-06,
+      "loss": 0.2927,
+      "step": 790
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 6.243478260869566e-06,
+      "loss": 0.0185,
+      "step": 791
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 6.226086956521739e-06,
+      "loss": 0.0043,
+      "step": 792
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 6.208695652173914e-06,
+      "loss": 0.0755,
+      "step": 793
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 6.191304347826087e-06,
+      "loss": 0.0218,
+      "step": 794
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 6.173913043478261e-06,
+      "loss": 0.0778,
+      "step": 795
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 6.156521739130436e-06,
+      "loss": 0.2174,
+      "step": 796
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 6.13913043478261e-06,
+      "loss": 0.0048,
+      "step": 797
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 6.1217391304347825e-06,
+      "loss": 0.0091,
+      "step": 798
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 6.104347826086957e-06,
+      "loss": 0.3581,
+      "step": 799
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 0.3201,
+      "step": 800
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 6.0695652173913045e-06,
+      "loss": 0.0028,
+      "step": 801
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 6.052173913043479e-06,
+      "loss": 0.0043,
+      "step": 802
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 6.034782608695653e-06,
+      "loss": 0.0505,
+      "step": 803
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 6.0173913043478264e-06,
+      "loss": 0.1975,
+      "step": 804
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 6e-06,
+      "loss": 0.298,
+      "step": 805
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5.982608695652175e-06,
+      "loss": 0.0066,
+      "step": 806
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5.965217391304348e-06,
+      "loss": 0.0156,
+      "step": 807
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5.947826086956522e-06,
+      "loss": 0.0058,
+      "step": 808
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.930434782608696e-06,
+      "loss": 0.099,
+      "step": 809
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.91304347826087e-06,
+      "loss": 0.289,
+      "step": 810
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.895652173913043e-06,
+      "loss": 0.0821,
+      "step": 811
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.878260869565218e-06,
+      "loss": 0.0364,
+      "step": 812
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.8608695652173915e-06,
+      "loss": 0.3247,
+      "step": 813
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.843478260869566e-06,
+      "loss": 0.0272,
+      "step": 814
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.826086956521739e-06,
+      "loss": 0.1369,
+      "step": 815
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.8086956521739135e-06,
+      "loss": 0.0065,
+      "step": 816
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.791304347826088e-06,
+      "loss": 0.7298,
+      "step": 817
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.773913043478261e-06,
+      "loss": 0.0323,
+      "step": 818
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.7565217391304355e-06,
+      "loss": 0.0042,
+      "step": 819
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.739130434782609e-06,
+      "loss": 0.173,
+      "step": 820
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.721739130434784e-06,
+      "loss": 0.042,
+      "step": 821
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.704347826086957e-06,
+      "loss": 0.1546,
+      "step": 822
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.686956521739131e-06,
+      "loss": 0.0082,
+      "step": 823
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.669565217391305e-06,
+      "loss": 0.005,
+      "step": 824
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 0.0053,
+      "step": 825
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.634782608695652e-06,
+      "loss": 0.2,
+      "step": 826
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.617391304347827e-06,
+      "loss": 0.0021,
+      "step": 827
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.0071,
+      "step": 828
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.582608695652174e-06,
+      "loss": 0.3061,
+      "step": 829
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.565217391304348e-06,
+      "loss": 0.018,
+      "step": 830
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.5478260869565226e-06,
+      "loss": 0.5691,
+      "step": 831
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.5304347826086954e-06,
+      "loss": 0.0036,
+      "step": 832
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.51304347826087e-06,
+      "loss": 0.4895,
+      "step": 833
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5.4956521739130446e-06,
+      "loss": 0.3744,
+      "step": 834
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5.478260869565217e-06,
+      "loss": 0.1054,
+      "step": 835
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5.460869565217391e-06,
+      "loss": 0.3853,
+      "step": 836
