diff --git "a/checkpoint-2646/trainer_state.json" "b/checkpoint-2646/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2646/trainer_state.json"
@@ -0,0 +1,15955 @@
+{
+  "best_metric": 0.40209120512008667,
+  "best_model_checkpoint": "./disaster-tweet-distilbert-classification/checkpoint-378",
+  "epoch": 7.0,
+  "global_step": 2646,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.7503,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7789,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.7344,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7709,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-06,
+      "loss": 0.6884,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-06,
+      "loss": 0.7087,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.6655,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6978,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9e-06,
+      "loss": 0.7435,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.719,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7129,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7249,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.7436,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.6886,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5e-05,
+      "loss": 0.702,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7105,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.6709,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6767,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6784,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 0.7016,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6308,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.645,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6845,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6891,
+      "step": 24
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5914,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.6916,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.5884,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.6587,
+      "step": 28
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9e-05,
+      "loss": 0.6008,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3e-05,
+      "loss": 0.6717,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.1e-05,
+      "loss": 0.5795,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6358,
+      "step": 32
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.3e-05,
+      "loss": 0.7508,
+      "step": 33
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.6601,
+      "step": 34
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.5e-05,
+      "loss": 0.6573,
+      "step": 35
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.6e-05,
+      "loss": 0.5695,
+      "step": 36
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.7e-05,
+      "loss": 0.5535,
+      "step": 37
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.8e-05,
+      "loss": 0.5813,
+      "step": 38
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.5224,
+      "step": 39
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4e-05,
+      "loss": 0.4757,
+      "step": 40
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5529,
+      "step": 41
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.2e-05,
+      "loss": 0.4964,
+      "step": 42
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.3e-05,
+      "loss": 0.4565,
+      "step": 43
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.5848,
+      "step": 44
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5e-05,
+      "loss": 0.7333,
+      "step": 45
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.5224,
+      "step": 46
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7e-05,
+      "loss": 0.4826,
+      "step": 47
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8e-05,
+      "loss": 0.4328,
+      "step": 48
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9e-05,
+      "loss": 0.2546,
+      "step": 49
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2636,
+      "step": 50
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.998659517426274e-05,
+      "loss": 0.4116,
+      "step": 51
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.997319034852547e-05,
+      "loss": 0.3191,
+      "step": 52
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.995978552278821e-05,
+      "loss": 0.2822,
+      "step": 53
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.994638069705094e-05,
+      "loss": 0.8821,
+      "step": 54
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.993297587131368e-05,
+      "loss": 0.483,
+      "step": 55
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9919571045576406e-05,
+      "loss": 0.7164,
+      "step": 56
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.990616621983915e-05,
+      "loss": 0.4161,
+      "step": 57
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.989276139410188e-05,
+      "loss": 0.2668,
+      "step": 58
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.987935656836462e-05,
+      "loss": 0.5255,
+      "step": 59
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.986595174262735e-05,
+      "loss": 0.3784,
+      "step": 60
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.985254691689008e-05,
+      "loss": 0.5065,
+      "step": 61
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.983914209115282e-05,
+      "loss": 0.1988,
+      "step": 62
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.982573726541555e-05,
+      "loss": 0.4362,
+      "step": 63
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.981233243967829e-05,
+      "loss": 0.6619,
+      "step": 64
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.979892761394102e-05,
+      "loss": 0.3217,
+      "step": 65
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.978552278820375e-05,
+      "loss": 0.2967,
+      "step": 66
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.977211796246649e-05,
+      "loss": 0.2429,
+      "step": 67
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.975871313672922e-05,
+      "loss": 0.6642,
+      "step": 68
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.974530831099196e-05,
+      "loss": 0.56,
+      "step": 69
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.973190348525469e-05,
+      "loss": 1.2979,
+      "step": 70
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9718498659517427e-05,
+      "loss": 0.5287,
+      "step": 71
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.970509383378016e-05,
+      "loss": 0.4684,
+      "step": 72
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.96916890080429e-05,
+      "loss": 0.283,
+      "step": 73
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.967828418230563e-05,
+      "loss": 0.6818,
+      "step": 74
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.966487935656837e-05,
+      "loss": 0.6141,
+      "step": 75
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.96514745308311e-05,
+      "loss": 0.5046,
+      "step": 76
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.963806970509384e-05,
+      "loss": 0.5266,
+      "step": 77
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.962466487935657e-05,
+      "loss": 0.5944,
+      "step": 78
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.961126005361931e-05,
+      "loss": 0.5631,
+      "step": 79
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.959785522788204e-05,
+      "loss": 0.4791,
+      "step": 80
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.958445040214477e-05,
+      "loss": 0.5645,
+      "step": 81
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.957104557640751e-05,
+      "loss": 0.4349,
+      "step": 82
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.955764075067024e-05,
+      "loss": 0.3865,
+      "step": 83
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.954423592493298e-05,
+      "loss": 0.486,
+      "step": 84
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.953083109919571e-05,
+      "loss": 0.2179,
+      "step": 85
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.951742627345845e-05,
+      "loss": 0.3896,
+      "step": 86
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.950402144772118e-05,
+      "loss": 0.4247,
+      "step": 87
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.949061662198392e-05,
+      "loss": 0.4906,
+      "step": 88
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.947721179624665e-05,
+      "loss": 0.4483,
+      "step": 89
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.946380697050939e-05,
+      "loss": 0.557,
+      "step": 90
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9450402144772116e-05,
+      "loss": 0.7521,
+      "step": 91
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.943699731903486e-05,
+      "loss": 0.3103,
+      "step": 92
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9423592493297586e-05,
+      "loss": 0.757,
+      "step": 93
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.941018766756033e-05,
+      "loss": 0.8248,
+      "step": 94
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9396782841823056e-05,
+      "loss": 0.4591,
+      "step": 95
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.938337801608579e-05,
+      "loss": 0.3912,
+      "step": 96
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9369973190348526e-05,
+      "loss": 0.5289,
+      "step": 97
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.935656836461126e-05,
+      "loss": 0.3264,
+      "step": 98
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9343163538874e-05,
+      "loss": 0.2947,
+      "step": 99
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.932975871313673e-05,
+      "loss": 0.2647,
+      "step": 100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.931635388739946e-05,
+      "loss": 0.3691,
+      "step": 101
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.93029490616622e-05,
+      "loss": 0.4796,
+      "step": 102
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.928954423592493e-05,
+      "loss": 0.4827,
+      "step": 103
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.927613941018767e-05,
+      "loss": 0.2672,
+      "step": 104
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.92627345844504e-05,
+      "loss": 0.7456,
+      "step": 105
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.9249329758713136e-05,
+      "loss": 0.5206,
+      "step": 106
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.923592493297587e-05,
+      "loss": 0.3576,
+      "step": 107
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.9222520107238606e-05,
+      "loss": 0.2596,
+      "step": 108
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.920911528150134e-05,
+      "loss": 0.4115,
+      "step": 109
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.9195710455764076e-05,
+      "loss": 0.3481,
+      "step": 110
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.918230563002681e-05,
+      "loss": 0.4387,
+      "step": 111
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.916890080428955e-05,
+      "loss": 0.5023,
+      "step": 112
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.915549597855228e-05,
+      "loss": 0.5916,
+      "step": 113
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.914209115281502e-05,
+      "loss": 0.5467,
+      "step": 114
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.912868632707775e-05,
+      "loss": 0.5631,
+      "step": 115
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.911528150134049e-05,
+      "loss": 0.5512,
+      "step": 116
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.910187667560322e-05,
+      "loss": 0.5546,
+      "step": 117
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.908847184986595e-05,
+      "loss": 0.4209,
+      "step": 118
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.907506702412869e-05,
+      "loss": 0.6064,
+      "step": 119
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.906166219839142e-05,
+      "loss": 0.5301,
+      "step": 120
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.904825737265416e-05,
+      "loss": 0.436,
+      "step": 121
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.903485254691689e-05,
+      "loss": 0.726,
+      "step": 122
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9021447721179626e-05,
+      "loss": 0.5288,
+      "step": 123
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.900804289544236e-05,
+      "loss": 0.4887,
+      "step": 124
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8994638069705097e-05,
+      "loss": 0.3752,
+      "step": 125
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.898123324396783e-05,
+      "loss": 0.4895,
+      "step": 126
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.896782841823057e-05,
+      "loss": 0.5046,
+      "step": 127
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8954423592493295e-05,
+      "loss": 0.3953,
+      "step": 128
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.894101876675604e-05,
+      "loss": 0.2015,
+      "step": 129
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8927613941018765e-05,
+      "loss": 0.5165,
+      "step": 130
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.891420911528151e-05,
+      "loss": 0.4237,
+      "step": 131
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8900804289544236e-05,
+      "loss": 0.239,
+      "step": 132
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.888739946380697e-05,
+      "loss": 0.5515,
+      "step": 133
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8873994638069706e-05,
+      "loss": 0.303,
+      "step": 134
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.886058981233244e-05,
+      "loss": 0.2867,
+      "step": 135
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8847184986595176e-05,
+      "loss": 0.6756,
+      "step": 136
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.883378016085791e-05,
+      "loss": 0.4996,
+      "step": 137
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8820375335120646e-05,
+      "loss": 0.2798,
+      "step": 138
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.880697050938338e-05,
+      "loss": 0.8877,
+      "step": 139
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.879356568364612e-05,
+      "loss": 0.5022,
+      "step": 140
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.878016085790885e-05,
+      "loss": 0.2177,
+      "step": 141
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.876675603217159e-05,
+      "loss": 0.8708,
+      "step": 142
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.8753351206434315e-05,
+      "loss": 0.719,
+      "step": 143
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.873994638069706e-05,
+      "loss": 0.562,
+      "step": 144
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.8726541554959786e-05,
+      "loss": 0.492,
+      "step": 145
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.871313672922253e-05,
+      "loss": 0.4637,
+      "step": 146
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8699731903485256e-05,
+      "loss": 0.4132,
+      "step": 147
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.868632707774799e-05,
+      "loss": 0.2889,
+      "step": 148
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8672922252010726e-05,
+      "loss": 0.2213,
+      "step": 149
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.865951742627346e-05,
+      "loss": 0.9268,
+      "step": 150
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.8646112600536196e-05,
+      "loss": 0.2852,
+      "step": 151
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.863270777479893e-05,
+      "loss": 0.4599,
+      "step": 152
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.861930294906166e-05,
+      "loss": 0.1913,
+      "step": 153
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.86058981233244e-05,
+      "loss": 0.4488,
+      "step": 154
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.859249329758713e-05,
+      "loss": 0.9022,
+      "step": 155
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.857908847184987e-05,
+      "loss": 0.5221,
+      "step": 156
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.85656836461126e-05,
+      "loss": 0.2394,
+      "step": 157
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8552278820375336e-05,
+      "loss": 0.3332,
+      "step": 158
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.853887399463807e-05,
+      "loss": 0.4015,
+      "step": 159
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8525469168900806e-05,
+      "loss": 0.4461,
+      "step": 160
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.851206434316354e-05,
+      "loss": 0.337,
+      "step": 161
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.8498659517426276e-05,
+      "loss": 0.4908,
+      "step": 162
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.848525469168901e-05,
+      "loss": 0.526,
+      "step": 163
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.8471849865951746e-05,
+      "loss": 0.5262,
+      "step": 164
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.845844504021448e-05,
+      "loss": 0.6818,
+      "step": 165
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.8445040214477217e-05,
+      "loss": 0.3154,
+      "step": 166
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.843163538873995e-05,
+      "loss": 0.5963,
+      "step": 167
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.841823056300268e-05,
+      "loss": 0.4451,
+      "step": 168
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.840482573726542e-05,
+      "loss": 0.5969,
+      "step": 169
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.839142091152815e-05,
+      "loss": 0.438,
+      "step": 170
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.837801608579089e-05,
+      "loss": 0.4827,
+      "step": 171
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.836461126005362e-05,
+      "loss": 0.2029,
+      "step": 172
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.8351206434316356e-05,
+      "loss": 0.5195,
+      "step": 173
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.833780160857909e-05,
+      "loss": 0.517,
+      "step": 174
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.8324396782841826e-05,
+      "loss": 0.5532,
+      "step": 175
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.831099195710456e-05,
+      "loss": 0.4198,
+      "step": 176
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.8297587131367296e-05,
+      "loss": 0.8386,
+      "step": 177
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.8284182305630025e-05,
+      "loss": 0.575,
+      "step": 178
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.8270777479892766e-05,
+      "loss": 0.6156,
+      "step": 179
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.8257372654155495e-05,
+      "loss": 0.7044,
+      "step": 180
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.824396782841824e-05,
+      "loss": 0.5712,
+      "step": 181
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.8230563002680965e-05,
+      "loss": 0.34,
+      "step": 182
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.82171581769437e-05,
+      "loss": 0.5773,
+      "step": 183
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.8203753351206435e-05,
+      "loss": 0.546,
+      "step": 184
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.819034852546917e-05,
+      "loss": 0.3955,
+      "step": 185
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.8176943699731906e-05,
+      "loss": 0.5921,
+      "step": 186
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.816353887399464e-05,
+      "loss": 0.3108,
+      "step": 187
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.8150134048257376e-05,
+      "loss": 0.5469,
+      "step": 188
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.813672922252011e-05,
+      "loss": 0.64,
+      "step": 189
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.8123324396782846e-05,
+      "loss": 0.5153,
+      "step": 190
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.810991957104558e-05,
+      "loss": 0.4719,
+      "step": 191
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.8096514745308316e-05,
+      "loss": 0.52,
+      "step": 192
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.8083109919571045e-05,
+      "loss": 0.5114,
+      "step": 193
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.806970509383379e-05,
+      "loss": 0.5469,
+      "step": 194
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.8056300268096515e-05,
+      "loss": 0.3435,
+      "step": 195
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.804289544235926e-05,
+      "loss": 0.6469,
+      "step": 196
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.8029490616621985e-05,
+      "loss": 0.6595,
+      "step": 197
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.801608579088472e-05,
+      "loss": 0.5503,
+      "step": 198
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8002680965147456e-05,
+      "loss": 0.3799,
+      "step": 199
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.798927613941019e-05,
+      "loss": 0.417,
+      "step": 200
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.7975871313672926e-05,
+      "loss": 0.5281,
+      "step": 201
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.796246648793566e-05,
+      "loss": 0.3439,
+      "step": 202
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.794906166219839e-05,
+      "loss": 0.5777,
+      "step": 203
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.793565683646113e-05,
+      "loss": 0.5286,
+      "step": 204
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.792225201072386e-05,
+      "loss": 0.4302,
+      "step": 205
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.79088471849866e-05,
+      "loss": 0.5413,
+      "step": 206
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.789544235924933e-05,
+      "loss": 0.3087,
+      "step": 207
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.7882037533512065e-05,
+      "loss": 0.4385,
+      "step": 208
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.78686327077748e-05,
+      "loss": 0.7137,
+      "step": 209
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.7855227882037535e-05,
+      "loss": 0.6278,
+      "step": 210
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.784182305630027e-05,
+      "loss": 0.2832,
+      "step": 211
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.7828418230563005e-05,
+      "loss": 0.5899,
+      "step": 212
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.7815013404825734e-05,
+      "loss": 0.4421,
+      "step": 213
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.7801608579088476e-05,
+      "loss": 0.3411,
+      "step": 214
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.7788203753351204e-05,
+      "loss": 0.4236,
+      "step": 215
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.7774798927613946e-05,
+      "loss": 0.4444,
+      "step": 216
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.7761394101876674e-05,
+      "loss": 0.4765,
+      "step": 217
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.774798927613941e-05,
+      "loss": 0.2453,
+      "step": 218
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.7734584450402145e-05,
+      "loss": 0.2765,
+      "step": 219
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.772117962466488e-05,
+      "loss": 0.2075,
+      "step": 220
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.7707774798927615e-05,
+      "loss": 0.5905,
+      "step": 221
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.769436997319035e-05,
+      "loss": 0.3457,
+      "step": 222
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.7680965147453085e-05,
+      "loss": 0.5986,
+      "step": 223
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.766756032171582e-05,
+      "loss": 0.3881,
+      "step": 224
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.7654155495978555e-05,
+      "loss": 0.5655,
+      "step": 225
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.764075067024129e-05,
+      "loss": 0.6156,
+      "step": 226
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.7627345844504026e-05,
+      "loss": 0.4784,
+      "step": 227
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.7613941018766754e-05,
+      "loss": 0.5195,
+      "step": 228
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7600536193029496e-05,
+      "loss": 0.66,
+      "step": 229
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7587131367292224e-05,
+      "loss": 0.4104,
+      "step": 230
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7573726541554966e-05,
+      "loss": 0.3478,
+      "step": 231
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.7560321715817695e-05,
+      "loss": 0.2037,
+      "step": 232
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.754691689008043e-05,
+      "loss": 0.75,
+      "step": 233
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7533512064343165e-05,
+      "loss": 0.4237,
+      "step": 234
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.75201072386059e-05,
+      "loss": 0.2372,
+      "step": 235
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.7506702412868635e-05,
+      "loss": 0.7874,
+      "step": 236
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.749329758713137e-05,
+      "loss": 0.5751,
+      "step": 237
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.7479892761394105e-05,
+      "loss": 0.5801,
+      "step": 238
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.746648793565684e-05,
+      "loss": 0.4983,
+      "step": 239
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.745308310991957e-05,
+      "loss": 0.4215,
+      "step": 240
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.743967828418231e-05,
+      "loss": 0.3655,
+      "step": 241
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.742627345844504e-05,
+      "loss": 0.523,
+      "step": 242
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.741286863270778e-05,
+      "loss": 0.5952,
+      "step": 243
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.739946380697051e-05,
+      "loss": 0.4226,
+      "step": 244
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.7386058981233244e-05,
+      "loss": 0.316,
+      "step": 245
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.737265415549598e-05,
+      "loss": 0.5096,
+      "step": 246
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.7359249329758715e-05,
+      "loss": 0.435,
+      "step": 247
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.734584450402145e-05,
+      "loss": 0.7516,
+      "step": 248
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.7332439678284185e-05,
+      "loss": 0.531,
+      "step": 249
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.731903485254692e-05,
+      "loss": 0.4096,
+      "step": 250
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.7305630026809655e-05,
+      "loss": 0.387,
+      "step": 251
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.729222520107239e-05,
+      "loss": 0.5468,
+      "step": 252
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.7278820375335125e-05,
+      "loss": 0.4613,