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5.443478260869566e-06,
+      "loss": 0.0026,
+      "step": 837
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5.42608695652174e-06,
+      "loss": 0.794,
+      "step": 838
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5.408695652173913e-06,
+      "loss": 0.0403,
+      "step": 839
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5.391304347826088e-06,
+      "loss": 0.0029,
+      "step": 840
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5.373913043478261e-06,
+      "loss": 0.288,
+      "step": 841
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5.356521739130435e-06,
+      "loss": 0.0066,
+      "step": 842
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5.339130434782609e-06,
+      "loss": 0.011,
+      "step": 843
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5.321739130434783e-06,
+      "loss": 0.003,
+      "step": 844
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5.304347826086957e-06,
+      "loss": 0.0118,
+      "step": 845
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5.286956521739131e-06,
+      "loss": 0.3774,
+      "step": 846
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5.2695652173913045e-06,
+      "loss": 0.0144,
+      "step": 847
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5.252173913043479e-06,
+      "loss": 0.0352,
+      "step": 848
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5.234782608695652e-06,
+      "loss": 0.084,
+      "step": 849
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 0.441,
+      "step": 850
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5.2e-06,
+      "loss": 0.317,
+      "step": 851
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5.182608695652175e-06,
+      "loss": 0.2945,
+      "step": 852
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5.165217391304348e-06,
+      "loss": 0.2824,
+      "step": 853
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5.147826086956522e-06,
+      "loss": 0.8795,
+      "step": 854
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5.130434782608697e-06,
+      "loss": 0.0024,
+      "step": 855
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5.11304347826087e-06,
+      "loss": 0.0132,
+      "step": 856
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.095652173913044e-06,
+      "loss": 0.0356,
+      "step": 857
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.078260869565218e-06,
+      "loss": 0.2225,
+      "step": 858
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.060869565217392e-06,
+      "loss": 0.0069,
+      "step": 859
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.043478260869565e-06,
+      "loss": 0.1165,
+      "step": 860
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.02608695652174e-06,
+      "loss": 0.0145,
+      "step": 861
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.0086956521739136e-06,
+      "loss": 0.0388,
+      "step": 862
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 4.991304347826087e-06,
+      "loss": 0.5738,
+      "step": 863
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 4.973913043478261e-06,
+      "loss": 0.0033,
+      "step": 864
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 4.9565217391304355e-06,
+      "loss": 0.0046,
+      "step": 865
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 4.939130434782609e-06,
+      "loss": 0.0078,
+      "step": 866
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 4.921739130434783e-06,
+      "loss": 0.0036,
+      "step": 867
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 4.904347826086957e-06,
+      "loss": 0.008,
+      "step": 868
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 4.88695652173913e-06,
+      "loss": 0.0044,
+      "step": 869
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 4.869565217391305e-06,
+      "loss": 0.2957,
+      "step": 870
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 4.852173913043479e-06,
+      "loss": 0.2203,
+      "step": 871
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 4.834782608695652e-06,
+      "loss": 0.0065,
+      "step": 872
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 4.817391304347827e-06,
+      "loss": 0.3126,
+      "step": 873
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.002,
+      "step": 874
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 0.0055,
+      "step": 875
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 4.765217391304348e-06,
+      "loss": 0.0044,
+      "step": 876
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 4.747826086956523e-06,
+      "loss": 0.391,
+      "step": 877
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 4.730434782608696e-06,
+      "loss": 0.1417,
+      "step": 878
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 4.71304347826087e-06,
+      "loss": 0.0029,
+      "step": 879
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.695652173913044e-06,
+      "loss": 0.0027,
+      "step": 880
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.6782608695652175e-06,
+      "loss": 0.3686,
+      "step": 881
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.660869565217392e-06,
+      "loss": 0.0107,
+      "step": 882
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 4.643478260869566e-06,
+      "loss": 0.2161,
+      "step": 883
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 4.6260869565217394e-06,
+      "loss": 0.026,
+      "step": 884
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 4.608695652173913e-06,
+      "loss": 0.0789,
+      "step": 885
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 4.591304347826087e-06,
+      "loss": 0.0058,
+      "step": 886
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 4.573913043478261e-06,
+      "loss": 0.4275,
+      "step": 887
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 4.556521739130435e-06,
+      "loss": 0.0231,
+      "step": 888
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.53913043478261e-06,
+      "loss": 0.2705,
+      "step": 889
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.5217391304347826e-06,
+      "loss": 0.003,
+      "step": 890
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.504347826086956e-06,
+      "loss": 0.4858,
+      "step": 891
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.486956521739131e-06,
+      "loss": 0.041,
+      "step": 892
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.4695652173913045e-06,
+      "loss": 0.0057,
+      "step": 893
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.452173913043479e-06,
+      "loss": 0.2002,
+      "step": 894
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.434782608695653e-06,
+      "loss": 0.022,
+      "step": 895
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.4173913043478265e-06,
+      "loss": 0.0075,
+      "step": 896
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0029,
+      "step": 897
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.382608695652174e-06,
+      "loss": 0.0042,
+      "step": 898
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.3652173913043485e-06,
+      "loss": 0.0052,
+      "step": 899
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.0037,
+      "step": 900
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.330434782608696e-06,
+      "loss": 0.0178,
+      "step": 901
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.31304347826087e-06,
+      "loss": 0.0041,
+      "step": 902
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 4.295652173913043e-06,
+      "loss": 0.2758,
+      "step": 903
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 4.278260869565218e-06,
+      "loss": 0.2233,
+      "step": 904
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 4.260869565217392e-06,
+      "loss": 0.4452,
+      "step": 905
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 4.243478260869565e-06,
+      "loss": 0.391,
+      "step": 906
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 4.226086956521739e-06,
+      "loss": 0.1382,
+      "step": 907
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.208695652173914e-06,
+      "loss": 0.0026,
+      "step": 908
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.191304347826087e-06,
+      "loss": 0.0145,
+      "step": 909
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 4.173913043478261e-06,
+      "loss": 0.3204,
+      "step": 910
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 4.1565217391304356e-06,
+      "loss": 0.35,
+      "step": 911
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.139130434782609e-06,
+      "loss": 0.0435,
+      "step": 912
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.121739130434783e-06,
+      "loss": 0.407,
+      "step": 913
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.104347826086957e-06,
+      "loss": 0.0472,
+      "step": 914
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 4.086956521739131e-06,
+      "loss": 0.3783,
+      "step": 915
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 4.069565217391305e-06,
+      "loss": 0.6374,
+      "step": 916
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 4.052173913043479e-06,
+      "loss": 0.2607,
+      "step": 917
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 4.034782608695652e-06,
+      "loss": 0.109,
+      "step": 918
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.017391304347826e-06,
+      "loss": 0.0039,
+      "step": 919
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0024,
+      "step": 920
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8823529411764706,
+      "eval_combined_score": 0.8990712074303405,
+      "eval_f1": 0.9157894736842105,
+      "eval_loss": 0.4960119128227234,
+      "eval_runtime": 0.8346,
+      "eval_samples_per_second": 488.843,
+      "eval_steps_per_second": 31.152,
+      "step": 920
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.982608695652174e-06,
+      "loss": 0.0102,
+      "step": 921
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.965217391304348e-06,
+      "loss": 0.0049,
+      "step": 922
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.947826086956522e-06,
+      "loss": 0.0863,
+      "step": 923
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.9304347826086955e-06,
+      "loss": 0.1483,
+      "step": 924
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 0.0044,
+      "step": 925
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.895652173913044e-06,
+      "loss": 0.0109,