+      "step": 253
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.726541554959786e-05,
+      "loss": 0.8437,
+      "step": 254
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.725201072386059e-05,
+      "loss": 0.522,
+      "step": 255
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.723860589812333e-05,
+      "loss": 0.3922,
+      "step": 256
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.722520107238606e-05,
+      "loss": 0.5114,
+      "step": 257
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.72117962466488e-05,
+      "loss": 0.6148,
+      "step": 258
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.719839142091153e-05,
+      "loss": 0.4578,
+      "step": 259
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.7184986595174265e-05,
+      "loss": 0.6286,
+      "step": 260
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.7171581769437e-05,
+      "loss": 0.5883,
+      "step": 261
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.7158176943699735e-05,
+      "loss": 0.5634,
+      "step": 262
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.714477211796247e-05,
+      "loss": 0.4085,
+      "step": 263
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.7131367292225205e-05,
+      "loss": 0.2988,
+      "step": 264
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.7117962466487934e-05,
+      "loss": 0.6353,
+      "step": 265
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.7104557640750675e-05,
+      "loss": 0.4598,
+      "step": 266
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.7091152815013404e-05,
+      "loss": 0.5072,
+      "step": 267
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.7077747989276146e-05,
+      "loss": 0.49,
+      "step": 268
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.7064343163538874e-05,
+      "loss": 0.7225,
+      "step": 269
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.705093833780161e-05,
+      "loss": 0.5332,
+      "step": 270
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.7037533512064344e-05,
+      "loss": 0.6064,
+      "step": 271
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.702412868632708e-05,
+      "loss": 0.3518,
+      "step": 272
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.7010723860589815e-05,
+      "loss": 0.3673,
+      "step": 273
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.699731903485255e-05,
+      "loss": 0.4688,
+      "step": 274
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.6983914209115285e-05,
+      "loss": 0.5389,
+      "step": 275
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.697050938337802e-05,
+      "loss": 0.374,
+      "step": 276
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.6957104557640755e-05,
+      "loss": 0.475,
+      "step": 277
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.694369973190349e-05,
+      "loss": 0.5397,
+      "step": 278
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.6930294906166225e-05,
+      "loss": 0.3821,
+      "step": 279
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.6916890080428954e-05,
+      "loss": 0.3372,
+      "step": 280
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.6903485254691696e-05,
+      "loss": 0.6652,
+      "step": 281
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.6890080428954424e-05,
+      "loss": 0.2894,
+      "step": 282
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.6876675603217166e-05,
+      "loss": 0.5639,
+      "step": 283
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.6863270777479894e-05,
+      "loss": 0.353,
+      "step": 284
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.684986595174263e-05,
+      "loss": 0.2932,
+      "step": 285
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.6836461126005364e-05,
+      "loss": 0.467,
+      "step": 286
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.68230563002681e-05,
+      "loss": 0.4732,
+      "step": 287
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.6809651474530835e-05,
+      "loss": 0.1808,
+      "step": 288
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.679624664879357e-05,
+      "loss": 0.6031,
+      "step": 289
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.67828418230563e-05,
+      "loss": 0.2555,
+      "step": 290
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.676943699731904e-05,
+      "loss": 0.4041,
+      "step": 291
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.675603217158177e-05,
+      "loss": 0.7822,
+      "step": 292
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.674262734584451e-05,
+      "loss": 0.138,
+      "step": 293
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.672922252010724e-05,
+      "loss": 0.2746,
+      "step": 294
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.6715817694369974e-05,
+      "loss": 0.5835,
+      "step": 295
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.670241286863271e-05,
+      "loss": 0.2367,
+      "step": 296
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.6689008042895444e-05,
+      "loss": 0.3247,
+      "step": 297
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.667560321715818e-05,
+      "loss": 0.306,
+      "step": 298
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.6662198391420914e-05,
+      "loss": 0.2825,
+      "step": 299
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.664879356568364e-05,
+      "loss": 0.6102,
+      "step": 300
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.6635388739946385e-05,
+      "loss": 0.3613,
+      "step": 301
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.662198391420911e-05,
+      "loss": 0.6327,
+      "step": 302
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.6608579088471855e-05,
+      "loss": 0.4059,
+      "step": 303
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.659517426273458e-05,
+      "loss": 0.4027,
+      "step": 304
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.658176943699732e-05,
+      "loss": 0.9133,
+      "step": 305
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.6568364611260054e-05,
+      "loss": 0.1869,
+      "step": 306
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.655495978552279e-05,
+      "loss": 0.3987,
+      "step": 307
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.6541554959785524e-05,
+      "loss": 0.6114,
+      "step": 308
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.652815013404826e-05,
+      "loss": 0.4406,
+      "step": 309
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.6514745308310994e-05,
+      "loss": 0.2954,
+      "step": 310
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.650134048257373e-05,
+      "loss": 0.5067,
+      "step": 311
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.6487935656836464e-05,
+      "loss": 0.3985,
+      "step": 312
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.64745308310992e-05,
+      "loss": 0.3756,
+      "step": 313
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.6461126005361935e-05,
+      "loss": 0.2618,
+      "step": 314
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.644772117962466e-05,
+      "loss": 0.4992,
+      "step": 315
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.6434316353887405e-05,
+      "loss": 0.8224,
+      "step": 316
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.642091152815013e-05,
+      "loss": 0.3425,
+      "step": 317
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.6407506702412875e-05,
+      "loss": 0.4062,
+      "step": 318
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.6394101876675603e-05,
+      "loss": 0.4748,
+      "step": 319
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.638069705093834e-05,
+      "loss": 0.6857,
+      "step": 320
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.6367292225201074e-05,
+      "loss": 0.5368,
+      "step": 321
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.635388739946381e-05,
+      "loss": 0.5571,
+      "step": 322
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.6340482573726544e-05,
+      "loss": 0.3045,
+      "step": 323
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.632707774798928e-05,
+      "loss": 0.3189,
+      "step": 324
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.631367292225201e-05,
+      "loss": 0.3033,
+      "step": 325
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.630026809651475e-05,
+      "loss": 0.4659,
+      "step": 326
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.628686327077748e-05,
+      "loss": 0.4058,
+      "step": 327
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.627345844504022e-05,
+      "loss": 0.6745,
+      "step": 328
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.626005361930295e-05,
+      "loss": 0.3259,
+      "step": 329
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.624664879356568e-05,
+      "loss": 0.5126,
+      "step": 330
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.623324396782842e-05,
+      "loss": 0.2759,
+      "step": 331
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.621983914209115e-05,
+      "loss": 0.2512,
+      "step": 332
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.620643431635389e-05,
+      "loss": 0.3046,
+      "step": 333
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.6193029490616624e-05,
+      "loss": 0.3931,
+      "step": 334
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.617962466487936e-05,
+      "loss": 0.4838,
+      "step": 335
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.6166219839142094e-05,
+      "loss": 0.2925,
+      "step": 336
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.615281501340483e-05,
+      "loss": 0.4481,
+      "step": 337
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.6139410187667564e-05,
+      "loss": 0.4528,
+      "step": 338
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.61260053619303e-05,
+      "loss": 0.2934,
+      "step": 339
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.611260053619303e-05,
+      "loss": 0.609,
+      "step": 340
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.609919571045577e-05,
+      "loss": 0.8988,
+      "step": 341
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.60857908847185e-05,
+      "loss": 1.1222,
+      "step": 342
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.607238605898124e-05,
+      "loss": 0.3265,
+      "step": 343
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.605898123324397e-05,
+      "loss": 0.4722,
+      "step": 344
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.60455764075067e-05,
+      "loss": 0.2791,
+      "step": 345
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.603217158176944e-05,
+      "loss": 0.4183,
+      "step": 346
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.6018766756032174e-05,
+      "loss": 0.5323,
+      "step": 347
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.600536193029491e-05,
+      "loss": 0.6108,
+      "step": 348
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.5991957104557644e-05,
+      "loss": 0.2875,
+      "step": 349
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.597855227882037e-05,
+      "loss": 0.6642,
+      "step": 350
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.5965147453083114e-05,
+      "loss": 0.5244,
+      "step": 351
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.595174262734584e-05,
+      "loss": 0.5562,
+      "step": 352
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.5938337801608584e-05,
+      "loss": 0.3972,
+      "step": 353
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.592493297587131e-05,
+      "loss": 0.4532,
+      "step": 354
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.591152815013405e-05,
+      "loss": 0.3368,
+      "step": 355
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.589812332439678e-05,
+      "loss": 0.171,
+      "step": 356
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.588471849865952e-05,
+      "loss": 0.4036,
+      "step": 357
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.587131367292225e-05,
+      "loss": 0.4305,
+      "step": 358
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.585790884718499e-05,
+      "loss": 0.2643,
+      "step": 359
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.5844504021447723e-05,
+      "loss": 0.3782,
+      "step": 360
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.583109919571046e-05,
+      "loss": 0.1673,
+      "step": 361
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.5817694369973194e-05,
+      "loss": 0.3969,
+      "step": 362
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.580428954423593e-05,
+      "loss": 0.3249,
+      "step": 363
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.5790884718498664e-05,
+      "loss": 0.1656,
+      "step": 364
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.57774798927614e-05,
+      "loss": 0.4551,
+      "step": 365
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.5764075067024134e-05,
+      "loss": 0.6075,
+      "step": 366
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.575067024128686e-05,
+      "loss": 0.4699,
+      "step": 367
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.5737265415549605e-05,
+      "loss": 0.5752,
+      "step": 368
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.572386058981233e-05,
+      "loss": 0.3114,
+      "step": 369
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.5710455764075075e-05,
+      "loss": 0.7407,
+      "step": 370
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.56970509383378e-05,
+      "loss": 0.6427,
+      "step": 371
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.568364611260054e-05,
+      "loss": 0.5021,
+      "step": 372
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.5670241286863273e-05,
+      "loss": 0.4209,
+      "step": 373
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.565683646112601e-05,
+      "loss": 0.5957,
+      "step": 374
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.5643431635388744e-05,
+      "loss": 0.495,
+      "step": 375
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.563002680965148e-05,
+      "loss": 0.3101,
+      "step": 376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.561662198391421e-05,
+      "loss": 0.472,
+      "step": 377
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.560321715817695e-05,
+      "loss": 0.4607,
+      "step": 378
+    },
+    {
+      "epoch": 1.0,
+      "eval_f1": 0.7914963205233032,
+      "eval_loss": 0.40209120512008667,
+      "eval_runtime": 2.1585,
+      "eval_samples_per_second": 700.949,
+      "eval_steps_per_second": 44.012,
+      "step": 378
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.558981233243968e-05,
+      "loss": 0.2422,
+      "step": 379
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557640750670242e-05,
+      "loss": 0.2074,
+      "step": 380
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556300268096515e-05,
+      "loss": 0.4469,
+      "step": 381
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554959785522788e-05,
+      "loss": 0.3121,
+      "step": 382
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553619302949062e-05,
+      "loss": 0.3634,
+      "step": 383
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.552278820375335e-05,
+      "loss": 0.2139,
+      "step": 384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.550938337801609e-05,
+      "loss": 0.5273,
+      "step": 385
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549597855227882e-05,
+      "loss": 0.1228,
+      "step": 386
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548257372654156e-05,
+      "loss": 0.2459,
+      "step": 387
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.5469168900804294e-05,
+      "loss": 0.443,
+      "step": 388
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.545576407506703e-05,
+      "loss": 0.364,
+      "step": 389
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.5442359249329764e-05,
+      "loss": 0.4071,
+      "step": 390
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.54289544235925e-05,
+      "loss": 0.3876,
+      "step": 391
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.541554959785523e-05,
+      "loss": 0.5239,
+      "step": 392
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.540214477211797e-05,
+      "loss": 0.4548,
+      "step": 393
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.53887399463807e-05,
+      "loss": 0.3588,
+      "step": 394
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.537533512064344e-05,
+      "loss": 0.2693,
+      "step": 395
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.536193029490617e-05,
+      "loss": 0.3773,
+      "step": 396
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.53485254691689e-05,
+      "loss": 0.3982,
+      "step": 397
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.533512064343164e-05,
+      "loss": 0.2074,
+      "step": 398
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.532171581769437e-05,
+      "loss": 0.3322,
+      "step": 399
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.530831099195711e-05,
+      "loss": 0.118,
+      "step": 400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.5294906166219844e-05,
+      "loss": 0.6216,
+      "step": 401
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.528150134048257e-05,
+      "loss": 0.4028,
+      "step": 402
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.5268096514745314e-05,
+      "loss": 0.3179,
+      "step": 403
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.525469168900804e-05,
+      "loss": 0.2815,
+      "step": 404
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.5241286863270784e-05,
+      "loss": 0.2366,
+      "step": 405
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.522788203753351e-05,
+      "loss": 0.14,
+      "step": 406
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.521447721179625e-05,
+      "loss": 0.1255,
+      "step": 407
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.520107238605898e-05,
+      "loss": 0.1482,
+      "step": 408
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.518766756032172e-05,
+      "loss": 0.665,
+      "step": 409
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.517426273458445e-05,
+      "loss": 0.2535,
+      "step": 410
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.516085790884719e-05,
+      "loss": 0.1716,
+      "step": 411
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.5147453083109916e-05,
+      "loss": 0.0765,
+      "step": 412
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.513404825737266e-05,
+      "loss": 0.1238,
+      "step": 413
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.512064343163539e-05,
+      "loss": 0.3877,
+      "step": 414
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.510723860589813e-05,
+      "loss": 0.206,
+      "step": 415
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.509383378016086e-05,
+      "loss": 0.1822,
+      "step": 416
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.508042895442359e-05,
+      "loss": 0.2294,
+      "step": 417
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.506702412868633e-05,
+      "loss": 0.7083,
+      "step": 418
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.505361930294906e-05,
+      "loss": 0.325,
+      "step": 419
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.50402144772118e-05,
+      "loss": 0.2378,
+      "step": 420
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.502680965147453e-05,
+      "loss": 0.2494,
+      "step": 421
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.501340482573727e-05,
+      "loss": 0.3812,
+      "step": 422
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1491,
+      "step": 423
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.498659517426274e-05,
+      "loss": 0.0833,
+      "step": 424
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.497319034852547e-05,
+      "loss": 0.4637,
+      "step": 425
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.495978552278821e-05,
+      "loss": 0.2594,
+      "step": 426
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.4946380697050937e-05,
+      "loss": 0.4136,
+      "step": 427
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.493297587131368e-05,
+      "loss": 0.4962,
+      "step": 428
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.491957104557641e-05,
+      "loss": 0.0842,
+      "step": 429
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.490616621983915e-05,
+      "loss": 0.4761,
+      "step": 430
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.489276139410188e-05,
+      "loss": 0.4351,
+      "step": 431
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.487935656836461e-05,
+      "loss": 0.4226,
+      "step": 432
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.486595174262735e-05,
+      "loss": 0.5714,
+      "step": 433
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.485254691689008e-05,
+      "loss": 0.2034,
+      "step": 434
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.483914209115282e-05,
+      "loss": 0.3354,
+      "step": 435
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.482573726541555e-05,
+      "loss": 0.7503,
+      "step": 436
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.481233243967828e-05,
+      "loss": 0.1542,
+      "step": 437
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.479892761394102e-05,
+      "loss": 0.3067,
+      "step": 438
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.478552278820375e-05,
+      "loss": 0.4134,
+      "step": 439
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.477211796246649e-05,
+      "loss": 0.1458,
+      "step": 440
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.475871313672922e-05,
+      "loss": 0.2814,
+      "step": 441
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.474530831099196e-05,
+      "loss": 0.1751,
+      "step": 442
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.473190348525469e-05,
+      "loss": 0.4144,
+      "step": 443
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.471849865951743e-05,
+      "loss": 0.3275,
+      "step": 444
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.470509383378016e-05,
+      "loss": 0.3083,
+      "step": 445
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.46916890080429e-05,
+      "loss": 0.4185,
+      "step": 446
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.467828418230563e-05,
+      "loss": 0.3181,
+      "step": 447
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.466487935656837e-05,
+      "loss": 0.438,
+      "step": 448
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.46514745308311e-05,
+      "loss": 0.4509,
+      "step": 449
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.463806970509384e-05,
+      "loss": 0.4597,
+      "step": 450
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.462466487935657e-05,
+      "loss": 0.3123,
+      "step": 451
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.46112600536193e-05,
+      "loss": 0.376,
+      "step": 452
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.459785522788204e-05,
+      "loss": 0.2392,
+      "step": 453
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.458445040214477e-05,
+      "loss": 0.4451,
+      "step": 454
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.4571045576407513e-05,
+      "loss": 0.2436,
+      "step": 455
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.455764075067024e-05,
+      "loss": 0.2944,
+      "step": 456
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.454423592493298e-05,
+      "loss": 0.323,
+      "step": 457
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.453083109919571e-05,
+      "loss": 0.1966,
+      "step": 458
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.451742627345845e-05,
+      "loss": 0.273,
+      "step": 459
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.450402144772118e-05,
+      "loss": 0.3498,
+      "step": 460
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.449061662198392e-05,
+      "loss": 0.458,
+      "step": 461
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.4477211796246646e-05,
+      "loss": 0.1607,
+      "step": 462
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.446380697050939e-05,
+      "loss": 0.5712,
+      "step": 463
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.4450402144772116e-05,
+      "loss": 0.0908,
+      "step": 464
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.443699731903486e-05,
+      "loss": 0.2625,
+      "step": 465
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.4423592493297586e-05,
+      "loss": 0.368,
+      "step": 466
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.441018766756032e-05,
+      "loss": 0.5096,
+      "step": 467
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.4396782841823057e-05,
+      "loss": 0.209,
+      "step": 468
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.438337801608579e-05,
+      "loss": 0.5235,
+      "step": 469
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.436997319034853e-05,
+      "loss": 0.5116,
+      "step": 470
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.435656836461126e-05,
+      "loss": 0.49,
+      "step": 471
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.4343163538874e-05,
+      "loss": 0.4409,
+      "step": 472
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.432975871313673e-05,
+      "loss": 0.1324,
+      "step": 473
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.431635388739947e-05,
+      "loss": 0.1257,
+      "step": 474
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.43029490616622e-05,
+      "loss": 0.3844,
+      "step": 475
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.428954423592494e-05,
+      "loss": 0.2942,
+      "step": 476
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.4276139410187666e-05,
+      "loss": 0.316,
+      "step": 477
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.426273458445041e-05,
+      "loss": 0.3724,
+      "step": 478
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.4249329758713136e-05,
+      "loss": 0.4498,
+      "step": 479
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.423592493297588e-05,
+      "loss": 0.2677,
+      "step": 480
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.4222520107238607e-05,
+      "loss": 0.3331,
+      "step": 481
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.420911528150134e-05,
+      "loss": 0.4375,
+      "step": 482
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.419571045576408e-05,
+      "loss": 0.3097,
+      "step": 483
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.418230563002681e-05,
+      "loss": 0.4438,
+      "step": 484
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.416890080428955e-05,
+      "loss": 0.3648,
+      "step": 485
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.415549597855228e-05,
+      "loss": 0.3023,
+      "step": 486
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.414209115281502e-05,
+      "loss": 0.2963,
+      "step": 487
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.412868632707775e-05,
+      "loss": 0.1749,
+      "step": 488
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.411528150134048e-05,