+      "step": 926
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.878260869565218e-06,
+      "loss": 0.0054,
+      "step": 927
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.860869565217392e-06,
+      "loss": 0.1617,
+      "step": 928
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.843478260869565e-06,
+      "loss": 0.0309,
+      "step": 929
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.8260869565217395e-06,
+      "loss": 0.0034,
+      "step": 930
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.808695652173913e-06,
+      "loss": 0.0141,
+      "step": 931
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.7913043478260873e-06,
+      "loss": 0.0055,
+      "step": 932
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.773913043478261e-06,
+      "loss": 0.0034,
+      "step": 933
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.7565217391304347e-06,
+      "loss": 0.0033,
+      "step": 934
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.739130434782609e-06,
+      "loss": 0.0036,
+      "step": 935
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.7217391304347826e-06,
+      "loss": 0.314,
+      "step": 936
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.704347826086957e-06,
+      "loss": 0.0065,
+      "step": 937
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.686956521739131e-06,
+      "loss": 0.002,
+      "step": 938
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.669565217391305e-06,
+      "loss": 0.0025,
+      "step": 939
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.6521739130434787e-06,
+      "loss": 0.1703,
+      "step": 940
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.6347826086956524e-06,
+      "loss": 0.1488,
+      "step": 941
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.6173913043478265e-06,
+      "loss": 0.0022,
+      "step": 942
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.3252,
+      "step": 943
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.5826086956521744e-06,
+      "loss": 0.0016,
+      "step": 944
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.565217391304348e-06,
+      "loss": 0.0028,
+      "step": 945
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.547826086956522e-06,
+      "loss": 0.2527,
+      "step": 946
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.530434782608696e-06,
+      "loss": 0.0039,
+      "step": 947
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.5130434782608697e-06,
+      "loss": 0.0023,
+      "step": 948
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.495652173913044e-06,
+      "loss": 0.0077,
+      "step": 949
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 0.0035,
+      "step": 950
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.4608695652173916e-06,
+      "loss": 0.0046,
+      "step": 951
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.4434782608695654e-06,
+      "loss": 0.3692,
+      "step": 952
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.426086956521739e-06,
+      "loss": 0.2441,
+      "step": 953
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.408695652173913e-06,
+      "loss": 0.412,
+      "step": 954
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.391304347826087e-06,
+      "loss": 0.0075,
+      "step": 955
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 3.3739130434782615e-06,
+      "loss": 0.0019,
+      "step": 956
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 3.356521739130435e-06,
+      "loss": 0.1191,
+      "step": 957
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 3.3391304347826093e-06,
+      "loss": 0.2743,
+      "step": 958
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 3.321739130434783e-06,
+      "loss": 0.0284,
+      "step": 959
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 3.3043478260869567e-06,
+      "loss": 0.002,
+      "step": 960
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 3.286956521739131e-06,
+      "loss": 0.0027,
+      "step": 961
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 3.2695652173913046e-06,
+      "loss": 0.0025,
+      "step": 962
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.2521739130434787e-06,
+      "loss": 0.0096,
+      "step": 963
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.2347826086956524e-06,
+      "loss": 0.0231,
+      "step": 964
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.217391304347826e-06,
+      "loss": 0.0023,
+      "step": 965
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0029,
+      "step": 966
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.182608695652174e-06,
+      "loss": 0.0025,
+      "step": 967
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 3.165217391304348e-06,
+      "loss": 0.3091,
+      "step": 968
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 3.147826086956522e-06,
+      "loss": 0.0015,
+      "step": 969
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 3.130434782608696e-06,
+      "loss": 0.4621,
+      "step": 970
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 3.1130434782608697e-06,
+      "loss": 0.0018,
+      "step": 971
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 3.0956521739130434e-06,
+      "loss": 0.0172,
+      "step": 972
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 3.078260869565218e-06,
+      "loss": 0.0016,
+      "step": 973
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 3.0608695652173912e-06,
+      "loss": 0.0127,
+      "step": 974
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 0.0095,
+      "step": 975
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 3.0260869565217395e-06,
+      "loss": 0.0016,
+      "step": 976
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 3.0086956521739132e-06,
+      "loss": 0.0831,
+      "step": 977
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.9913043478260874e-06,
+      "loss": 0.0038,
+      "step": 978
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.973913043478261e-06,
+      "loss": 0.0125,
+      "step": 979
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.956521739130435e-06,
+      "loss": 0.313,
+      "step": 980
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.939130434782609e-06,
+      "loss": 0.0018,
+      "step": 981
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.921739130434783e-06,
+      "loss": 0.0014,
+      "step": 982
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.9043478260869568e-06,
+      "loss": 0.0097,
+      "step": 983
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.8869565217391305e-06,
+      "loss": 0.0014,
+      "step": 984
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.8695652173913046e-06,
+      "loss": 0.0017,
+      "step": 985
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.8521739130434783e-06,
+      "loss": 0.0061,
+      "step": 986
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.8347826086956525e-06,
+      "loss": 0.0018,
+      "step": 987
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.817391304347826e-06,
+      "loss": 0.4488,
+      "step": 988
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.0022,
+      "step": 989
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.782608695652174e-06,
+      "loss": 0.4577,
+      "step": 990
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.7652173913043477e-06,
+      "loss": 0.0016,
+      "step": 991
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.7478260869565223e-06,
+      "loss": 0.0013,
+      "step": 992
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 2.7304347826086956e-06,
+      "loss": 0.004,
+      "step": 993
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 2.71304347826087e-06,
+      "loss": 0.0048,
+      "step": 994
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.695652173913044e-06,
+      "loss": 0.0364,
+      "step": 995
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.6782608695652175e-06,
+      "loss": 0.0013,
+      "step": 996
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.6608695652173917e-06,
+      "loss": 0.3396,
+      "step": 997
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.6434782608695654e-06,
+      "loss": 0.0025,
+      "step": 998
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.6260869565217395e-06,
+      "loss": 0.0018,
+      "step": 999
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 0.2721,
+      "step": 1000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.5913043478260874e-06,
+      "loss": 0.2693,
+      "step": 1001
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.573913043478261e-06,
+      "loss": 0.3681,
+      "step": 1002
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.556521739130435e-06,
+      "loss": 0.1959,
+      "step": 1003
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.539130434782609e-06,
+      "loss": 0.4282,
+      "step": 1004
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.5217391304347826e-06,
+      "loss": 0.0015,
+      "step": 1005
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.5043478260869568e-06,
+      "loss": 0.253,
+      "step": 1006
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.4869565217391305e-06,
+      "loss": 0.2982,
+      "step": 1007
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.4695652173913046e-06,
+      "loss": 0.0312,
+      "step": 1008
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.4521739130434783e-06,
+      "loss": 0.1157,
+      "step": 1009
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.4347826086956525e-06,
+      "loss": 0.003,
+      "step": 1010
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.417391304347826e-06,
+      "loss": 0.0014,
+      "step": 1011
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.0015,
+      "step": 1012
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.382608695652174e-06,
+      "loss": 0.3158,
+      "step": 1013
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 2.365217391304348e-06,
+      "loss": 0.0022,
+      "step": 1014
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 2.347826086956522e-06,
+      "loss": 0.5545,
+      "step": 1015
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.330434782608696e-06,
+      "loss": 0.0032,
+      "step": 1016
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.3130434782608697e-06,
+      "loss": 0.002,
+      "step": 1017
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.2956521739130434e-06,
+      "loss": 0.0018,
+      "step": 1018
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.2782608695652176e-06,