+      "loss": 0.2006,
+      "step": 489
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.410187667560322e-05,
+      "loss": 0.0923,
+      "step": 490
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.408847184986595e-05,
+      "loss": 0.2665,
+      "step": 491
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.407506702412869e-05,
+      "loss": 0.3378,
+      "step": 492
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.406166219839142e-05,
+      "loss": 0.2543,
+      "step": 493
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.4048257372654156e-05,
+      "loss": 0.2521,
+      "step": 494
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.403485254691689e-05,
+      "loss": 0.1616,
+      "step": 495
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.402144772117963e-05,
+      "loss": 0.1445,
+      "step": 496
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.400804289544236e-05,
+      "loss": 0.5352,
+      "step": 497
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.39946380697051e-05,
+      "loss": 0.0654,
+      "step": 498
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.398123324396783e-05,
+      "loss": 0.4354,
+      "step": 499
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.396782841823057e-05,
+      "loss": 0.3893,
+      "step": 500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.39544235924933e-05,
+      "loss": 0.2779,
+      "step": 501
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.394101876675604e-05,
+      "loss": 0.3702,
+      "step": 502
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.392761394101877e-05,
+      "loss": 0.3899,
+      "step": 503
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.39142091152815e-05,
+      "loss": 0.3091,
+      "step": 504
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.390080428954424e-05,
+      "loss": 0.2143,
+      "step": 505
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.388739946380697e-05,
+      "loss": 0.2551,
+      "step": 506
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.387399463806971e-05,
+      "loss": 0.2227,
+      "step": 507
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.386058981233244e-05,
+      "loss": 0.1383,
+      "step": 508
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.384718498659518e-05,
+      "loss": 0.3463,
+      "step": 509
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.383378016085791e-05,
+      "loss": 0.1104,
+      "step": 510
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.382037533512065e-05,
+      "loss": 0.2439,
+      "step": 511
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.380697050938338e-05,
+      "loss": 0.0725,
+      "step": 512
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.379356568364612e-05,
+      "loss": 0.4928,
+      "step": 513
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.3780160857908846e-05,
+      "loss": 0.0903,
+      "step": 514
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.376675603217159e-05,
+      "loss": 0.1808,
+      "step": 515
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.3753351206434316e-05,
+      "loss": 0.617,
+      "step": 516
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.373994638069706e-05,
+      "loss": 0.0333,
+      "step": 517
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.3726541554959786e-05,
+      "loss": 0.1113,
+      "step": 518
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.371313672922252e-05,
+      "loss": 0.4604,
+      "step": 519
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.3699731903485256e-05,
+      "loss": 0.4267,
+      "step": 520
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.368632707774799e-05,
+      "loss": 0.2621,
+      "step": 521
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.3672922252010727e-05,
+      "loss": 0.3236,
+      "step": 522
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.365951742627346e-05,
+      "loss": 0.0559,
+      "step": 523
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.364611260053619e-05,
+      "loss": 0.0801,
+      "step": 524
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.363270777479893e-05,
+      "loss": 0.0518,
+      "step": 525
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.361930294906166e-05,
+      "loss": 0.0618,
+      "step": 526
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.36058981233244e-05,
+      "loss": 0.3043,
+      "step": 527
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.359249329758713e-05,
+      "loss": 0.4693,
+      "step": 528
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3579088471849866e-05,
+      "loss": 0.2978,
+      "step": 529
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.35656836461126e-05,
+      "loss": 0.0817,
+      "step": 530
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3552278820375336e-05,
+      "loss": 0.0784,
+      "step": 531
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.353887399463807e-05,
+      "loss": 0.5389,
+      "step": 532
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3525469168900806e-05,
+      "loss": 0.2337,
+      "step": 533
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.351206434316354e-05,
+      "loss": 0.3189,
+      "step": 534
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.3498659517426276e-05,
+      "loss": 0.2503,
+      "step": 535
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.348525469168901e-05,
+      "loss": 0.333,
+      "step": 536
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.347184986595175e-05,
+      "loss": 0.2311,
+      "step": 537
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.345844504021448e-05,
+      "loss": 0.5196,
+      "step": 538
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.344504021447721e-05,
+      "loss": 0.5466,
+      "step": 539
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.343163538873995e-05,
+      "loss": 0.4296,
+      "step": 540
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.341823056300268e-05,
+      "loss": 0.6559,
+      "step": 541
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.340482573726542e-05,
+      "loss": 0.6134,
+      "step": 542
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.339142091152815e-05,
+      "loss": 0.5614,
+      "step": 543
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.3378016085790886e-05,
+      "loss": 0.3143,
+      "step": 544
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.336461126005362e-05,
+      "loss": 0.564,
+      "step": 545
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.3351206434316356e-05,
+      "loss": 0.2141,
+      "step": 546
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.333780160857909e-05,
+      "loss": 0.5886,
+      "step": 547
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.3324396782841826e-05,
+      "loss": 0.2129,
+      "step": 548
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.3310991957104555e-05,
+      "loss": 0.2491,
+      "step": 549
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.32975871313673e-05,
+      "loss": 0.4023,
+      "step": 550
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.3284182305630025e-05,
+      "loss": 0.2159,
+      "step": 551
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.327077747989277e-05,
+      "loss": 0.2975,
+      "step": 552
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.3257372654155495e-05,
+      "loss": 0.165,
+      "step": 553
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.324396782841823e-05,
+      "loss": 0.063,
+      "step": 554
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.3230563002680966e-05,
+      "loss": 0.4209,
+      "step": 555
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.32171581769437e-05,
+      "loss": 0.2704,
+      "step": 556
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.3203753351206436e-05,
+      "loss": 0.6455,
+      "step": 557
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.319034852546917e-05,
+      "loss": 0.3354,
+      "step": 558
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.3176943699731906e-05,
+      "loss": 0.4069,
+      "step": 559
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.316353887399464e-05,
+      "loss": 0.2334,
+      "step": 560
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.3150134048257376e-05,
+      "loss": 0.4192,
+      "step": 561
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.313672922252011e-05,
+      "loss": 0.2942,
+      "step": 562
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.3123324396782847e-05,
+      "loss": 0.2503,
+      "step": 563
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.3109919571045575e-05,
+      "loss": 0.15,
+      "step": 564
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.309651474530832e-05,
+      "loss": 0.6895,
+      "step": 565
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.3083109919571045e-05,
+      "loss": 0.2462,
+      "step": 566
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.306970509383379e-05,
+      "loss": 0.142,
+      "step": 567
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.3056300268096515e-05,
+      "loss": 0.4838,
+      "step": 568
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.304289544235925e-05,
+      "loss": 0.4714,
+      "step": 569
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.3029490616621986e-05,
+      "loss": 0.4916,
+      "step": 570
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.301608579088472e-05,
+      "loss": 0.2001,
+      "step": 571
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.3002680965147456e-05,
+      "loss": 0.3897,
+      "step": 572
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.298927613941019e-05,
+      "loss": 0.1723,
+      "step": 573
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.297587131367292e-05,
+      "loss": 0.4095,
+      "step": 574
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.296246648793566e-05,
+      "loss": 0.2615,
+      "step": 575
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.294906166219839e-05,
+      "loss": 0.221,
+      "step": 576
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.293565683646113e-05,
+      "loss": 0.1373,
+      "step": 577
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.292225201072386e-05,
+      "loss": 0.313,
+      "step": 578
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.2908847184986595e-05,
+      "loss": 0.4608,
+      "step": 579
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.289544235924933e-05,
+      "loss": 0.2894,
+      "step": 580
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.2882037533512065e-05,
+      "loss": 0.5123,
+      "step": 581
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.28686327077748e-05,
+      "loss": 0.3287,
+      "step": 582
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.2855227882037536e-05,
+      "loss": 0.4634,
+      "step": 583
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.284182305630027e-05,
+      "loss": 0.0753,
+      "step": 584
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.2828418230563006e-05,
+      "loss": 0.5287,
+      "step": 585
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.281501340482574e-05,
+      "loss": 0.5483,
+      "step": 586
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.2801608579088476e-05,
+      "loss": 0.1076,
+      "step": 587
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.278820375335121e-05,
+      "loss": 0.232,
+      "step": 588
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.277479892761394e-05,
+      "loss": 0.4706,
+      "step": 589
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.276139410187668e-05,
+      "loss": 0.5791,
+      "step": 590
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.274798927613941e-05,
+      "loss": 0.4995,
+      "step": 591
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.273458445040215e-05,
+      "loss": 0.3419,
+      "step": 592
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.272117962466488e-05,
+      "loss": 0.2728,
+      "step": 593
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.2707774798927615e-05,
+      "loss": 0.3598,
+      "step": 594
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.269436997319035e-05,
+      "loss": 0.2669,
+      "step": 595
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.2680965147453086e-05,
+      "loss": 0.4091,
+      "step": 596
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.266756032171582e-05,
+      "loss": 0.7653,
+      "step": 597
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.2654155495978556e-05,
+      "loss": 0.4323,
+      "step": 598
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.2640750670241284e-05,
+      "loss": 0.2147,
+      "step": 599
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.2627345844504026e-05,
+      "loss": 0.2229,
+      "step": 600
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.2613941018766754e-05,
+      "loss": 0.1477,
+      "step": 601
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.2600536193029496e-05,
+      "loss": 0.4813,
+      "step": 602
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.2587131367292225e-05,
+      "loss": 0.2719,
+      "step": 603
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.257372654155496e-05,
+      "loss": 0.1577,
+      "step": 604
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.2560321715817695e-05,
+      "loss": 0.5718,
+      "step": 605
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.254691689008043e-05,
+      "loss": 0.1481,
+      "step": 606
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.2533512064343165e-05,
+      "loss": 0.3548,
+      "step": 607
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.25201072386059e-05,
+      "loss": 0.1907,
+      "step": 608
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.250670241286863e-05,
+      "loss": 0.4215,
+      "step": 609
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.249329758713137e-05,
+      "loss": 0.0681,
+      "step": 610
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.24798927613941e-05,
+      "loss": 0.514,
+      "step": 611
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.246648793565684e-05,
+      "loss": 0.1242,
+      "step": 612
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.245308310991957e-05,
+      "loss": 0.217,
+      "step": 613
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.243967828418231e-05,
+      "loss": 0.1641,
+      "step": 614
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.242627345844504e-05,
+      "loss": 0.7095,
+      "step": 615
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.2412868632707775e-05,
+      "loss": 0.359,
+      "step": 616
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.239946380697051e-05,
+      "loss": 0.94,
+      "step": 617
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.2386058981233245e-05,
+      "loss": 0.5194,
+      "step": 618
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.237265415549598e-05,
+      "loss": 0.3398,
+      "step": 619
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.2359249329758715e-05,
+      "loss": 0.2305,
+      "step": 620
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.234584450402145e-05,
+      "loss": 0.6045,
+      "step": 621
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.2332439678284185e-05,
+      "loss": 0.3196,
+      "step": 622
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.231903485254692e-05,
+      "loss": 0.0753,
+      "step": 623
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.2305630026809656e-05,
+      "loss": 0.2732,
+      "step": 624
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.229222520107239e-05,
+      "loss": 0.4372,
+      "step": 625
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.227882037533512e-05,
+      "loss": 0.1299,
+      "step": 626
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.226541554959786e-05,
+      "loss": 0.2928,
+      "step": 627
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.225201072386059e-05,
+      "loss": 0.2028,
+      "step": 628
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.223860589812333e-05,
+      "loss": 0.2725,
+      "step": 629
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.222520107238606e-05,
+      "loss": 0.0851,
+      "step": 630
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.2211796246648795e-05,
+      "loss": 0.1471,
+      "step": 631
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.219839142091153e-05,
+      "loss": 0.1986,
+      "step": 632
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.2184986595174265e-05,
+      "loss": 0.5156,
+      "step": 633
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.2171581769437e-05,
+      "loss": 0.5286,
+      "step": 634
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.2158176943699735e-05,
+      "loss": 0.3635,
+      "step": 635
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.2144772117962464e-05,
+      "loss": 0.1407,
+      "step": 636
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.2131367292225206e-05,
+      "loss": 0.1042,
+      "step": 637
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.2117962466487934e-05,
+      "loss": 0.1553,
+      "step": 638
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.2104557640750676e-05,
+      "loss": 0.1665,
+      "step": 639
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.2091152815013404e-05,
+      "loss": 0.3706,
+      "step": 640
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.207774798927614e-05,
+      "loss": 0.6195,
+      "step": 641
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.2064343163538874e-05,
+      "loss": 0.1341,
+      "step": 642
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.205093833780161e-05,
+      "loss": 0.5384,
+      "step": 643
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.2037533512064345e-05,
+      "loss": 0.2802,
+      "step": 644
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.202412868632708e-05,
+      "loss": 0.3812,
+      "step": 645
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.2010723860589815e-05,
+      "loss": 0.2433,
+      "step": 646
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.199731903485255e-05,
+      "loss": 0.0642,
+      "step": 647
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.1983914209115285e-05,
+      "loss": 0.0547,
+      "step": 648
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.197050938337802e-05,
+      "loss": 0.4388,
+      "step": 649
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.1957104557640756e-05,
+      "loss": 0.8228,
+      "step": 650
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.1943699731903484e-05,
+      "loss": 0.6453,
+      "step": 651
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.1930294906166226e-05,
+      "loss": 0.3367,
+      "step": 652
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.1916890080428954e-05,
+      "loss": 0.2139,
+      "step": 653
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.1903485254691696e-05,
+      "loss": 0.2144,
+      "step": 654
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.1890080428954424e-05,
+      "loss": 0.3894,
+      "step": 655
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.187667560321716e-05,
+      "loss": 0.0891,
+      "step": 656
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.1863270777479895e-05,
+      "loss": 0.1287,
+      "step": 657
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.184986595174263e-05,
+      "loss": 0.2826,
+      "step": 658
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.1836461126005365e-05,
+      "loss": 0.474,
+      "step": 659
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.18230563002681e-05,
+      "loss": 0.4228,
+      "step": 660
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.180965147453083e-05,
+      "loss": 0.4952,
+      "step": 661
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.179624664879357e-05,
+      "loss": 0.173,
+      "step": 662
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.17828418230563e-05,
+      "loss": 0.363,
+      "step": 663
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.176943699731904e-05,
+      "loss": 0.4404,
+      "step": 664
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.175603217158177e-05,
+      "loss": 0.486,
+      "step": 665
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.1742627345844504e-05,
+      "loss": 0.4463,
+      "step": 666
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.172922252010724e-05,
+      "loss": 0.2409,
+      "step": 667
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.1715817694369974e-05,
+      "loss": 0.5291,
+      "step": 668
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.170241286863271e-05,
+      "loss": 0.069,
+      "step": 669
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.1689008042895445e-05,
+      "loss": 0.4162,
+      "step": 670
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.167560321715818e-05,
+      "loss": 0.6171,
+      "step": 671
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.1662198391420915e-05,
+      "loss": 0.3097,
+      "step": 672
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.164879356568365e-05,
+      "loss": 0.5109,
+      "step": 673
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.1635388739946385e-05,
+      "loss": 0.2169,
+      "step": 674
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.162198391420912e-05,
+      "loss": 0.2406,
+      "step": 675
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.160857908847185e-05,
+      "loss": 0.1853,
+      "step": 676
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.159517426273459e-05,
+      "loss": 0.5743,
+      "step": 677
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.158176943699732e-05,
+      "loss": 0.5432,
+      "step": 678
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.156836461126006e-05,
+      "loss": 0.2033,
+      "step": 679
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.155495978552279e-05,
+      "loss": 0.3848,
+      "step": 680
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.1541554959785524e-05,
+      "loss": 0.1721,
+      "step": 681
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.152815013404826e-05,
+      "loss": 0.3793,
+      "step": 682
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.1514745308310994e-05,
+      "loss": 0.3848,
+      "step": 683
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.150134048257373e-05,
+      "loss": 0.186,
+      "step": 684
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.1487935656836465e-05,
+      "loss": 0.2692,
+      "step": 685
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.147453083109919e-05,
+      "loss": 0.3839,
+      "step": 686
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.1461126005361935e-05,
+      "loss": 0.3037,
+      "step": 687
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.144772117962466e-05,
+      "loss": 0.6401,
+      "step": 688
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.1434316353887405e-05,
+      "loss": 0.1173,
+      "step": 689
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.1420911528150134e-05,
+      "loss": 0.3217,
+      "step": 690
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.140750670241287e-05,
+      "loss": 0.2358,
+      "step": 691
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.1394101876675604e-05,
+      "loss": 0.7696,
+      "step": 692
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.138069705093834e-05,
+      "loss": 0.2288,
+      "step": 693
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.1367292225201074e-05,
+      "loss": 0.2575,
+      "step": 694
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.135388739946381e-05,
+      "loss": 0.1201,
+      "step": 695
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.1340482573726544e-05,
+      "loss": 0.2034,
+      "step": 696
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.132707774798928e-05,
+      "loss": 0.1142,
+      "step": 697
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.1313672922252015e-05,
+      "loss": 0.5671,
+      "step": 698
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.130026809651475e-05,
+      "loss": 0.3132,
+      "step": 699
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.1286863270777485e-05,
+      "loss": 0.4266,
+      "step": 700
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.127345844504021e-05,
+      "loss": 0.1354,
+      "step": 701
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 4.1260053619302955e-05,
+      "loss": 0.2867,
+      "step": 702
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 4.1246648793565684e-05,
+      "loss": 0.1839,
+      "step": 703
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 4.1233243967828425e-05,
+      "loss": 0.4741,
+      "step": 704
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.1219839142091154e-05,
+      "loss": 0.2909,
+      "step": 705
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.120643431635389e-05,
+      "loss": 0.2705,
+      "step": 706
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.1193029490616624e-05,
+      "loss": 0.1354,
+      "step": 707
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.117962466487936e-05,
+      "loss": 0.4801,
+      "step": 708
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.1166219839142094e-05,
+      "loss": 0.189,
+      "step": 709
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.115281501340483e-05,
+      "loss": 0.3204,
+      "step": 710
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.113941018766756e-05,
+      "loss": 0.4358,
+      "step": 711
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.11260053619303e-05,
+      "loss": 0.9474,
+      "step": 712
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.111260053619303e-05,
+      "loss": 0.2102,
+      "step": 713
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.109919571045577e-05,
+      "loss": 0.3927,
+      "step": 714
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.10857908847185e-05,
+      "loss": 0.139,
+      "step": 715
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.1072386058981233e-05,
+      "loss": 0.3575,
+      "step": 716
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 4.105898123324397e-05,
+      "loss": 0.7534,
+      "step": 717
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 4.1045576407506704e-05,
+      "loss": 0.1134,
+      "step": 718
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 4.103217158176944e-05,
+      "loss": 0.2136,
+      "step": 719
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 4.1018766756032174e-05,
+      "loss": 0.4344,
+      "step": 720
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 4.10053619302949e-05,
+      "loss": 0.0695,
+      "step": 721
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 4.0991957104557644e-05,
+      "loss": 0.2286,
+      "step": 722
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 4.097855227882037e-05,
+      "loss": 0.1189,
+      "step": 723
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 4.0965147453083115e-05,
+      "loss": 0.2882,
+      "step": 724
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 4.095174262734584e-05,
+      "loss": 0.2623,
+      "step": 725
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 4.093833780160858e-05,