+      "loss": 0.0023,
+      "step": 1019
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.2608695652173913e-06,
+      "loss": 0.2615,
+      "step": 1020
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.2434782608695654e-06,
+      "loss": 0.333,
+      "step": 1021
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.2260869565217395e-06,
+      "loss": 0.0015,
+      "step": 1022
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.2086956521739133e-06,
+      "loss": 0.3583,
+      "step": 1023
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.191304347826087e-06,
+      "loss": 0.0164,
+      "step": 1024
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.0021,
+      "step": 1025
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.156521739130435e-06,
+      "loss": 0.0053,
+      "step": 1026
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.139130434782609e-06,
+      "loss": 0.0021,
+      "step": 1027
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.1217391304347827e-06,
+      "loss": 0.0026,
+      "step": 1028
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.104347826086957e-06,
+      "loss": 0.0016,
+      "step": 1029
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.0869565217391305e-06,
+      "loss": 0.0247,
+      "step": 1030
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.0695652173913046e-06,
+      "loss": 0.0023,
+      "step": 1031
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.0521739130434784e-06,
+      "loss": 0.0975,
+      "step": 1032
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.0347826086956525e-06,
+      "loss": 0.1697,
+      "step": 1033
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.017391304347826e-06,
+      "loss": 0.2039,
+      "step": 1034
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.0024,
+      "step": 1035
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.982608695652174e-06,
+      "loss": 0.0699,
+      "step": 1036
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1.9652173913043478e-06,
+      "loss": 0.4081,
+      "step": 1037
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1.947826086956522e-06,
+      "loss": 0.3488,
+      "step": 1038
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.930434782608696e-06,
+      "loss": 0.0019,
+      "step": 1039
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.9130434782608697e-06,
+      "loss": 0.0019,
+      "step": 1040
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.8956521739130437e-06,
+      "loss": 0.7029,
+      "step": 1041
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.8782608695652174e-06,
+      "loss": 0.0365,
+      "step": 1042
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.8608695652173913e-06,
+      "loss": 0.0019,
+      "step": 1043
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1.8434782608695654e-06,
+      "loss": 0.2362,
+      "step": 1044
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1.8260869565217394e-06,
+      "loss": 0.036,
+      "step": 1045
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1.8086956521739133e-06,
+      "loss": 0.0013,
+      "step": 1046
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1.7913043478260872e-06,
+      "loss": 0.0024,
+      "step": 1047
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.773913043478261e-06,
+      "loss": 0.7819,
+      "step": 1048
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.7565217391304348e-06,
+      "loss": 0.4246,
+      "step": 1049
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 0.5042,
+      "step": 1050
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.7217391304347827e-06,
+      "loss": 0.2364,
+      "step": 1051
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.7043478260869566e-06,
+      "loss": 0.0129,
+      "step": 1052
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.6869565217391307e-06,
+      "loss": 0.2911,
+      "step": 1053
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.6695652173913047e-06,
+      "loss": 0.0025,
+      "step": 1054
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.6521739130434784e-06,
+      "loss": 0.0024,
+      "step": 1055
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.6347826086956523e-06,
+      "loss": 0.0602,
+      "step": 1056
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.6173913043478262e-06,
+      "loss": 0.0581,
+      "step": 1057
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.004,
+      "step": 1058
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.582608695652174e-06,
+      "loss": 0.2232,
+      "step": 1059
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.565217391304348e-06,
+      "loss": 0.0023,
+      "step": 1060
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.5478260869565217e-06,
+      "loss": 0.0023,
+      "step": 1061
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.5304347826086956e-06,
+      "loss": 0.0029,
+      "step": 1062
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.5130434782608698e-06,
+      "loss": 0.0018,
+      "step": 1063
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.4956521739130437e-06,
+      "loss": 0.0054,
+      "step": 1064
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.4782608695652176e-06,
+      "loss": 0.2773,
+      "step": 1065
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.4608695652173915e-06,
+      "loss": 0.8426,
+      "step": 1066
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.4434782608695652e-06,
+      "loss": 0.0034,
+      "step": 1067
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.4260869565217392e-06,
+      "loss": 0.0032,
+      "step": 1068
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.408695652173913e-06,
+      "loss": 0.3156,
+      "step": 1069
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.391304347826087e-06,
+      "loss": 0.3228,
+      "step": 1070
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.3739130434782611e-06,
+      "loss": 0.1484,
+      "step": 1071
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.356521739130435e-06,
+      "loss": 0.0035,
+      "step": 1072
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.3391304347826088e-06,
+      "loss": 0.3541,
+      "step": 1073
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.3217391304347827e-06,
+      "loss": 0.4588,
+      "step": 1074
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 0.0032,
+      "step": 1075
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.2869565217391305e-06,
+      "loss": 0.5475,
+      "step": 1076
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.2695652173913045e-06,
+      "loss": 0.002,
+      "step": 1077
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.2521739130434784e-06,
+      "loss": 0.0019,
+      "step": 1078
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.2347826086956523e-06,
+      "loss": 0.0035,
+      "step": 1079
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.2173913043478262e-06,
+      "loss": 0.1288,
+      "step": 1080
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.0163,
+      "step": 1081
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.182608695652174e-06,
+      "loss": 0.3959,
+      "step": 1082
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.165217391304348e-06,
+      "loss": 0.0072,
+      "step": 1083
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.1478260869565217e-06,
+      "loss": 0.5006,
+      "step": 1084
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.1304347826086956e-06,
+      "loss": 0.2139,
+      "step": 1085
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.1130434782608698e-06,
+      "loss": 0.0058,
+      "step": 1086
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.0956521739130435e-06,
+      "loss": 0.002,
+      "step": 1087
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.0782608695652174e-06,
+      "loss": 0.0023,
+      "step": 1088
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.0608695652173913e-06,
+      "loss": 0.3971,
+      "step": 1089
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.0434782608695653e-06,
+      "loss": 0.0048,
+      "step": 1090
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.0260869565217392e-06,
+      "loss": 0.354,
+      "step": 1091
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.008695652173913e-06,
+      "loss": 0.0025,
+      "step": 1092
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 9.91304347826087e-07,
+      "loss": 0.3845,
+      "step": 1093
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 9.73913043478261e-07,
+      "loss": 0.4051,
+      "step": 1094
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 9.565217391304349e-07,
+      "loss": 0.4048,
+      "step": 1095
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 9.391304347826087e-07,
+      "loss": 0.0034,
+      "step": 1096
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 9.217391304347827e-07,
+      "loss": 0.0287,
+      "step": 1097
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 9.043478260869566e-07,
+      "loss": 0.0432,
+      "step": 1098
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 8.869565217391305e-07,
+      "loss": 0.1,
+      "step": 1099
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 0.0044,
+      "step": 1100
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 8.521739130434783e-07,
+      "loss": 0.0358,
+      "step": 1101
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 8.347826086956523e-07,
+      "loss": 0.0267,
+      "step": 1102
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 8.173913043478261e-07,
+      "loss": 0.2655,
+      "step": 1103
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.0116,
+      "step": 1104
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.82608695652174e-07,
+      "loss": 0.0128,
+      "step": 1105
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.652173913043478e-07,
+      "loss": 0.0019,
+      "step": 1106
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.478260869565218e-07,
+      "loss": 0.6324,
+      "step": 1107
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.304347826086958e-07,
+      "loss": 0.0022,
+      "step": 1108
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.130434782608696e-07,
+      "loss": 0.0023,
+      "step": 1109
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.956521739130435e-07,
+      "loss": 0.0026,
+      "step": 1110
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.782608695652175e-07,