+      "loss": 0.2473,
+      "step": 726
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 4.092493297587131e-05,
+      "loss": 0.4846,
+      "step": 727
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 4.091152815013405e-05,
+      "loss": 0.1689,
+      "step": 728
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 4.0898123324396783e-05,
+      "loss": 0.3481,
+      "step": 729
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 4.088471849865952e-05,
+      "loss": 0.3447,
+      "step": 730
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 4.0871313672922254e-05,
+      "loss": 0.2959,
+      "step": 731
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.085790884718499e-05,
+      "loss": 0.3387,
+      "step": 732
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.0844504021447724e-05,
+      "loss": 0.3742,
+      "step": 733
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.083109919571046e-05,
+      "loss": 0.3245,
+      "step": 734
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.0817694369973194e-05,
+      "loss": 0.4891,
+      "step": 735
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.080428954423593e-05,
+      "loss": 0.1444,
+      "step": 736
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.0790884718498664e-05,
+      "loss": 0.3678,
+      "step": 737
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.077747989276139e-05,
+      "loss": 0.3772,
+      "step": 738
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.0764075067024135e-05,
+      "loss": 0.43,
+      "step": 739
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.075067024128686e-05,
+      "loss": 0.2463,
+      "step": 740
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.0737265415549605e-05,
+      "loss": 0.2277,
+      "step": 741
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.072386058981233e-05,
+      "loss": 0.2153,
+      "step": 742
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.071045576407507e-05,
+      "loss": 0.1052,
+      "step": 743
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.0697050938337804e-05,
+      "loss": 0.5657,
+      "step": 744
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.068364611260054e-05,
+      "loss": 0.2664,
+      "step": 745
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.0670241286863274e-05,
+      "loss": 0.1369,
+      "step": 746
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.065683646112601e-05,
+      "loss": 0.4972,
+      "step": 747
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.064343163538874e-05,
+      "loss": 0.4131,
+      "step": 748
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.063002680965148e-05,
+      "loss": 0.3824,
+      "step": 749
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.061662198391421e-05,
+      "loss": 0.323,
+      "step": 750
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.060321715817695e-05,
+      "loss": 0.3698,
+      "step": 751
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.058981233243968e-05,
+      "loss": 0.1708,
+      "step": 752
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.057640750670241e-05,
+      "loss": 0.2941,
+      "step": 753
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.056300268096515e-05,
+      "loss": 0.3224,
+      "step": 754
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.054959785522788e-05,
+      "loss": 0.0851,
+      "step": 755
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.053619302949062e-05,
+      "loss": 0.4694,
+      "step": 756
+    },
+    {
+      "epoch": 2.0,
+      "eval_f1": 0.7882736156351792,
+      "eval_loss": 0.4484867751598358,
+      "eval_runtime": 1.8734,
+      "eval_samples_per_second": 807.638,
+      "eval_steps_per_second": 50.711,
+      "step": 756
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.0522788203753354e-05,
+      "loss": 0.4739,
+      "step": 757
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.050938337801609e-05,
+      "loss": 0.4117,
+      "step": 758
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0495978552278824e-05,
+      "loss": 0.2169,
+      "step": 759
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.048257372654156e-05,
+      "loss": 0.1848,
+      "step": 760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.0469168900804294e-05,
+      "loss": 0.5066,
+      "step": 761
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.045576407506703e-05,
+      "loss": 0.1784,
+      "step": 762
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.044235924932976e-05,
+      "loss": 0.3869,
+      "step": 763
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.04289544235925e-05,
+      "loss": 0.1132,
+      "step": 764
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.041554959785523e-05,
+      "loss": 0.2724,
+      "step": 765
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.040214477211797e-05,
+      "loss": 0.0983,
+      "step": 766
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.03887399463807e-05,
+      "loss": 0.1831,
+      "step": 767
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.037533512064343e-05,
+      "loss": 0.1954,
+      "step": 768
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.036193029490617e-05,
+      "loss": 0.4738,
+      "step": 769
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.0348525469168903e-05,
+      "loss": 0.3375,
+      "step": 770
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.033512064343164e-05,
+      "loss": 0.3991,
+      "step": 771
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.0321715817694374e-05,
+      "loss": 0.0696,
+      "step": 772
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.03083109919571e-05,
+      "loss": 0.0982,
+      "step": 773
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.0294906166219844e-05,
+      "loss": 0.2236,
+      "step": 774
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.028150134048257e-05,
+      "loss": 0.4225,
+      "step": 775
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.0268096514745314e-05,
+      "loss": 0.0583,
+      "step": 776
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.025469168900804e-05,
+      "loss": 0.1643,
+      "step": 777
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.024128686327078e-05,
+      "loss": 0.0351,
+      "step": 778
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.022788203753351e-05,
+      "loss": 0.4496,
+      "step": 779
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.021447721179625e-05,
+      "loss": 0.0372,
+      "step": 780
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.020107238605898e-05,
+      "loss": 0.4198,
+      "step": 781
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.018766756032172e-05,
+      "loss": 0.1968,
+      "step": 782
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.017426273458445e-05,
+      "loss": 0.515,
+      "step": 783
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.016085790884719e-05,
+      "loss": 0.408,
+      "step": 784
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.0147453083109924e-05,
+      "loss": 0.3693,
+      "step": 785
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.013404825737266e-05,
+      "loss": 0.0561,
+      "step": 786
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.0120643431635394e-05,
+      "loss": 0.4011,
+      "step": 787
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.010723860589812e-05,
+      "loss": 0.0872,
+      "step": 788
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.0093833780160864e-05,
+      "loss": 0.0768,
+      "step": 789
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.008042895442359e-05,
+      "loss": 0.0184,
+      "step": 790
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.0067024128686334e-05,
+      "loss": 0.3287,
+      "step": 791
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.005361930294906e-05,
+      "loss": 0.0262,
+      "step": 792
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.00402144772118e-05,
+      "loss": 0.0248,
+      "step": 793
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.002680965147453e-05,
+      "loss": 0.3853,
+      "step": 794
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.001340482573727e-05,
+      "loss": 0.3512,
+      "step": 795
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4e-05,
+      "loss": 0.4188,
+      "step": 796
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.998659517426274e-05,
+      "loss": 0.1834,
+      "step": 797
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.997319034852547e-05,
+      "loss": 0.2074,
+      "step": 798
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 3.995978552278821e-05,
+      "loss": 0.7317,
+      "step": 799
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.994638069705094e-05,
+      "loss": 0.3534,
+      "step": 800
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.993297587131368e-05,
+      "loss": 0.3184,
+      "step": 801
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.991957104557641e-05,
+      "loss": 0.1088,
+      "step": 802
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.990616621983914e-05,
+      "loss": 0.0429,
+      "step": 803
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.989276139410188e-05,
+      "loss": 0.4518,
+      "step": 804
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.987935656836461e-05,
+      "loss": 0.1746,
+      "step": 805
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.986595174262735e-05,
+      "loss": 0.1881,
+      "step": 806
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.985254691689008e-05,
+      "loss": 0.4111,
+      "step": 807
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.983914209115281e-05,
+      "loss": 0.059,
+      "step": 808
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.982573726541555e-05,
+      "loss": 0.0495,
+      "step": 809
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.981233243967828e-05,
+      "loss": 0.1134,
+      "step": 810
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9798927613941023e-05,
+      "loss": 0.5469,
+      "step": 811
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.978552278820375e-05,
+      "loss": 0.0581,
+      "step": 812
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.977211796246649e-05,
+      "loss": 0.1254,
+      "step": 813
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.975871313672922e-05,
+      "loss": 0.0679,
+      "step": 814
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.974530831099196e-05,
+      "loss": 0.0463,
+      "step": 815
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.973190348525469e-05,
+      "loss": 0.33,
+      "step": 816
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.971849865951743e-05,
+      "loss": 0.2931,
+      "step": 817
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.970509383378016e-05,
+      "loss": 0.1034,
+      "step": 818
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.96916890080429e-05,
+      "loss": 0.0379,
+      "step": 819
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.967828418230563e-05,
+      "loss": 0.0456,
+      "step": 820
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.966487935656837e-05,
+      "loss": 0.4862,
+      "step": 821
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.96514745308311e-05,
+      "loss": 0.0512,
+      "step": 822
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.963806970509383e-05,
+      "loss": 0.0879,
+      "step": 823
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.962466487935657e-05,
+      "loss": 0.3664,
+      "step": 824
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.96112600536193e-05,
+      "loss": 0.0975,
+      "step": 825
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.9597855227882044e-05,
+      "loss": 0.162,
+      "step": 826
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.958445040214477e-05,
+      "loss": 0.1076,
+      "step": 827
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.957104557640751e-05,
+      "loss": 0.1185,
+      "step": 828
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.955764075067024e-05,
+      "loss": 0.0868,
+      "step": 829
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.954423592493298e-05,
+      "loss": 0.0246,
+      "step": 830
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.953083109919571e-05,
+      "loss": 0.2233,
+      "step": 831
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.951742627345845e-05,
+      "loss": 0.2113,
+      "step": 832
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.9504021447721176e-05,
+      "loss": 0.0587,
+      "step": 833
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.949061662198392e-05,
+      "loss": 0.0521,
+      "step": 834
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.9477211796246646e-05,
+      "loss": 0.3845,
+      "step": 835
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.946380697050939e-05,
+      "loss": 0.1096,
+      "step": 836
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.9450402144772117e-05,
+      "loss": 0.1488,
+      "step": 837
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.943699731903485e-05,
+      "loss": 0.1937,
+      "step": 838
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.942359249329759e-05,
+      "loss": 0.1309,
+      "step": 839
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.941018766756032e-05,
+      "loss": 0.3271,
+      "step": 840
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.939678284182306e-05,
+      "loss": 0.3318,
+      "step": 841
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.938337801608579e-05,
+      "loss": 0.3516,
+      "step": 842
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.936997319034853e-05,
+      "loss": 0.1641,
+      "step": 843
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.935656836461126e-05,
+      "loss": 0.064,
+      "step": 844
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.9343163538874e-05,
+      "loss": 0.1971,
+      "step": 845
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.932975871313673e-05,
+      "loss": 0.1166,
+      "step": 846
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.931635388739947e-05,
+      "loss": 0.0384,
+      "step": 847
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.9302949061662196e-05,
+      "loss": 0.0462,
+      "step": 848
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.928954423592494e-05,
+      "loss": 0.1073,
+      "step": 849
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.9276139410187666e-05,
+      "loss": 0.0227,
+      "step": 850
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.926273458445041e-05,
+      "loss": 0.0683,
+      "step": 851
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.924932975871314e-05,
+      "loss": 0.7962,
+      "step": 852
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.923592493297587e-05,
+      "loss": 0.0635,
+      "step": 853
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.922252010723861e-05,
+      "loss": 0.0454,
+      "step": 854
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.920911528150134e-05,
+      "loss": 0.5389,
+      "step": 855
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.919571045576408e-05,
+      "loss": 0.1181,
+      "step": 856
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.918230563002681e-05,
+      "loss": 0.1372,
+      "step": 857
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.916890080428954e-05,
+      "loss": 0.3192,
+      "step": 858
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.915549597855228e-05,
+      "loss": 0.3419,
+      "step": 859
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.914209115281501e-05,
+      "loss": 0.0114,
+      "step": 860
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.912868632707775e-05,
+      "loss": 0.1905,
+      "step": 861
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.911528150134048e-05,
+      "loss": 0.0218,
+      "step": 862
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.910187667560322e-05,
+      "loss": 0.4374,
+      "step": 863
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.908847184986595e-05,
+      "loss": 0.1844,
+      "step": 864
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.907506702412869e-05,
+      "loss": 0.2427,
+      "step": 865
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.906166219839142e-05,
+      "loss": 0.2749,
+      "step": 866
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 3.904825737265416e-05,
+      "loss": 0.2089,
+      "step": 867
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.903485254691689e-05,
+      "loss": 0.06,
+      "step": 868
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.902144772117963e-05,
+      "loss": 0.1951,
+      "step": 869
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.900804289544236e-05,
+      "loss": 0.0252,
+      "step": 870
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 3.89946380697051e-05,
+      "loss": 0.0299,
+      "step": 871
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 3.898123324396783e-05,
+      "loss": 0.0298,
+      "step": 872
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 3.896782841823057e-05,
+      "loss": 0.5186,
+      "step": 873
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 3.89544235924933e-05,
+      "loss": 0.2704,
+      "step": 874
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 3.894101876675603e-05,
+      "loss": 0.2435,
+      "step": 875
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 3.892761394101877e-05,
+      "loss": 0.039,
+      "step": 876
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 3.89142091152815e-05,
+      "loss": 0.0275,
+      "step": 877
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 3.890080428954424e-05,
+      "loss": 0.1164,
+      "step": 878
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 3.888739946380697e-05,
+      "loss": 0.1551,
+      "step": 879
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 3.887399463806971e-05,
+      "loss": 0.0215,
+      "step": 880
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 3.886058981233244e-05,
+      "loss": 0.0379,
+      "step": 881
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 3.884718498659518e-05,
+      "loss": 0.0553,
+      "step": 882
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 3.883378016085791e-05,
+      "loss": 0.1073,
+      "step": 883
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 3.882037533512065e-05,
+      "loss": 0.3525,
+      "step": 884
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 3.8806970509383376e-05,
+      "loss": 0.2646,
+      "step": 885
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 3.879356568364612e-05,
+      "loss": 0.5758,
+      "step": 886
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.8780160857908846e-05,
+      "loss": 0.9312,
+      "step": 887
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.876675603217159e-05,
+      "loss": 0.2748,
+      "step": 888
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.8753351206434316e-05,
+      "loss": 0.33,
+      "step": 889
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 3.873994638069705e-05,
+      "loss": 0.0312,
+      "step": 890
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 3.8726541554959786e-05,
+      "loss": 0.0449,
+      "step": 891
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 3.871313672922252e-05,
+      "loss": 0.1197,
+      "step": 892
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 3.869973190348526e-05,
+      "loss": 0.0913,
+      "step": 893
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.868632707774799e-05,
+      "loss": 0.0284,
+      "step": 894
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.867292225201073e-05,
+      "loss": 0.3769,
+      "step": 895
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.865951742627346e-05,
+      "loss": 0.0947,
+      "step": 896
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 3.86461126005362e-05,
+      "loss": 0.4282,
+      "step": 897
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.863270777479893e-05,
+      "loss": 0.0049,
+      "step": 898
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.861930294906167e-05,
+      "loss": 0.3632,
+      "step": 899
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.8605898123324396e-05,
+      "loss": 0.0421,
+      "step": 900
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.859249329758714e-05,
+      "loss": 0.5793,
+      "step": 901
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.8579088471849866e-05,
+      "loss": 0.1695,
+      "step": 902
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.856568364611261e-05,
+      "loss": 0.3082,
+      "step": 903
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.8552278820375336e-05,
+      "loss": 0.0151,
+      "step": 904
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 3.853887399463807e-05,
+      "loss": 0.3463,
+      "step": 905
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.852546916890081e-05,
+      "loss": 0.4573,
+      "step": 906
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.851206434316354e-05,
+      "loss": 0.1281,
+      "step": 907
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.849865951742628e-05,
+      "loss": 0.3168,
+      "step": 908
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 3.848525469168901e-05,
+      "loss": 0.0331,
+      "step": 909
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.847184986595174e-05,
+      "loss": 0.1825,
+      "step": 910
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.845844504021448e-05,
+      "loss": 0.6238,
+      "step": 911
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 3.844504021447721e-05,
+      "loss": 0.0663,
+      "step": 912
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.843163538873995e-05,
+      "loss": 0.2827,
+      "step": 913
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.841823056300268e-05,
+      "loss": 0.0133,
+      "step": 914
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.8404825737265416e-05,
+      "loss": 0.0415,
+      "step": 915
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 3.839142091152815e-05,
+      "loss": 0.2969,
+      "step": 916
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.8378016085790886e-05,
+      "loss": 0.0145,
+      "step": 917
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.836461126005362e-05,
+      "loss": 0.0223,
+      "step": 918
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.8351206434316357e-05,
+      "loss": 1.042,
+      "step": 919
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 3.8337801608579085e-05,
+      "loss": 0.5061,
+      "step": 920
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.832439678284183e-05,
+      "loss": 0.406,
+      "step": 921
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.8310991957104555e-05,
+      "loss": 0.1337,
+      "step": 922
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.82975871313673e-05,
+      "loss": 0.2057,
+      "step": 923
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.8284182305630025e-05,
+      "loss": 0.0987,
+      "step": 924
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.827077747989276e-05,
+      "loss": 0.4611,
+      "step": 925
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.8257372654155496e-05,
+      "loss": 0.0152,
+      "step": 926
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 3.824396782841823e-05,
+      "loss": 0.4478,
+      "step": 927
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.8230563002680966e-05,
+      "loss": 0.0579,
+      "step": 928
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.82171581769437e-05,
+      "loss": 0.1506,
+      "step": 929
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.8203753351206436e-05,
+      "loss": 0.5514,
+      "step": 930
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 3.819034852546917e-05,
+      "loss": 0.1505,
+      "step": 931
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.8176943699731906e-05,
+      "loss": 0.1413,
+      "step": 932
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.816353887399464e-05,
+      "loss": 0.4758,
+      "step": 933
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.815013404825738e-05,
+      "loss": 0.0863,
+      "step": 934
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.8136729222520105e-05,
+      "loss": 0.3561,
+      "step": 935
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.812332439678285e-05,
+      "loss": 0.7784,
+      "step": 936
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.8109919571045575e-05,
+      "loss": 0.2243,
+      "step": 937
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.809651474530832e-05,
+      "loss": 0.1013,
+      "step": 938
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 3.8083109919571046e-05,
+      "loss": 0.132,
+      "step": 939
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.806970509383378e-05,
+      "loss": 0.3971,
+      "step": 940
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.8056300268096516e-05,
+      "loss": 0.0637,
+      "step": 941
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.804289544235925e-05,
+      "loss": 0.5178,
+      "step": 942
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 3.8029490616621986e-05,
+      "loss": 0.1982,
+      "step": 943
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.801608579088472e-05,
+      "loss": 0.2225,
+      "step": 944
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.800268096514745e-05,
+      "loss": 0.1425,
+      "step": 945
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 3.798927613941019e-05,
+      "loss": 0.0621,
+      "step": 946
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.797587131367292e-05,
+      "loss": 0.1556,
+      "step": 947
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.796246648793566e-05,
+      "loss": 0.3805,
+      "step": 948
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.794906166219839e-05,
+      "loss": 0.7049,
+      "step": 949
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 3.7935656836461125e-05,
+      "loss": 0.1055,
+      "step": 950
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.792225201072386e-05,
+      "loss": 0.0489,
+      "step": 951
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.7908847184986596e-05,
+      "loss": 0.1881,
+      "step": 952
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.789544235924933e-05,
+      "loss": 0.0202,
+      "step": 953
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 3.7882037533512066e-05,
+      "loss": 0.1043,
+      "step": 954
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.78686327077748e-05,
+      "loss": 0.2093,
+      "step": 955
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.7855227882037536e-05,
+      "loss": 0.0395,
+      "step": 956
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.784182305630027e-05,
+      "loss": 0.1459,
+      "step": 957
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 3.7828418230563006e-05,
+      "loss": 0.0338,
+      "step": 958
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.781501340482574e-05,
+      "loss": 0.4741,
+      "step": 959
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.780160857908847e-05,
+      "loss": 0.2049,
+      "step": 960
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.778820375335121e-05,
+      "loss": 0.309,
+      "step": 961
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 3.777479892761394e-05,
+      "loss": 0.0253,
+      "step": 962
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.776139410187668e-05,
+      "loss": 0.4832,
+      "step": 963
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.774798927613941e-05,
+      "loss": 0.2111,
+      "step": 964
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 3.7734584450402145e-05,
+      "loss": 0.1788,
+      "step": 965
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.772117962466488e-05,
+      "loss": 0.5252,
+      "step": 966
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.7707774798927616e-05,
+      "loss": 0.4711,
+      "step": 967
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.769436997319035e-05,
+      "loss": 0.5184,
+      "step": 968
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 3.7680965147453086e-05,
+      "loss": 0.2164,
+      "step": 969
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.7667560321715814e-05,
+      "loss": 0.5393,
+      "step": 970
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.7654155495978556e-05,
+      "loss": 0.2588,
+      "step": 971
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.7640750670241285e-05,
+      "loss": 0.164,
+      "step": 972
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.7627345844504027e-05,
+      "loss": 0.2896,
+      "step": 973
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.7613941018766755e-05,
+      "loss": 0.039,
+      "step": 974
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.760053619302949e-05,
+      "loss": 0.16,
+      "step": 975
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.7587131367292225e-05,
+      "loss": 0.1832,
+      "step": 976
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 3.757372654155496e-05,
+      "loss": 0.0812,
+      "step": 977
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 3.7560321715817695e-05,
+      "loss": 0.1476,
+      "step": 978
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 3.754691689008043e-05,
+      "loss": 0.1853,
+      "step": 979
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 3.7533512064343166e-05,
+      "loss": 0.2875,
+      "step": 980
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 3.75201072386059e-05,
+      "loss": 0.1918,
+      "step": 981
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 3.7506702412868636e-05,
+      "loss": 0.2445,
+      "step": 982
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 3.749329758713137e-05,
+      "loss": 0.4653,
+      "step": 983
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 3.7479892761394106e-05,
+      "loss": 0.0614,
+      "step": 984
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 3.746648793565684e-05,
+      "loss": 0.2818,
+      "step": 985
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 3.7453083109919576e-05,
+      "loss": 0.1363,
+      "step": 986
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 3.7439678284182305e-05,
+      "loss": 0.3244,
+      "step": 987
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 3.742627345844505e-05,
+      "loss": 0.081,
+      "step": 988
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 3.7412868632707775e-05,
+      "loss": 0.0488,
+      "step": 989
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 3.739946380697052e-05,
+      "loss": 0.2057,
+      "step": 990
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 3.7386058981233245e-05,
+      "loss": 0.2598,
+      "step": 991
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 3.737265415549598e-05,
+      "loss": 0.1318,
+      "step": 992
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.7359249329758716e-05,
+      "loss": 0.0482,
+      "step": 993
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.734584450402145e-05,
+      "loss": 0.2586,
+      "step": 994
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.7332439678284186e-05,
+      "loss": 0.2533,
+      "step": 995
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.731903485254692e-05,
+      "loss": 0.5292,
+      "step": 996
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.730563002680965e-05,
+      "loss": 0.1677,
+      "step": 997
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.729222520107239e-05,
+      "loss": 0.1869,
+      "step": 998
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.727882037533512e-05,
+      "loss": 0.0645,
+      "step": 999
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.726541554959786e-05,
+      "loss": 0.4107,
+      "step": 1000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.725201072386059e-05,
+      "loss": 0.0484,
+      "step": 1001
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.7238605898123325e-05,
+      "loss": 0.0813,
+      "step": 1002
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.722520107238606e-05,
+      "loss": 0.2467,
+      "step": 1003
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.7211796246648795e-05,
+      "loss": 0.0324,
+      "step": 1004
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.719839142091153e-05,
+      "loss": 0.0536,
+      "step": 1005
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.7184986595174266e-05,
+      "loss": 0.0399,
+      "step": 1006
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.7171581769436994e-05,
+      "loss": 0.0257,
+      "step": 1007
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.7158176943699736e-05,
+      "loss": 0.0407,
+      "step": 1008
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.7144772117962464e-05,
+      "loss": 0.0173,
+      "step": 1009
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.7131367292225206e-05,
+      "loss": 0.0166,
+      "step": 1010
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.7117962466487934e-05,
+      "loss": 0.1898,
+      "step": 1011
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 3.710455764075067e-05,
+      "loss": 0.0525,
+      "step": 1012
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 3.7091152815013405e-05,
+      "loss": 0.043,
+      "step": 1013
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 3.707774798927614e-05,
+      "loss": 0.3994,
+      "step": 1014
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 3.7064343163538875e-05,
+      "loss": 0.0372,
+      "step": 1015
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 3.705093833780161e-05,
+      "loss": 0.2909,
+      "step": 1016
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 3.7037533512064345e-05,
+      "loss": 0.8221,
+      "step": 1017
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 3.702412868632708e-05,
+      "loss": 0.0084,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 3.7010723860589815e-05,
+      "loss": 0.3058,
+      "step": 1019
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 3.699731903485255e-05,
+      "loss": 1.0774,
+      "step": 1020
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 3.6983914209115286e-05,
+      "loss": 0.2018,
+      "step": 1021
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 3.6970509383378014e-05,
+      "loss": 0.0537,
+      "step": 1022
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 3.6957104557640756e-05,
+      "loss": 0.444,
+      "step": 1023
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 3.6943699731903484e-05,
+      "loss": 0.6497,
+      "step": 1024
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 3.6930294906166226e-05,
+      "loss": 0.0097,
+      "step": 1025
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 3.6916890080428955e-05,
+      "loss": 0.0082,
+      "step": 1026
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 3.690348525469169e-05,
+      "loss": 0.0387,
+      "step": 1027
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 3.6890080428954425e-05,
+      "loss": 0.3969,
+      "step": 1028
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 3.687667560321716e-05,
+      "loss": 0.0136,
+      "step": 1029
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 3.6863270777479895e-05,
+      "loss": 0.0099,
+      "step": 1030
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.684986595174263e-05,
+      "loss": 0.3509,
+      "step": 1031
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.683646112600536e-05,
+      "loss": 0.0257,
+      "step": 1032
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.68230563002681e-05,
+      "loss": 0.3158,
+      "step": 1033
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.680965147453083e-05,
+      "loss": 0.0677,
+      "step": 1034
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.679624664879357e-05,
+      "loss": 0.3887,
+      "step": 1035
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.67828418230563e-05,
+      "loss": 0.0714,
+      "step": 1036
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.6769436997319034e-05,
+      "loss": 0.1066,
+      "step": 1037
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.675603217158177e-05,
+      "loss": 0.6238,
+      "step": 1038
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.6742627345844504e-05,
+      "loss": 0.0405,
+      "step": 1039
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.672922252010724e-05,
+      "loss": 0.0223,
+      "step": 1040
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.6715817694369975e-05,
+      "loss": 0.2737,
+      "step": 1041
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.670241286863271e-05,
+      "loss": 0.015,
+      "step": 1042
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.6689008042895445e-05,
+      "loss": 0.1709,
+      "step": 1043
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.667560321715818e-05,
+      "loss": 0.2649,
+      "step": 1044
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.6662198391420915e-05,
+      "loss": 0.1524,
+      "step": 1045
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.664879356568365e-05,
+      "loss": 0.2461,
+      "step": 1046
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.663538873994638e-05,
+      "loss": 0.3425,
+      "step": 1047
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.662198391420912e-05,
+      "loss": 0.2689,
+      "step": 1048
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.660857908847185e-05,
+      "loss": 0.0066,
+      "step": 1049
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.659517426273459e-05,
+      "loss": 0.0328,
+      "step": 1050
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.658176943699732e-05,
+      "loss": 0.1273,
+      "step": 1051
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.6568364611260054e-05,
+      "loss": 0.2346,
+      "step": 1052
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.655495978552279e-05,
+      "loss": 0.0118,
+      "step": 1053
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.6541554959785525e-05,
+      "loss": 0.0287,
+      "step": 1054
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.652815013404826e-05,
+      "loss": 0.264,
+      "step": 1055
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.6514745308310995e-05,
+      "loss": 0.0216,
+      "step": 1056
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.650134048257372e-05,
+      "loss": 0.0261,
+      "step": 1057
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.6487935656836465e-05,
+      "loss": 0.1911,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.6474530831099194e-05,
+      "loss": 0.029,
+      "step": 1059
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.6461126005361935e-05,
+      "loss": 0.0393,
+      "step": 1060
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.6447721179624664e-05,
+      "loss": 0.1044,
+      "step": 1061
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.64343163538874e-05,
+      "loss": 0.6364,
+      "step": 1062
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.6420911528150134e-05,
+      "loss": 0.3589,
+      "step": 1063
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.640750670241287e-05,
+      "loss": 0.1128,
+      "step": 1064
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.6394101876675604e-05,
+      "loss": 0.2219,
+      "step": 1065
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.638069705093834e-05,
+      "loss": 0.211,
+      "step": 1066
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.6367292225201075e-05,
+      "loss": 0.387,
+      "step": 1067
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.635388739946381e-05,
+      "loss": 0.0725,
+      "step": 1068
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.6340482573726545e-05,
+      "loss": 0.0268,
+      "step": 1069
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.632707774798928e-05,
+      "loss": 0.516,
+      "step": 1070
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.6313672922252015e-05,
+      "loss": 0.0746,
+      "step": 1071
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 3.6300268096514743e-05,
+      "loss": 0.2486,
+      "step": 1072
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 3.6286863270777485e-05,
+      "loss": 0.1584,
+      "step": 1073
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 3.6273458445040214e-05,
+      "loss": 0.1301,
+      "step": 1074
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 3.6260053619302956e-05,
+      "loss": 0.0122,
+      "step": 1075
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 3.6246648793565684e-05,
+      "loss": 0.0215,
+      "step": 1076
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 3.623324396782842e-05,
+      "loss": 0.2068,
+      "step": 1077
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 3.6219839142091154e-05,
+      "loss": 0.1882,
+      "step": 1078
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 3.620643431635389e-05,
+      "loss": 0.368,
+      "step": 1079
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 3.6193029490616625e-05,
+      "loss": 0.7537,
+      "step": 1080
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 3.617962466487936e-05,
+      "loss": 0.1614,
+      "step": 1081
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 3.616621983914209e-05,
+      "loss": 0.0585,
+      "step": 1082
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.615281501340483e-05,
+      "loss": 0.0535,
+      "step": 1083
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.613941018766756e-05,
+      "loss": 0.6518,
+      "step": 1084
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.61260053619303e-05,
+      "loss": 0.4885,
+      "step": 1085
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.611260053619303e-05,
+      "loss": 0.4078,
+      "step": 1086
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.6099195710455764e-05,
+      "loss": 0.2101,
+      "step": 1087
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.60857908847185e-05,
+      "loss": 0.0192,
+      "step": 1088
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.6072386058981234e-05,
+      "loss": 0.3885,
+      "step": 1089
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.605898123324397e-05,
+      "loss": 0.0393,
+      "step": 1090
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.6045576407506704e-05,
+      "loss": 0.2179,
+      "step": 1091
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.603217158176944e-05,
+      "loss": 0.1814,
+      "step": 1092
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.6018766756032174e-05,
+      "loss": 0.0647,
+      "step": 1093
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.600536193029491e-05,
+      "loss": 0.0657,
+      "step": 1094
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.5991957104557645e-05,
+      "loss": 0.1062,
+      "step": 1095
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.597855227882038e-05,
+      "loss": 0.4314,
+      "step": 1096
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.596514745308311e-05,
+      "loss": 0.3074,
+      "step": 1097
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.595174262734585e-05,
+      "loss": 0.0159,
+      "step": 1098
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.593833780160858e-05,
+      "loss": 0.3829,
+      "step": 1099
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.592493297587132e-05,
+      "loss": 0.3277,
+      "step": 1100
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.591152815013405e-05,
+      "loss": 0.3785,
+      "step": 1101
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.5898123324396784e-05,
+      "loss": 0.0162,
+      "step": 1102
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.588471849865952e-05,
+      "loss": 0.0772,
+      "step": 1103
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.5871313672922254e-05,
+      "loss": 0.0292,
+      "step": 1104
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.585790884718499e-05,
+      "loss": 0.0748,
+      "step": 1105
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.5844504021447724e-05,
+      "loss": 0.2276,
+      "step": 1106
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.583109919571046e-05,
+      "loss": 0.3174,
+      "step": 1107
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.5817694369973195e-05,
+      "loss": 0.134,
+      "step": 1108
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.580428954423592e-05,
+      "loss": 0.3488,
+      "step": 1109
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.5790884718498665e-05,
+      "loss": 0.1003,
+      "step": 1110
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.577747989276139e-05,
+      "loss": 0.1972,
+      "step": 1111
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.5764075067024135e-05,
+      "loss": 0.2151,
+      "step": 1112
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.5750670241286863e-05,
+      "loss": 0.3725,
+      "step": 1113
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.57372654155496e-05,
+      "loss": 0.0227,
+      "step": 1114
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.5723860589812334e-05,
+      "loss": 0.2063,
+      "step": 1115
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.571045576407507e-05,
+      "loss": 0.0316,
+      "step": 1116
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.5697050938337804e-05,
+      "loss": 0.1545,
+      "step": 1117
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.568364611260054e-05,
+      "loss": 0.4582,
+      "step": 1118
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.567024128686327e-05,
+      "loss": 0.5452,
+      "step": 1119
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.565683646112601e-05,
+      "loss": 0.2009,
+      "step": 1120
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.564343163538874e-05,
+      "loss": 0.1201,
+      "step": 1121
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.563002680965148e-05,
+      "loss": 0.5343,
+      "step": 1122
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.561662198391421e-05,
+      "loss": 0.3925,
+      "step": 1123
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.560321715817694e-05,
+      "loss": 0.0375,
+      "step": 1124
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.558981233243968e-05,
+      "loss": 0.0411,
+      "step": 1125
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.5576407506702413e-05,
+      "loss": 0.0338,
+      "step": 1126
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.556300268096515e-05,
+      "loss": 0.035,
+      "step": 1127
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.5549597855227884e-05,
+      "loss": 0.0283,
+      "step": 1128
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.553619302949062e-05,
+      "loss": 0.034,
+      "step": 1129
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.5522788203753354e-05,
+      "loss": 0.518,
+      "step": 1130
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.550938337801609e-05,
+      "loss": 0.0241,
+      "step": 1131
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.5495978552278824e-05,
+      "loss": 0.3143,
+      "step": 1132
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.548257372654156e-05,
+      "loss": 0.8011,
+      "step": 1133
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.546916890080429e-05,
+      "loss": 0.4365,
+      "step": 1134
+    },
+    {
+      "epoch": 3.0,
+      "eval_f1": 0.7867219917012448,
+      "eval_loss": 0.613310694694519,
+      "eval_runtime": 1.9007,
+      "eval_samples_per_second": 796.041,
+      "eval_steps_per_second": 49.983,
+      "step": 1134
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.545576407506703e-05,
+      "loss": 0.0111,
+      "step": 1135
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.544235924932976e-05,
+      "loss": 0.0166,
+      "step": 1136
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.54289544235925e-05,
+      "loss": 0.2159,
+      "step": 1137
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.541554959785523e-05,
+      "loss": 0.0096,
+      "step": 1138
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.540214477211796e-05,
+      "loss": 0.1352,
+      "step": 1139
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.53887399463807e-05,
+      "loss": 0.0195,
+      "step": 1140
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.5375335120643434e-05,
+      "loss": 0.1579,
+      "step": 1141
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.536193029490617e-05,
+      "loss": 0.0078,
+      "step": 1142
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.5348525469168904e-05,
+      "loss": 0.0111,
+      "step": 1143
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 3.533512064343163e-05,
+      "loss": 0.2457,
+      "step": 1144
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 3.5321715817694374e-05,
+      "loss": 0.014,
+      "step": 1145
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 3.53083109919571e-05,
+      "loss": 0.2021,
+      "step": 1146
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 3.5294906166219844e-05,
+      "loss": 0.5334,
+      "step": 1147
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 3.528150134048257e-05,
+      "loss": 0.0116,
+      "step": 1148
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 3.526809651474531e-05,
+      "loss": 0.0099,
+      "step": 1149
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 3.525469168900804e-05,
+      "loss": 0.2102,
+      "step": 1150
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 3.524128686327078e-05,
+      "loss": 0.0093,
+      "step": 1151
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 3.522788203753351e-05,
+      "loss": 0.0112,
+      "step": 1152
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 3.521447721179625e-05,
+      "loss": 0.1761,
+      "step": 1153
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 3.5201072386058984e-05,
+      "loss": 0.1608,
+      "step": 1154
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 3.518766756032172e-05,
+      "loss": 0.2883,
+      "step": 1155
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 3.5174262734584454e-05,
+      "loss": 0.0304,
+      "step": 1156
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 3.516085790884719e-05,
+      "loss": 0.0623,
+      "step": 1157
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 3.5147453083109924e-05,
+      "loss": 0.1824,
+      "step": 1158
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 3.513404825737265e-05,
+      "loss": 0.2527,
+      "step": 1159
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 3.5120643431635394e-05,
+      "loss": 0.0877,
+      "step": 1160
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 3.510723860589812e-05,
+      "loss": 0.2735,
+      "step": 1161
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 3.5093833780160865e-05,
+      "loss": 0.1126,
+      "step": 1162
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 3.508042895442359e-05,
+      "loss": 0.2498,
+      "step": 1163
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 3.506702412868633e-05,
+      "loss": 0.022,
+      "step": 1164
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 3.505361930294906e-05,
+      "loss": 0.2768,
+      "step": 1165
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 3.50402144772118e-05,
+      "loss": 0.0429,
+      "step": 1166
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 3.5026809651474533e-05,
+      "loss": 0.0198,
+      "step": 1167
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 3.501340482573727e-05,
+      "loss": 0.0097,
+      "step": 1168
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 3.5e-05,
+      "loss": 0.0276,
+      "step": 1169
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 3.498659517426274e-05,
+      "loss": 0.2276,
+      "step": 1170
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 3.497319034852547e-05,
+      "loss": 0.0461,
+      "step": 1171
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 3.495978552278821e-05,
+      "loss": 0.0103,
+      "step": 1172
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 3.494638069705094e-05,
+      "loss": 0.1455,
+      "step": 1173
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 3.493297587131367e-05,
+      "loss": 0.0865,
+      "step": 1174
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 3.491957104557641e-05,
+      "loss": 0.3226,
+      "step": 1175
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 3.490616621983914e-05,
+      "loss": 0.1744,
+      "step": 1176
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 3.489276139410188e-05,
+      "loss": 0.0148,
+      "step": 1177
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 3.487935656836461e-05,
+      "loss": 0.2582,
+      "step": 1178
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 3.486595174262735e-05,
+      "loss": 0.2782,
+      "step": 1179
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 3.485254691689008e-05,
+      "loss": 0.143,
+      "step": 1180
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 3.483914209115282e-05,
+      "loss": 0.0853,
+      "step": 1181
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 3.4825737265415554e-05,
+      "loss": 0.1361,
+      "step": 1182
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 3.481233243967829e-05,
+      "loss": 0.0883,
+      "step": 1183
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 3.479892761394102e-05,
+      "loss": 0.0116,
+      "step": 1184
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 3.478552278820376e-05,
+      "loss": 0.0531,
+      "step": 1185
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 3.477211796246649e-05,
+      "loss": 0.0184,
+      "step": 1186
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 3.475871313672923e-05,
+      "loss": 0.1601,
+      "step": 1187
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 3.474530831099196e-05,
+      "loss": 0.007,
+      "step": 1188
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 3.473190348525469e-05,
+      "loss": 0.0101,
+      "step": 1189
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 3.471849865951743e-05,
+      "loss": 0.2385,
+      "step": 1190
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 3.470509383378016e-05,
+      "loss": 0.0075,
+      "step": 1191
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 3.46916890080429e-05,
+      "loss": 0.0919,
+      "step": 1192
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.467828418230563e-05,