+      "loss": 0.0029,
+      "step": 1111
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.608695652173913e-07,
+      "loss": 0.4185,
+      "step": 1112
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 6.434782608695653e-07,
+      "loss": 0.0023,
+      "step": 1113
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 6.260869565217392e-07,
+      "loss": 0.0024,
+      "step": 1114
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 6.086956521739131e-07,
+      "loss": 0.0073,
+      "step": 1115
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5.91304347826087e-07,
+      "loss": 0.003,
+      "step": 1116
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 5.739130434782609e-07,
+      "loss": 0.0033,
+      "step": 1117
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 5.565217391304349e-07,
+      "loss": 0.0139,
+      "step": 1118
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 5.391304347826087e-07,
+      "loss": 0.002,
+      "step": 1119
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 5.217391304347826e-07,
+      "loss": 0.0099,
+      "step": 1120
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 5.043478260869565e-07,
+      "loss": 0.0037,
+      "step": 1121
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 4.869565217391305e-07,
+      "loss": 0.0628,
+      "step": 1122
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 4.6956521739130434e-07,
+      "loss": 0.0018,
+      "step": 1123
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 4.521739130434783e-07,
+      "loss": 0.7079,
+      "step": 1124
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 0.1734,
+      "step": 1125
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 4.1739130434782616e-07,
+      "loss": 0.5442,
+      "step": 1126
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.1084,
+      "step": 1127
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 3.826086956521739e-07,
+      "loss": 0.0026,
+      "step": 1128
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 3.652173913043479e-07,
+      "loss": 0.0051,
+      "step": 1129
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 3.4782608695652175e-07,
+      "loss": 0.0029,
+      "step": 1130
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 3.304347826086957e-07,
+      "loss": 0.0028,
+      "step": 1131
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 3.130434782608696e-07,
+      "loss": 0.3837,
+      "step": 1132
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.956521739130435e-07,
+      "loss": 0.0676,
+      "step": 1133
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.7826086956521744e-07,
+      "loss": 0.0023,
+      "step": 1134
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.608695652173913e-07,
+      "loss": 0.0022,
+      "step": 1135
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.4347826086956524e-07,
+      "loss": 0.0044,
+      "step": 1136
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.2608695652173916e-07,
+      "loss": 0.0058,
+      "step": 1137
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.0869565217391308e-07,
+      "loss": 0.3331,
+      "step": 1138
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.9130434782608695e-07,
+      "loss": 0.2143,
+      "step": 1139
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.7391304347826088e-07,
+      "loss": 0.0073,
+      "step": 1140
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.565217391304348e-07,
+      "loss": 0.0035,
+      "step": 1141
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.3913043478260872e-07,
+      "loss": 0.0945,
+      "step": 1142
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.2173913043478262e-07,
+      "loss": 0.002,
+      "step": 1143
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.0434782608695654e-07,
+      "loss": 0.005,
+      "step": 1144
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 8.695652173913044e-08,
+      "loss": 0.002,
+      "step": 1145
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 6.956521739130436e-08,
+      "loss": 0.0028,
+      "step": 1146
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 5.217391304347827e-08,
+      "loss": 0.3331,
+      "step": 1147
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 3.478260869565218e-08,
+      "loss": 0.0026,
+      "step": 1148
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.739130434782609e-08,
+      "loss": 0.3649,
+      "step": 1149
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 0.0046,
+      "step": 1150
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8995098039215687,
+      "eval_combined_score": 0.9130822722648988,
+      "eval_f1": 0.9266547406082289,
+      "eval_loss": 0.5354248285293579,
+      "eval_runtime": 0.8349,
+      "eval_samples_per_second": 488.668,
+      "eval_steps_per_second": 31.141,
+      "step": 1150
+    },
+    {
+      "epoch": 5.0,
+      "step": 1150,
+      "total_flos": 1755670347110400.0,
+      "train_loss": 0.2699521788773532,
+      "train_runtime": 359.8356,
+      "train_samples_per_second": 50.968,
+      "train_steps_per_second": 3.196
+    }
+  ],
+  "max_steps": 1150,
+  "num_train_epochs": 5,
+  "total_flos": 1755670347110400.0,
+  "trial_name": null,
+  "trial_params": null
+}