+      "loss": 0.0162,
+      "step": 1193
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.466487935656836e-05,
+      "loss": 0.2239,
+      "step": 1194
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.4651474530831104e-05,
+      "loss": 0.5757,
+      "step": 1195
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.463806970509383e-05,
+      "loss": 0.0774,
+      "step": 1196
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 3.4624664879356574e-05,
+      "loss": 0.2124,
+      "step": 1197
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 3.46112600536193e-05,
+      "loss": 0.0107,
+      "step": 1198
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 3.459785522788204e-05,
+      "loss": 0.3179,
+      "step": 1199
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 3.458445040214477e-05,
+      "loss": 0.0138,
+      "step": 1200
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 3.457104557640751e-05,
+      "loss": 0.0094,
+      "step": 1201
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 3.455764075067024e-05,
+      "loss": 0.0039,
+      "step": 1202
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 3.454423592493298e-05,
+      "loss": 0.0745,
+      "step": 1203
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 3.453083109919571e-05,
+      "loss": 0.0387,
+      "step": 1204
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 3.451742627345845e-05,
+      "loss": 0.1378,
+      "step": 1205
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 3.450402144772118e-05,
+      "loss": 0.1299,
+      "step": 1206
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 3.449061662198392e-05,
+      "loss": 0.2821,
+      "step": 1207
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 3.4477211796246653e-05,
+      "loss": 0.2236,
+      "step": 1208
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 3.446380697050938e-05,
+      "loss": 0.1436,
+      "step": 1209
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 3.4450402144772124e-05,
+      "loss": 0.1504,
+      "step": 1210
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 3.443699731903485e-05,
+      "loss": 0.0415,
+      "step": 1211
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 3.4423592493297594e-05,
+      "loss": 0.023,
+      "step": 1212
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 3.441018766756032e-05,
+      "loss": 0.2128,
+      "step": 1213
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 3.439678284182306e-05,
+      "loss": 0.0066,
+      "step": 1214
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 3.438337801608579e-05,
+      "loss": 0.4345,
+      "step": 1215
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.436997319034853e-05,
+      "loss": 0.0214,
+      "step": 1216
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.435656836461126e-05,
+      "loss": 0.2094,
+      "step": 1217
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.4343163538874e-05,
+      "loss": 0.0822,
+      "step": 1218
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.4329758713136726e-05,
+      "loss": 0.1153,
+      "step": 1219
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.431635388739947e-05,
+      "loss": 0.0059,
+      "step": 1220
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.43029490616622e-05,
+      "loss": 0.0069,
+      "step": 1221
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.428954423592494e-05,
+      "loss": 0.044,
+      "step": 1222
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.427613941018767e-05,
+      "loss": 0.1975,
+      "step": 1223
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.42627345844504e-05,
+      "loss": 0.3294,
+      "step": 1224
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.424932975871314e-05,
+      "loss": 0.026,
+      "step": 1225
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.423592493297587e-05,
+      "loss": 0.2666,
+      "step": 1226
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.422252010723861e-05,
+      "loss": 0.0628,
+      "step": 1227
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.420911528150134e-05,
+      "loss": 0.0068,
+      "step": 1228
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.419571045576407e-05,
+      "loss": 0.0144,
+      "step": 1229
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.418230563002681e-05,
+      "loss": 0.0029,
+      "step": 1230
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.416890080428954e-05,
+      "loss": 0.606,
+      "step": 1231
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.415549597855228e-05,
+      "loss": 0.2162,
+      "step": 1232
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.414209115281501e-05,
+      "loss": 0.146,
+      "step": 1233
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.412868632707775e-05,
+      "loss": 0.3649,
+      "step": 1234
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.411528150134048e-05,
+      "loss": 0.0062,
+      "step": 1235
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.410187667560322e-05,
+      "loss": 0.4097,
+      "step": 1236
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.408847184986595e-05,
+      "loss": 0.5354,
+      "step": 1237
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.407506702412869e-05,
+      "loss": 0.6222,
+      "step": 1238
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.406166219839142e-05,
+      "loss": 0.0023,
+      "step": 1239
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.404825737265416e-05,
+      "loss": 0.0247,
+      "step": 1240
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.403485254691689e-05,
+      "loss": 0.0051,
+      "step": 1241
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.402144772117963e-05,
+      "loss": 0.2504,
+      "step": 1242
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.400804289544236e-05,
+      "loss": 0.0195,
+      "step": 1243
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.39946380697051e-05,
+      "loss": 0.3706,
+      "step": 1244
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.398123324396783e-05,
+      "loss": 0.0174,
+      "step": 1245
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.396782841823056e-05,
+      "loss": 0.0068,
+      "step": 1246
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.39544235924933e-05,
+      "loss": 0.3938,
+      "step": 1247
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.394101876675603e-05,
+      "loss": 0.0114,
+      "step": 1248
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.3927613941018774e-05,
+      "loss": 0.0088,
+      "step": 1249
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.39142091152815e-05,
+      "loss": 0.0126,
+      "step": 1250
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.390080428954424e-05,
+      "loss": 0.0091,
+      "step": 1251
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.388739946380697e-05,
+      "loss": 0.0232,
+      "step": 1252
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.387399463806971e-05,
+      "loss": 0.3704,
+      "step": 1253
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.386058981233244e-05,
+      "loss": 0.0112,
+      "step": 1254
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.384718498659518e-05,
+      "loss": 0.1709,
+      "step": 1255
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.3833780160857906e-05,
+      "loss": 0.0109,
+      "step": 1256
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.382037533512065e-05,
+      "loss": 0.2874,
+      "step": 1257
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.3806970509383376e-05,
+      "loss": 0.024,
+      "step": 1258
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.379356568364612e-05,
+      "loss": 0.0131,
+      "step": 1259
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.3780160857908846e-05,
+      "loss": 0.2076,
+      "step": 1260
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.376675603217158e-05,
+      "loss": 0.0083,
+      "step": 1261
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.375335120643432e-05,
+      "loss": 0.0234,
+      "step": 1262
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.373994638069705e-05,
+      "loss": 0.0066,
+      "step": 1263
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.372654155495979e-05,
+      "loss": 0.3983,
+      "step": 1264
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.371313672922252e-05,
+      "loss": 0.0648,
+      "step": 1265
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.369973190348526e-05,
+      "loss": 0.006,
+      "step": 1266
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.368632707774799e-05,
+      "loss": 0.0807,
+      "step": 1267
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.367292225201073e-05,
+      "loss": 0.0975,
+      "step": 1268
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.365951742627346e-05,
+      "loss": 0.2934,
+      "step": 1269
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.36461126005362e-05,
+      "loss": 0.0869,
+      "step": 1270
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.3632707774798926e-05,
+      "loss": 0.1374,
+      "step": 1271
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.361930294906167e-05,
+      "loss": 0.3314,
+      "step": 1272
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.3605898123324396e-05,
+      "loss": 0.0045,
+      "step": 1273
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.359249329758714e-05,
+      "loss": 0.0536,
+      "step": 1274
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.3579088471849867e-05,
+      "loss": 0.0564,
+      "step": 1275
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.35656836461126e-05,
+      "loss": 0.0689,
+      "step": 1276
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.355227882037534e-05,
+      "loss": 0.5177,
+      "step": 1277
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.353887399463807e-05,
+      "loss": 0.0689,
+      "step": 1278
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.352546916890081e-05,
+      "loss": 0.0664,
+      "step": 1279
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.351206434316354e-05,
+      "loss": 0.0614,
+      "step": 1280
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.349865951742627e-05,
+      "loss": 0.1994,
+      "step": 1281
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.348525469168901e-05,
+      "loss": 0.4769,
+      "step": 1282
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.347184986595174e-05,
+      "loss": 0.1851,
+      "step": 1283
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.345844504021448e-05,
+      "loss": 0.0092,
+      "step": 1284
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.344504021447721e-05,
+      "loss": 0.0052,
+      "step": 1285
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.3431635388739946e-05,
+      "loss": 0.0095,
+      "step": 1286
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.341823056300268e-05,
+      "loss": 0.0242,
+      "step": 1287
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.3404825737265416e-05,
+      "loss": 0.0565,
+      "step": 1288
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.339142091152815e-05,
+      "loss": 0.2645,
+      "step": 1289
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.337801608579089e-05,
+      "loss": 0.0049,
+      "step": 1290
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 3.336461126005362e-05,
+      "loss": 0.0929,
+      "step": 1291
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 3.335120643431636e-05,
+      "loss": 0.3968,
+      "step": 1292
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 3.333780160857909e-05,
+      "loss": 0.033,
+      "step": 1293
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 3.332439678284183e-05,
+      "loss": 0.007,
+      "step": 1294
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 3.331099195710456e-05,
+      "loss": 0.2552,
+      "step": 1295
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 3.329758713136729e-05,
+      "loss": 0.004,
+      "step": 1296
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 3.328418230563003e-05,
+      "loss": 0.136,
+      "step": 1297
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 3.327077747989276e-05,
+      "loss": 0.1407,
+      "step": 1298
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 3.32573726541555e-05,
+      "loss": 0.0354,
+      "step": 1299
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 3.324396782841823e-05,
+      "loss": 0.6141,
+      "step": 1300
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 3.3230563002680966e-05,
+      "loss": 0.2544,
+      "step": 1301
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 3.32171581769437e-05,
+      "loss": 0.0046,
+      "step": 1302
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 3.320375335120644e-05,
+      "loss": 0.0126,
+      "step": 1303
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 3.319034852546917e-05,
+      "loss": 0.3506,
+      "step": 1304
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 3.317694369973191e-05,
+      "loss": 0.3512,
+      "step": 1305
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.3163538873994635e-05,
+      "loss": 0.3675,
+      "step": 1306
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.315013404825738e-05,
+      "loss": 0.1676,
+      "step": 1307
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.3136729222520106e-05,
+      "loss": 0.0307,
+      "step": 1308
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.312332439678285e-05,
+      "loss": 0.0084,
+      "step": 1309
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 3.3109919571045576e-05,
+      "loss": 0.1977,
+      "step": 1310
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 3.309651474530831e-05,
+      "loss": 0.1645,
+      "step": 1311
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 3.3083109919571046e-05,
+      "loss": 0.2579,
+      "step": 1312
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 3.306970509383378e-05,
+      "loss": 0.1656,
+      "step": 1313
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.3056300268096516e-05,
+      "loss": 0.0168,
+      "step": 1314
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.304289544235925e-05,
+      "loss": 0.0291,
+      "step": 1315
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.302949061662198e-05,
+      "loss": 0.0146,
+      "step": 1316
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.301608579088472e-05,
+      "loss": 0.0037,
+      "step": 1317
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 3.300268096514745e-05,
+      "loss": 0.0113,
+      "step": 1318
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 3.298927613941019e-05,
+      "loss": 0.0734,
+      "step": 1319
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 3.297587131367292e-05,
+      "loss": 0.0292,
+      "step": 1320
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 3.2962466487935655e-05,
+      "loss": 0.3875,
+      "step": 1321
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 3.294906166219839e-05,
+      "loss": 0.0138,
+      "step": 1322
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 3.2935656836461126e-05,
+      "loss": 0.4653,
+      "step": 1323
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 3.292225201072386e-05,
+      "loss": 0.1864,
+      "step": 1324
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.2908847184986596e-05,
+      "loss": 0.0116,
+      "step": 1325
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.289544235924933e-05,
+      "loss": 0.014,
+      "step": 1326
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.2882037533512066e-05,
+      "loss": 0.3344,
+      "step": 1327
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.28686327077748e-05,
+      "loss": 0.1544,
+      "step": 1328
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.2855227882037537e-05,
+      "loss": 0.0065,
+      "step": 1329
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.284182305630027e-05,
+      "loss": 0.0041,
+      "step": 1330
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.2828418230563e-05,
+      "loss": 0.0044,
+      "step": 1331
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.281501340482574e-05,
+      "loss": 0.1808,
+      "step": 1332
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.280160857908847e-05,
+      "loss": 0.0521,
+      "step": 1333
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.278820375335121e-05,
+      "loss": 0.3505,
+      "step": 1334
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.277479892761394e-05,
+      "loss": 0.2032,
+      "step": 1335
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.2761394101876676e-05,
+      "loss": 0.004,
+      "step": 1336
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.274798927613941e-05,
+      "loss": 0.0343,
+      "step": 1337
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.2734584450402146e-05,
+      "loss": 0.278,
+      "step": 1338
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.272117962466488e-05,
+      "loss": 0.0056,
+      "step": 1339
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.2707774798927616e-05,
+      "loss": 0.1673,
+      "step": 1340
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 3.2694369973190345e-05,
+      "loss": 0.0092,
+      "step": 1341
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 3.2680965147453086e-05,
+      "loss": 0.0058,
+      "step": 1342
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 3.2667560321715815e-05,
+      "loss": 0.097,
+      "step": 1343
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.265415549597856e-05,
+      "loss": 0.2138,
+      "step": 1344
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.2640750670241285e-05,
+      "loss": 0.0077,
+      "step": 1345
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.262734584450402e-05,
+      "loss": 0.2294,
+      "step": 1346
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.2613941018766755e-05,
+      "loss": 0.3282,
+      "step": 1347
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.260053619302949e-05,
+      "loss": 0.233,
+      "step": 1348
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.2587131367292226e-05,
+      "loss": 0.0379,
+      "step": 1349
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.257372654155496e-05,
+      "loss": 0.2168,
+      "step": 1350
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.2560321715817696e-05,
+      "loss": 0.0443,
+      "step": 1351
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.254691689008043e-05,
+      "loss": 0.2665,
+      "step": 1352
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.2533512064343166e-05,
+      "loss": 0.0136,
+      "step": 1353
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.25201072386059e-05,
+      "loss": 0.0035,
+      "step": 1354
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.2506702412868636e-05,
+      "loss": 0.2153,
+      "step": 1355
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 3.249329758713137e-05,
+      "loss": 0.088,
+      "step": 1356
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 3.247989276139411e-05,
+      "loss": 0.0074,
+      "step": 1357
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 3.2466487935656835e-05,
+      "loss": 0.0924,
+      "step": 1358
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.245308310991958e-05,
+      "loss": 0.0171,
+      "step": 1359
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.2439678284182305e-05,
+      "loss": 0.0132,
+      "step": 1360
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.242627345844505e-05,
+      "loss": 0.0583,
+      "step": 1361
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.2412868632707776e-05,
+      "loss": 0.0038,
+      "step": 1362
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.239946380697051e-05,
+      "loss": 0.0846,
+      "step": 1363
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.2386058981233246e-05,
+      "loss": 0.0058,
+      "step": 1364
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.237265415549598e-05,
+      "loss": 0.4456,
+      "step": 1365
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.2359249329758716e-05,
+      "loss": 0.0029,
+      "step": 1366
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.234584450402145e-05,
+      "loss": 0.2553,
+      "step": 1367
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.233243967828418e-05,
+      "loss": 0.0936,
+      "step": 1368
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.231903485254692e-05,
+      "loss": 0.1017,
+      "step": 1369
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.230563002680965e-05,
+      "loss": 0.0379,
+      "step": 1370
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.229222520107239e-05,
+      "loss": 0.0069,
+      "step": 1371
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.227882037533512e-05,
+      "loss": 0.3235,
+      "step": 1372
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.2265415549597855e-05,
+      "loss": 0.3796,
+      "step": 1373
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.225201072386059e-05,
+      "loss": 0.3246,
+      "step": 1374
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.2238605898123325e-05,
+      "loss": 0.0059,
+      "step": 1375
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.222520107238606e-05,
+      "loss": 0.0405,
+      "step": 1376
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.2211796246648796e-05,
+      "loss": 0.0142,
+      "step": 1377
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.219839142091153e-05,
+      "loss": 0.4426,
+      "step": 1378
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.2184986595174266e-05,
+      "loss": 0.0249,
+      "step": 1379
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.2171581769437e-05,
+      "loss": 0.1053,
+      "step": 1380
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.2158176943699736e-05,
+      "loss": 0.0179,
+      "step": 1381
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.214477211796247e-05,
+      "loss": 0.0718,
+      "step": 1382
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.21313672922252e-05,
+      "loss": 0.1431,
+      "step": 1383
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.211796246648794e-05,
+      "loss": 0.2391,
+      "step": 1384
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.210455764075067e-05,
+      "loss": 0.0053,
+      "step": 1385
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.209115281501341e-05,
+      "loss": 0.2935,
+      "step": 1386
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.207774798927614e-05,
+      "loss": 0.0071,
+      "step": 1387
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.2064343163538875e-05,
+      "loss": 0.031,
+      "step": 1388
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.205093833780161e-05,
+      "loss": 0.1989,
+      "step": 1389
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.2037533512064346e-05,
+      "loss": 0.0533,
+      "step": 1390
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.202412868632708e-05,
+      "loss": 0.2408,
+      "step": 1391
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.2010723860589816e-05,
+      "loss": 0.3158,
+      "step": 1392
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.1997319034852544e-05,
+      "loss": 0.3629,
+      "step": 1393
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.1983914209115286e-05,
+      "loss": 0.0122,
+      "step": 1394
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.1970509383378014e-05,
+      "loss": 0.0449,
+      "step": 1395
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.1957104557640756e-05,
+      "loss": 0.1273,
+      "step": 1396
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.1943699731903485e-05,
+      "loss": 0.3401,
+      "step": 1397
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.193029490616622e-05,
+      "loss": 0.0183,
+      "step": 1398
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.1916890080428955e-05,
+      "loss": 0.0526,
+      "step": 1399
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.190348525469169e-05,
+      "loss": 0.5037,
+      "step": 1400
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.1890080428954425e-05,
+      "loss": 0.0059,
+      "step": 1401
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.187667560321716e-05,
+      "loss": 0.0266,
+      "step": 1402
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.1863270777479896e-05,
+      "loss": 0.4095,
+      "step": 1403
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.184986595174263e-05,
+      "loss": 0.1802,
+      "step": 1404
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.1836461126005366e-05,
+      "loss": 0.3586,
+      "step": 1405
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.18230563002681e-05,
+      "loss": 0.2058,
+      "step": 1406
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.1809651474530836e-05,
+      "loss": 0.008,
+      "step": 1407
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.1796246648793564e-05,
+      "loss": 0.0282,
+      "step": 1408
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.1782841823056306e-05,
+      "loss": 0.0077,
+      "step": 1409
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.1769436997319035e-05,
+      "loss": 0.3461,
+      "step": 1410
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.1756032171581777e-05,
+      "loss": 0.0038,
+      "step": 1411
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.1742627345844505e-05,
+      "loss": 0.0087,
+      "step": 1412
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.172922252010724e-05,
+      "loss": 0.8254,
+      "step": 1413
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.1715817694369975e-05,
+      "loss": 0.017,
+      "step": 1414
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.170241286863271e-05,
+      "loss": 0.2954,
+      "step": 1415
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.1689008042895445e-05,
+      "loss": 0.0286,
+      "step": 1416
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.167560321715818e-05,
+      "loss": 0.0454,
+      "step": 1417
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.166219839142091e-05,
+      "loss": 0.222,
+      "step": 1418
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.164879356568365e-05,
+      "loss": 0.0225,
+      "step": 1419
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.163538873994638e-05,
+      "loss": 0.2599,
+      "step": 1420
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.162198391420912e-05,
+      "loss": 0.2343,
+      "step": 1421
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.160857908847185e-05,
+      "loss": 0.0274,
+      "step": 1422
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.1595174262734585e-05,
+      "loss": 0.0109,
+      "step": 1423
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.158176943699732e-05,
+      "loss": 0.012,
+      "step": 1424
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.1568364611260055e-05,
+      "loss": 0.0267,
+      "step": 1425
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.155495978552279e-05,
+      "loss": 0.0116,
+      "step": 1426
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.1541554959785525e-05,
+      "loss": 0.2563,
+      "step": 1427
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.1528150134048253e-05,
+      "loss": 0.2149,
+      "step": 1428
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.1514745308310995e-05,
+      "loss": 0.2099,
+      "step": 1429
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.1501340482573724e-05,
+      "loss": 0.1445,
+      "step": 1430
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.1487935656836466e-05,
+      "loss": 0.0069,
+      "step": 1431
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.1474530831099194e-05,
+      "loss": 0.3583,
+      "step": 1432
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.146112600536193e-05,
+      "loss": 0.1112,
+      "step": 1433
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.1447721179624664e-05,
+      "loss": 0.5379,
+      "step": 1434
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.14343163538874e-05,
+      "loss": 0.0248,
+      "step": 1435
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.1420911528150135e-05,
+      "loss": 0.0255,
+      "step": 1436
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.140750670241287e-05,
+      "loss": 0.3363,
+      "step": 1437
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.1394101876675605e-05,
+      "loss": 0.2952,
+      "step": 1438
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.138069705093834e-05,
+      "loss": 0.0337,
+      "step": 1439
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.1367292225201075e-05,
+      "loss": 0.0157,
+      "step": 1440
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.135388739946381e-05,
+      "loss": 0.0204,
+      "step": 1441
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.1340482573726545e-05,
+      "loss": 0.7707,
+      "step": 1442
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.1327077747989274e-05,
+      "loss": 0.4232,
+      "step": 1443
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.1313672922252016e-05,
+      "loss": 0.116,
+      "step": 1444
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.1300268096514744e-05,
+      "loss": 0.421,
+      "step": 1445
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.1286863270777486e-05,
+      "loss": 0.0267,
+      "step": 1446
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.1273458445040214e-05,
+      "loss": 0.0078,
+      "step": 1447
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.126005361930295e-05,
+      "loss": 0.0996,
+      "step": 1448
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.1246648793565684e-05,
+      "loss": 0.0389,
+      "step": 1449
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.123324396782842e-05,
+      "loss": 0.0482,
+      "step": 1450
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.1219839142091155e-05,
+      "loss": 0.0053,
+      "step": 1451
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.120643431635389e-05,
+      "loss": 0.0153,
+      "step": 1452
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.119302949061662e-05,
+      "loss": 0.008,
+      "step": 1453
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.117962466487936e-05,
+      "loss": 0.0166,
+      "step": 1454
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.116621983914209e-05,
+      "loss": 0.0889,
+      "step": 1455
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.115281501340483e-05,
+      "loss": 0.0695,
+      "step": 1456
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.113941018766756e-05,
+      "loss": 0.3353,
+      "step": 1457
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.1126005361930294e-05,
+      "loss": 0.0729,
+      "step": 1458
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.111260053619303e-05,
+      "loss": 0.0187,
+      "step": 1459
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.1099195710455764e-05,
+      "loss": 0.2512,
+      "step": 1460
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 3.10857908847185e-05,
+      "loss": 0.3837,
+      "step": 1461
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 3.1072386058981234e-05,
+      "loss": 0.2543,
+      "step": 1462
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 3.105898123324397e-05,
+      "loss": 0.1797,
+      "step": 1463
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 3.1045576407506705e-05,
+      "loss": 0.3097,
+      "step": 1464
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.103217158176944e-05,
+      "loss": 0.268,
+      "step": 1465
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.1018766756032175e-05,
+      "loss": 0.1773,
+      "step": 1466
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.100536193029491e-05,
+      "loss": 0.2055,
+      "step": 1467
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.099195710455764e-05,
+      "loss": 0.0279,
+      "step": 1468
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 3.097855227882038e-05,
+      "loss": 0.1263,
+      "step": 1469
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 3.096514745308311e-05,
+      "loss": 0.0449,
+      "step": 1470
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 3.095174262734585e-05,
+      "loss": 0.2429,
+      "step": 1471
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 3.093833780160858e-05,
+      "loss": 0.1245,
+      "step": 1472
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.0924932975871314e-05,
+      "loss": 0.1303,
+      "step": 1473
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.091152815013405e-05,
+      "loss": 0.0303,
+      "step": 1474
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.0898123324396784e-05,
+      "loss": 0.3279,
+      "step": 1475
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.088471849865952e-05,
+      "loss": 0.134,
+      "step": 1476
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 3.0871313672922255e-05,
+      "loss": 0.5138,
+      "step": 1477
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 3.085790884718498e-05,
+      "loss": 0.0476,
+      "step": 1478
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 3.0844504021447725e-05,
+      "loss": 0.1956,
+      "step": 1479
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.083109919571045e-05,
+      "loss": 0.2061,
+      "step": 1480
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.0817694369973195e-05,
+      "loss": 0.269,
+      "step": 1481
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.0804289544235923e-05,
+      "loss": 0.0708,
+      "step": 1482
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.0790884718498665e-05,
+      "loss": 0.0389,
+      "step": 1483
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.0777479892761394e-05,
+      "loss": 0.2566,
+      "step": 1484
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.076407506702413e-05,
+      "loss": 0.0581,
+      "step": 1485
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.0750670241286864e-05,
+      "loss": 0.1527,
+      "step": 1486
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.07372654155496e-05,
+      "loss": 0.3963,
+      "step": 1487
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 3.0723860589812334e-05,
+      "loss": 0.2241,
+      "step": 1488
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 3.071045576407507e-05,
+      "loss": 0.1275,
+      "step": 1489
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 3.0697050938337804e-05,
+      "loss": 0.3148,
+      "step": 1490
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 3.068364611260054e-05,
+      "loss": 0.1474,
+      "step": 1491
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 3.0670241286863275e-05,
+      "loss": 0.0233,
+      "step": 1492
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 3.065683646112601e-05,
+      "loss": 0.1721,
+      "step": 1493
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 3.0643431635388745e-05,
+      "loss": 0.6024,
+      "step": 1494
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.063002680965147e-05,
+      "loss": 0.1425,
+      "step": 1495
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.0616621983914215e-05,
+      "loss": 0.0311,
+      "step": 1496
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.0603217158176944e-05,
+      "loss": 0.0197,
+      "step": 1497
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.0589812332439686e-05,
+      "loss": 0.0406,
+      "step": 1498
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.0576407506702414e-05,
+      "loss": 0.054,
+      "step": 1499
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.056300268096515e-05,
+      "loss": 0.161,
+      "step": 1500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.0549597855227884e-05,
+      "loss": 0.0549,
+      "step": 1501
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.053619302949062e-05,
+      "loss": 0.1667,
+      "step": 1502
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.0522788203753354e-05,
+      "loss": 0.1264,
+      "step": 1503
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.0509383378016086e-05,
+      "loss": 0.0133,
+      "step": 1504
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.049597855227882e-05,
+      "loss": 0.0655,
+      "step": 1505
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.0482573726541556e-05,
+      "loss": 0.1054,
+      "step": 1506
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.046916890080429e-05,
+      "loss": 0.0053,
+      "step": 1507
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.0455764075067027e-05,
+      "loss": 0.0347,
+      "step": 1508
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.0442359249329762e-05,
+      "loss": 0.6095,
+      "step": 1509
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.0428954423592494e-05,
+      "loss": 0.1339,
+      "step": 1510
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.0415549597855232e-05,
+      "loss": 0.0088,
+      "step": 1511
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.0402144772117964e-05,
+      "loss": 0.4356,
+      "step": 1512
+    },
+    {
+      "epoch": 4.0,
+      "eval_f1": 0.7822580645161291,
+      "eval_loss": 0.6966613531112671,
+      "eval_runtime": 1.8703,
+      "eval_samples_per_second": 808.957,
+      "eval_steps_per_second": 50.794,
+      "step": 1512
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.0388739946380702e-05,
+      "loss": 0.003,
+      "step": 1513
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.0375335120643434e-05,
+      "loss": 0.0067,
+      "step": 1514
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.0361930294906166e-05,
+      "loss": 0.0488,
+      "step": 1515
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.0348525469168904e-05,
+      "loss": 0.0106,
+      "step": 1516
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.0335120643431636e-05,
+      "loss": 0.0098,
+      "step": 1517
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.0321715817694375e-05,
+      "loss": 0.274,
+      "step": 1518
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.0308310991957106e-05,
+      "loss": 0.2007,
+      "step": 1519
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.0294906166219838e-05,
+      "loss": 0.0121,
+      "step": 1520
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.0281501340482577e-05,
+      "loss": 0.0632,
+      "step": 1521
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.026809651474531e-05,
+      "loss": 0.0062,
+      "step": 1522
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.0254691689008047e-05,
+      "loss": 0.0123,
+      "step": 1523
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.024128686327078e-05,
+      "loss": 0.0063,
+      "step": 1524
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.022788203753351e-05,
+      "loss": 0.0102,
+      "step": 1525
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.021447721179625e-05,
+      "loss": 0.0082,
+      "step": 1526
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.020107238605898e-05,
+      "loss": 0.3369,
+      "step": 1527
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.018766756032172e-05,
+      "loss": 0.2587,
+      "step": 1528
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.017426273458445e-05,
+      "loss": 0.0067,
+      "step": 1529
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.0160857908847186e-05,
+      "loss": 0.0021,
+      "step": 1530
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.014745308310992e-05,
+      "loss": 0.0724,
+      "step": 1531
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.0134048257372656e-05,
+      "loss": 0.0074,
+      "step": 1532
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.012064343163539e-05,
+      "loss": 0.0202,
+      "step": 1533
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.0107238605898126e-05,
+      "loss": 0.1435,
+      "step": 1534
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.0093833780160858e-05,
+      "loss": 0.0074,
+      "step": 1535
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.0080428954423597e-05,
+      "loss": 0.4145,
+      "step": 1536
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.006702412868633e-05,
+      "loss": 0.0186,
+      "step": 1537
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.0053619302949067e-05,
+      "loss": 0.1648,
+      "step": 1538
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.00402144772118e-05,
+      "loss": 0.2545,
+      "step": 1539
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.002680965147453e-05,
+      "loss": 0.0016,
+      "step": 1540
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.001340482573727e-05,
+      "loss": 0.0184,
+      "step": 1541
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3e-05,
+      "loss": 0.1208,
+      "step": 1542
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 2.998659517426274e-05,
+      "loss": 0.0021,
+      "step": 1543
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 2.997319034852547e-05,
+      "loss": 0.0092,
+      "step": 1544
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 2.9959785522788203e-05,
+      "loss": 0.1514,
+      "step": 1545
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 2.994638069705094e-05,
+      "loss": 0.0773,
+      "step": 1546
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 2.9932975871313673e-05,
+      "loss": 0.0093,
+      "step": 1547
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.991957104557641e-05,
+      "loss": 0.0022,
+      "step": 1548
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.9906166219839143e-05,
+      "loss": 0.1765,
+      "step": 1549
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.9892761394101875e-05,
+      "loss": 0.1766,
+      "step": 1550
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.9879356568364614e-05,
+      "loss": 0.0024,
+      "step": 1551
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9865951742627345e-05,
+      "loss": 0.012,
+      "step": 1552
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9852546916890084e-05,
+      "loss": 0.0055,
+      "step": 1553
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9839142091152816e-05,
+      "loss": 0.0088,
+      "step": 1554
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 2.9825737265415547e-05,
+      "loss": 0.0019,
+      "step": 1555
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.9812332439678286e-05,
+      "loss": 0.0186,
+      "step": 1556
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.9798927613941018e-05,
+      "loss": 0.25,
+      "step": 1557
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.9785522788203756e-05,
+      "loss": 0.0129,
+      "step": 1558
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.9772117962466488e-05,
+      "loss": 0.0048,
+      "step": 1559
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.9758713136729223e-05,
+      "loss": 0.1153,
+      "step": 1560
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.9745308310991958e-05,
+      "loss": 0.1871,
+      "step": 1561
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.9731903485254693e-05,
+      "loss": 0.0087,
+      "step": 1562
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.971849865951743e-05,
+      "loss": 0.0048,
+      "step": 1563
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.9705093833780163e-05,
+      "loss": 0.026,
+      "step": 1564
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.9691689008042895e-05,
+      "loss": 0.3336,
+      "step": 1565
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.9678284182305634e-05,
+      "loss": 0.0015,
+      "step": 1566
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.9664879356568365e-05,
+      "loss": 0.0044,
+      "step": 1567
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.9651474530831104e-05,
+      "loss": 0.0035,
+      "step": 1568
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.9638069705093836e-05,
+      "loss": 0.1206,
+      "step": 1569
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.9624664879356567e-05,
+      "loss": 0.1247,
+      "step": 1570
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.9611260053619306e-05,
+      "loss": 0.0011,
+      "step": 1571
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.9597855227882038e-05,
+      "loss": 0.0023,
+      "step": 1572
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.9584450402144776e-05,
+      "loss": 0.0014,
+      "step": 1573
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.9571045576407508e-05,
+      "loss": 0.2967,
+      "step": 1574
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.955764075067024e-05,
+      "loss": 0.0373,
+      "step": 1575
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.9544235924932978e-05,
+      "loss": 0.3351,
+      "step": 1576
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.953083109919571e-05,
+      "loss": 0.0025,
+      "step": 1577
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.951742627345845e-05,
+      "loss": 0.0025,
+      "step": 1578
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.950402144772118e-05,
+      "loss": 0.0182,
+      "step": 1579
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.9490616621983912e-05,
+      "loss": 0.001,
+      "step": 1580
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.947721179624665e-05,
+      "loss": 0.003,
+      "step": 1581
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.9463806970509382e-05,
+      "loss": 0.0038,
+      "step": 1582
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.945040214477212e-05,
+      "loss": 0.002,
+      "step": 1583
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.9436997319034853e-05,
+      "loss": 0.1688,
+      "step": 1584
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.9423592493297584e-05,
+      "loss": 0.0014,
+      "step": 1585
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.9410187667560323e-05,
+      "loss": 0.2664,
+      "step": 1586
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.9396782841823055e-05,
+      "loss": 0.0012,
+      "step": 1587
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.9383378016085793e-05,
+      "loss": 0.0022,
+      "step": 1588
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.9369973190348525e-05,
+      "loss": 0.0959,
+      "step": 1589
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.935656836461126e-05,
+      "loss": 0.0839,
+      "step": 1590
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.9343163538873995e-05,
+      "loss": 0.7405,
+      "step": 1591
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.932975871313673e-05,
+      "loss": 0.0351,
+      "step": 1592
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.9316353887399465e-05,
+      "loss": 0.0025,
+      "step": 1593
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.93029490616622e-05,
+      "loss": 0.0054,
+      "step": 1594
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.9289544235924932e-05,
+      "loss": 0.0043,
+      "step": 1595
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.927613941018767e-05,
+      "loss": 0.1828,
+      "step": 1596
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.9262734584450402e-05,
+      "loss": 0.0022,
+      "step": 1597
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.924932975871314e-05,
+      "loss": 0.0051,
+      "step": 1598
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.9235924932975873e-05,
+      "loss": 0.0025,
+      "step": 1599
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.9222520107238604e-05,
+      "loss": 0.0018,
+      "step": 1600
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.9209115281501343e-05,
+      "loss": 0.0348,
+      "step": 1601
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.9195710455764075e-05,
+      "loss": 0.207,
+      "step": 1602
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.9182305630026813e-05,
+      "loss": 0.0249,
+      "step": 1603
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.9168900804289545e-05,
+      "loss": 0.0028,
+      "step": 1604
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.9155495978552283e-05,
+      "loss": 0.2604,
+      "step": 1605
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.9142091152815015e-05,
+      "loss": 0.2808,
+      "step": 1606
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.9128686327077747e-05,
+      "loss": 0.0289,
+      "step": 1607
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.9115281501340486e-05,
+      "loss": 0.005,
+      "step": 1608
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.9101876675603217e-05,
+      "loss": 0.7931,
+      "step": 1609
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.9088471849865956e-05,
+      "loss": 0.335,
+      "step": 1610
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.9075067024128688e-05,
+      "loss": 0.2779,
+      "step": 1611
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.906166219839142e-05,
+      "loss": 0.1649,
+      "step": 1612
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.9048257372654158e-05,
+      "loss": 0.0081,
+      "step": 1613
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.903485254691689e-05,
+      "loss": 0.0638,
+      "step": 1614
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.9021447721179628e-05,
+      "loss": 0.016,
+      "step": 1615
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.900804289544236e-05,
+      "loss": 0.0025,
+      "step": 1616
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.8994638069705095e-05,
+      "loss": 0.0249,
+      "step": 1617
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.898123324396783e-05,
+      "loss": 0.0291,
+      "step": 1618
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.8967828418230565e-05,
+      "loss": 0.1773,
+      "step": 1619
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.89544235924933e-05,
+      "loss": 0.3452,
+      "step": 1620
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.8941018766756035e-05,
+      "loss": 0.006,
+      "step": 1621
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.8927613941018767e-05,
+      "loss": 0.0054,
+      "step": 1622
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 2.8914209115281506e-05,
+      "loss": 0.1852,
+      "step": 1623
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.8900804289544237e-05,
+      "loss": 0.4424,
+      "step": 1624
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.8887399463806976e-05,
+      "loss": 0.0063,
+      "step": 1625
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.8873994638069708e-05,
+      "loss": 0.43,
+      "step": 1626
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 2.886058981233244e-05,
+      "loss": 0.2283,
+      "step": 1627
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.8847184986595178e-05,
+      "loss": 0.0519,
+      "step": 1628
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.883378016085791e-05,
+      "loss": 0.1797,
+      "step": 1629
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.8820375335120648e-05,
+      "loss": 0.2569,
+      "step": 1630
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 2.880697050938338e-05,
+      "loss": 0.0024,
+      "step": 1631
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 2.8793565683646112e-05,
+      "loss": 0.1727,
+      "step": 1632
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 2.878016085790885e-05,
+      "loss": 0.0091,
+      "step": 1633
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 2.8766756032171582e-05,
+      "loss": 0.2002,
+      "step": 1634
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.875335120643432e-05,
+      "loss": 0.0217,
+      "step": 1635
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.8739946380697052e-05,
+      "loss": 0.2163,
+      "step": 1636
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.8726541554959784e-05,
+      "loss": 0.0065,
+      "step": 1637
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 2.8713136729222522e-05,
+      "loss": 0.1567,
+      "step": 1638
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.8699731903485254e-05,
+      "loss": 0.1775,
+      "step": 1639
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.8686327077747993e-05,
+      "loss": 0.0116,
+      "step": 1640
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.8672922252010724e-05,
+      "loss": 0.0114,
+      "step": 1641
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 2.8659517426273456e-05,
+      "loss": 0.0264,
+      "step": 1642
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.8646112600536195e-05,
+      "loss": 0.0172,
+      "step": 1643
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.8632707774798926e-05,
+      "loss": 0.187,
+      "step": 1644
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.8619302949061665e-05,
+      "loss": 0.009,
+      "step": 1645
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.8605898123324397e-05,
+      "loss": 0.014,
+      "step": 1646
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.8592493297587132e-05,
+      "loss": 0.1643,
+      "step": 1647
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.8579088471849867e-05,
+      "loss": 0.2763,
+      "step": 1648
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.8565683646112602e-05,
+      "loss": 0.0641,
+      "step": 1649
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.8552278820375337e-05,
+      "loss": 0.6128,
+      "step": 1650
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.8538873994638072e-05,
+      "loss": 0.0229,
+      "step": 1651
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.8525469168900804e-05,
+      "loss": 0.0344,
+      "step": 1652
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 2.8512064343163543e-05,
+      "loss": 0.018,
+      "step": 1653
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.8498659517426274e-05,
+      "loss": 0.191,
+      "step": 1654
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.8485254691689013e-05,
+      "loss": 0.0397,
+      "step": 1655
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.8471849865951745e-05,
+      "loss": 0.0029,
+      "step": 1656
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 2.8458445040214476e-05,
+      "loss": 0.0034,
+      "step": 1657
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.8445040214477215e-05,
+      "loss": 0.0031,
+      "step": 1658
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.8431635388739947e-05,
+      "loss": 0.4272,
+      "step": 1659
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.8418230563002685e-05,
+      "loss": 0.0042,
+      "step": 1660
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 2.8404825737265417e-05,
+      "loss": 0.0224,
+      "step": 1661
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.839142091152815e-05,
+      "loss": 0.1021,
+      "step": 1662
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.8378016085790887e-05,
+      "loss": 0.0076,
+      "step": 1663
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.836461126005362e-05,
+      "loss": 0.084,
+      "step": 1664
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 2.8351206434316357e-05,
+      "loss": 0.0321,
+      "step": 1665
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 2.833780160857909e-05,
+      "loss": 0.1369,
+      "step": 1666
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 2.832439678284182e-05,
+      "loss": 0.018,
+      "step": 1667
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 2.831099195710456e-05,
+      "loss": 0.1886,
+      "step": 1668
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.829758713136729e-05,
+      "loss": 0.0016,
+      "step": 1669
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.828418230563003e-05,
+      "loss": 0.0031,
+      "step": 1670
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.827077747989276e-05,
+      "loss": 0.0043,
+      "step": 1671
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 2.8257372654155497e-05,
+      "loss": 0.1202,
+      "step": 1672
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.8243967828418232e-05,
+      "loss": 0.1409,
+      "step": 1673
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.8230563002680967e-05,
+      "loss": 0.0821,
+      "step": 1674
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.8217158176943702e-05,
+      "loss": 0.0468,
+      "step": 1675
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.8203753351206437e-05,
+      "loss": 0.0559,
+      "step": 1676
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.819034852546917e-05,
+      "loss": 0.0192,
+      "step": 1677
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.8176943699731907e-05,
+      "loss": 0.0024,
+      "step": 1678
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.816353887399464e-05,
+      "loss": 0.0021,
+      "step": 1679
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 2.8150134048257378e-05,
+      "loss": 0.0139,
+      "step": 1680
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.813672922252011e-05,
+      "loss": 0.0042,
+      "step": 1681
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.812332439678284e-05,
+      "loss": 0.1666,
+      "step": 1682
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.810991957104558e-05,
+      "loss": 0.5925,
+      "step": 1683
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.809651474530831e-05,
+      "loss": 0.1689,
+      "step": 1684
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.808310991957105e-05,
+      "loss": 0.0053,
+      "step": 1685
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.806970509383378e-05,
+      "loss": 0.0019,
+      "step": 1686
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 2.8056300268096513e-05,
+      "loss": 0.0632,
+      "step": 1687
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8042895442359252e-05,
+      "loss": 0.0115,
+      "step": 1688
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8029490616621984e-05,
+      "loss": 0.002,
+      "step": 1689
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8016085790884722e-05,
+      "loss": 0.0021,
+      "step": 1690
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.8002680965147454e-05,
+      "loss": 0.0079,
+      "step": 1691
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.7989276139410186e-05,
+      "loss": 0.0016,
+      "step": 1692
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.7975871313672924e-05,
+      "loss": 0.1824,
+      "step": 1693
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.7962466487935656e-05,
+      "loss": 0.1025,
+      "step": 1694
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 2.7949061662198394e-05,
+      "loss": 0.4274,
+      "step": 1695
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.7935656836461126e-05,
+      "loss": 0.0834,
+      "step": 1696
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.7922252010723858e-05,
+      "loss": 0.6412,
+      "step": 1697
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.7908847184986596e-05,
+      "loss": 0.3051,
+      "step": 1698
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 2.7895442359249328e-05,
+      "loss": 0.0909,
+      "step": 1699
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.7882037533512067e-05,
+      "loss": 0.2655,
+      "step": 1700
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.78686327077748e-05,
+      "loss": 0.305,
+      "step": 1701
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 2.7855227882037534e-05,
+      "loss": 0.2733,
+      "step": 1702
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.784182305630027e-05,
+      "loss": 0.0021,
+      "step": 1703
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.7828418230563004e-05,
+      "loss": 0.0072,
+      "step": 1704
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.781501340482574e-05,
+      "loss": 0.0027,
+      "step": 1705
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 2.7801608579088474e-05,
+      "loss": 0.184,
+      "step": 1706
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.7788203753351206e-05,
+      "loss": 0.0143,
+      "step": 1707
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.7774798927613944e-05,
+      "loss": 0.0297,
+      "step": 1708
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.7761394101876676e-05,
+      "loss": 0.0739,
+      "step": 1709
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.7747989276139415e-05,
+      "loss": 0.0188,
+      "step": 1710
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.7734584450402146e-05,
+      "loss": 0.2487,
+      "step": 1711
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.7721179624664878e-05,
+      "loss": 0.0222,
+      "step": 1712
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.7707774798927617e-05,
+      "loss": 0.0041,
+      "step": 1713
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.769436997319035e-05,
+      "loss": 0.0164,
+      "step": 1714
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.7680965147453087e-05,
+      "loss": 0.0985,
+      "step": 1715
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.766756032171582e-05,
+      "loss": 0.0067,
+      "step": 1716
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.765415549597855e-05,
+      "loss": 0.3304,
+      "step": 1717
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.764075067024129e-05,
+      "loss": 0.006,
+      "step": 1718
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.762734584450402e-05,
+      "loss": 0.0142,
+      "step": 1719
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.761394101876676e-05,
+      "loss": 0.2205,
+      "step": 1720
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.760053619302949e-05,
+      "loss": 0.298,
+      "step": 1721
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.7587131367292223e-05,
+      "loss": 0.0041,
+      "step": 1722
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.757372654155496e-05,
+      "loss": 0.0018,
+      "step": 1723
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.7560321715817693e-05,
+      "loss": 0.0185,
+      "step": 1724
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.754691689008043e-05,
+      "loss": 0.0042,
+      "step": 1725
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.7533512064343163e-05,
+      "loss": 0.036,
+      "step": 1726
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.7520107238605898e-05,
+      "loss": 0.2593,
+      "step": 1727
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.7506702412868633e-05,
+      "loss": 0.0062,
+      "step": 1728
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.749329758713137e-05,
+      "loss": 0.1759,
+      "step": 1729
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.7479892761394104e-05,
+      "loss": 0.0202,
+      "step": 1730
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.746648793565684e-05,
+      "loss": 0.2156,
+      "step": 1731
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.7453083109919574e-05,
+      "loss": 0.4112,
+      "step": 1732
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.743967828418231e-05,
+      "loss": 0.0037,
+      "step": 1733
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.742627345844504e-05,
+      "loss": 0.0186,
+      "step": 1734
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.741286863270778e-05,
+      "loss": 0.0117,
+      "step": 1735
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.739946380697051e-05,
+      "loss": 0.0039,
+      "step": 1736
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.738605898123325e-05,
+      "loss": 0.1185,
+      "step": 1737
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.737265415549598e-05,
+      "loss": 0.0276,
+      "step": 1738
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.7359249329758713e-05,
+      "loss": 0.0041,
+      "step": 1739
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.734584450402145e-05,
+      "loss": 0.0133,
+      "step": 1740
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.7332439678284183e-05,
+      "loss": 0.1042,
+      "step": 1741
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.7319034852546922e-05,
+      "loss": 0.0023,
+      "step": 1742
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.7305630026809654e-05,
+      "loss": 0.1586,
+      "step": 1743
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.7292225201072385e-05,
+      "loss": 0.0258,
+      "step": 1744
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.7278820375335124e-05,
+      "loss": 0.1119,
+      "step": 1745
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.7265415549597856e-05,
+      "loss": 0.1115,
+      "step": 1746
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.7252010723860594e-05,
+      "loss": 0.4607,
+      "step": 1747
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.7238605898123326e-05,
+      "loss": 0.0296,
+      "step": 1748
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.7225201072386058e-05,
+      "loss": 0.0277,
+      "step": 1749
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.7211796246648796e-05,
+      "loss": 0.0777,
+      "step": 1750
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.7198391420911528e-05,
+      "loss": 0.0031,
+      "step": 1751
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.7184986595174266e-05,
+      "loss": 0.2238,
+      "step": 1752
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.7171581769436998e-05,
+      "loss": 0.0409,
+      "step": 1753
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.715817694369973e-05,
+      "loss": 0.0032,
+      "step": 1754
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.714477211796247e-05,
+      "loss": 0.0113,
+      "step": 1755
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.71313672922252e-05,
+      "loss": 0.0204,
+      "step": 1756
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.711796246648794e-05,
+      "loss": 0.0022,
+      "step": 1757
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.710455764075067e-05,
+      "loss": 0.0018,
+      "step": 1758
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.7091152815013406e-05,
+      "loss": 0.263,
+      "step": 1759
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.707774798927614e-05,
+      "loss": 0.0109,
+      "step": 1760
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.7064343163538876e-05,
+      "loss": 0.0653,
+      "step": 1761
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.705093833780161e-05,
+      "loss": 0.0116,
+      "step": 1762
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.7037533512064346e-05,
+      "loss": 0.0063,
+      "step": 1763
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.7024128686327078e-05,
+      "loss": 0.0034,
+      "step": 1764
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.7010723860589816e-05,
+      "loss": 0.0395,
+      "step": 1765
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.6997319034852548e-05,
+      "loss": 0.0014,
+      "step": 1766
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.6983914209115287e-05,
+      "loss": 0.0057,
+      "step": 1767
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.697050938337802e-05,
+      "loss": 0.0018,
+      "step": 1768
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.695710455764075e-05,
+      "loss": 0.012,
+      "step": 1769
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.694369973190349e-05,
+      "loss": 0.0017,
+      "step": 1770
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.693029490616622e-05,
+      "loss": 0.0654,
+      "step": 1771
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.691689008042896e-05,
+      "loss": 0.8002,
+      "step": 1772
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.690348525469169e-05,
+      "loss": 0.0035,
+      "step": 1773
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.6890080428954422e-05,
+      "loss": 0.0051,
+      "step": 1774
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.687667560321716e-05,
+      "loss": 0.0031,
+      "step": 1775
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.6863270777479893e-05,
+      "loss": 0.0142,
+      "step": 1776
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.684986595174263e-05,
+      "loss": 0.0009,
+      "step": 1777
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.6836461126005363e-05,
+      "loss": 0.0015,
+      "step": 1778
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.6823056300268095e-05,
+      "loss": 0.3481,
+      "step": 1779
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.6809651474530833e-05,
+      "loss": 0.3095,
+      "step": 1780
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.6796246648793565e-05,
+      "loss": 0.2567,
+      "step": 1781
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.6782841823056303e-05,
+      "loss": 0.0037,
+      "step": 1782
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.6769436997319035e-05,
+      "loss": 0.001,
+      "step": 1783
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.675603217158177e-05,
+      "loss": 0.0065,
+      "step": 1784
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.6742627345844505e-05,
+      "loss": 0.0029,
+      "step": 1785
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.672922252010724e-05,
+      "loss": 0.6096,
+      "step": 1786
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.6715817694369976e-05,
+      "loss": 0.0127,
+      "step": 1787
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.670241286863271e-05,
+      "loss": 0.0031,
+      "step": 1788
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.6689008042895443e-05,
+      "loss": 0.2463,
+      "step": 1789
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.667560321715818e-05,
+      "loss": 0.1022,
+      "step": 1790
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.6662198391420913e-05,
+      "loss": 0.002,
+      "step": 1791
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.664879356568365e-05,
+      "loss": 0.1576,
+      "step": 1792
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.6635388739946383e-05,
+      "loss": 0.1099,
+      "step": 1793
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.6621983914209115e-05,
+      "loss": 0.1482,
+      "step": 1794
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.6608579088471853e-05,
+      "loss": 0.0007,
+      "step": 1795
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.6595174262734585e-05,
+      "loss": 0.0009,
+      "step": 1796
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.6581769436997324e-05,
+      "loss": 0.005,
+      "step": 1797
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.6568364611260055e-05,
+      "loss": 0.1808,
+      "step": 1798
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.6554959785522787e-05,
+      "loss": 0.0351,
+      "step": 1799
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.6541554959785526e-05,
+      "loss": 0.2555,
+      "step": 1800
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.6528150134048257e-05,
+      "loss": 0.2236,
+      "step": 1801
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.6514745308310996e-05,
+      "loss": 0.3208,
+      "step": 1802
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.6501340482573728e-05,
+      "loss": 0.0202,
+      "step": 1803
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.648793565683646e-05,
+      "loss": 0.0033,
+      "step": 1804
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.6474530831099198e-05,
+      "loss": 0.001,
+      "step": 1805
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.646112600536193e-05,
+      "loss": 0.0019,
+      "step": 1806
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.6447721179624668e-05,
+      "loss": 0.0027,
+      "step": 1807
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.64343163538874e-05,
+      "loss": 0.0051,
+      "step": 1808
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.642091152815013e-05,
+      "loss": 0.1994,
+      "step": 1809
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.640750670241287e-05,
+      "loss": 0.0372,
+      "step": 1810
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.6394101876675602e-05,
+      "loss": 0.0678,
+      "step": 1811
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.638069705093834e-05,
+      "loss": 0.0252,
+      "step": 1812
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.6367292225201072e-05,
+      "loss": 0.0065,
+      "step": 1813
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.6353887399463807e-05,
+      "loss": 0.0045,
+      "step": 1814
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.6340482573726542e-05,
+      "loss": 0.0037,
+      "step": 1815
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.6327077747989277e-05,
+      "loss": 0.0251,
+      "step": 1816
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.6313672922252013e-05,
+      "loss": 0.4196,
+      "step": 1817
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.6300268096514748e-05,
+      "loss": 0.0071,
+      "step": 1818
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.628686327077748e-05,
+      "loss": 0.0787,
+      "step": 1819
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.6273458445040218e-05,
+      "loss": 0.0145,
+      "step": 1820
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.626005361930295e-05,
+      "loss": 0.009,
+      "step": 1821
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.6246648793565688e-05,
+      "loss": 0.0027,
+      "step": 1822
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.623324396782842e-05,
+      "loss": 0.0017,
+      "step": 1823
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.6219839142091152e-05,
+      "loss": 0.4824,
+      "step": 1824
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.620643431635389e-05,
+      "loss": 0.0022,
+      "step": 1825
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.6193029490616622e-05,
+      "loss": 0.3223,
+      "step": 1826
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.617962466487936e-05,
+      "loss": 0.2195,
+      "step": 1827
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.6166219839142092e-05,
+      "loss": 0.0013,
+      "step": 1828
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.6152815013404824e-05,
+      "loss": 0.0343,
+      "step": 1829
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.6139410187667563e-05,
+      "loss": 0.0022,
+      "step": 1830
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.6126005361930294e-05,
+      "loss": 0.0022,
+      "step": 1831
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.6112600536193033e-05,
+      "loss": 0.4116,
+      "step": 1832
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.6099195710455765e-05,
+      "loss": 0.0048,
+      "step": 1833
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.6085790884718496e-05,
+      "loss": 0.5819,
+      "step": 1834
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.6072386058981235e-05,
+      "loss": 0.1985,
+      "step": 1835
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.6058981233243967e-05,
+      "loss": 0.0989,
+      "step": 1836
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.6045576407506705e-05,
+      "loss": 0.341,
+      "step": 1837
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.6032171581769437e-05,
+      "loss": 0.0044,
+      "step": 1838
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.601876675603217e-05,
+      "loss": 0.004,
+      "step": 1839
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.6005361930294907e-05,
+      "loss": 0.2858,
+      "step": 1840
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.599195710455764e-05,
+      "loss": 0.0009,
+      "step": 1841
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.5978552278820377e-05,
+      "loss": 0.0042,
+      "step": 1842
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.596514745308311e-05,
+      "loss": 0.0045,
+      "step": 1843
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.5951742627345844e-05,
+      "loss": 0.0144,
+      "step": 1844
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.593833780160858e-05,
+      "loss": 0.0084,
+      "step": 1845
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.5924932975871314e-05,
+      "loss": 0.4276,
+      "step": 1846
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.591152815013405e-05,
+      "loss": 0.0122,
+      "step": 1847
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.5898123324396785e-05,
+      "loss": 0.0776,
+      "step": 1848
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.5884718498659516e-05,
+      "loss": 0.0117,
+      "step": 1849
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.5871313672922255e-05,
+      "loss": 0.2809,
+      "step": 1850
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.5857908847184987e-05,
+      "loss": 0.0413,
+      "step": 1851
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.5844504021447725e-05,
+      "loss": 0.0187,
+      "step": 1852
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.5831099195710457e-05,
+      "loss": 0.452,
+      "step": 1853
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.5817694369973195e-05,
+      "loss": 0.0206,
+      "step": 1854
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.5804289544235927e-05,
+      "loss": 0.1639,
+      "step": 1855
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.579088471849866e-05,
+      "loss": 0.1865,
+      "step": 1856
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.5777479892761398e-05,
+      "loss": 0.0022,
+      "step": 1857
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.576407506702413e-05,
+      "loss": 0.1167,
+      "step": 1858
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.5750670241286868e-05,
+      "loss": 0.4013,
+      "step": 1859
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.57372654155496e-05,
+      "loss": 0.2355,
+      "step": 1860
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.572386058981233e-05,
+      "loss": 0.0076,
+      "step": 1861
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.571045576407507e-05,
+      "loss": 0.1612,
+      "step": 1862
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.56970509383378e-05,
+      "loss": 0.0047,
+      "step": 1863
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.568364611260054e-05,
+      "loss": 0.1511,
+      "step": 1864
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.5670241286863272e-05,
+      "loss": 0.011,
+      "step": 1865
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.5656836461126004e-05,
+      "loss": 0.1761,
+      "step": 1866
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.5643431635388742e-05,
+      "loss": 0.004,
+      "step": 1867
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.5630026809651474e-05,
+      "loss": 0.0036,
+      "step": 1868
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.5616621983914212e-05,
+      "loss": 0.4345,
+      "step": 1869
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.5603217158176944e-05,
+      "loss": 0.0034,
+      "step": 1870
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.558981233243968e-05,
+      "loss": 0.1269,
+      "step": 1871
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.5576407506702414e-05,
+      "loss": 0.183,
+      "step": 1872
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.556300268096515e-05,
+      "loss": 0.008,
+      "step": 1873
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.5549597855227885e-05,
+      "loss": 0.0035,
+      "step": 1874
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.553619302949062e-05,
+      "loss": 0.0133,
+      "step": 1875
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.552278820375335e-05,
+      "loss": 0.2156,
+      "step": 1876
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.550938337801609e-05,
+      "loss": 0.0043,
+      "step": 1877
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.549597855227882e-05,
+      "loss": 0.2614,
+      "step": 1878
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.548257372654156e-05,
+      "loss": 0.0208,
+      "step": 1879
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.5469168900804292e-05,
+      "loss": 0.0228,
+      "step": 1880
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.5455764075067024e-05,
+      "loss": 0.0105,
+      "step": 1881
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.5442359249329762e-05,
+      "loss": 0.0108,
+      "step": 1882
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.5428954423592494e-05,
+      "loss": 0.3828,
+      "step": 1883
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.5415549597855232e-05,
+      "loss": 0.0093,
+      "step": 1884
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.5402144772117964e-05,
+      "loss": 0.0231,
+      "step": 1885
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.5388739946380696e-05,
+      "loss": 0.0082,
+      "step": 1886
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.5375335120643434e-05,
+      "loss": 0.1796,
+      "step": 1887
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.5361930294906166e-05,
+      "loss": 0.0753,
+      "step": 1888
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.5348525469168905e-05,
+      "loss": 0.0142,
+      "step": 1889
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.5335120643431636e-05,
+      "loss": 0.0047,
+      "step": 1890
+    },
+    {
+      "epoch": 5.0,
+      "eval_f1": 0.7775974025974025,
+      "eval_loss": 0.953689694404602,
+      "eval_runtime": 1.8696,
+      "eval_samples_per_second": 809.285,
+      "eval_steps_per_second": 50.814,
+      "step": 1890
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.5321715817694368e-05,
+      "loss": 0.0014,
+      "step": 1891
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.5308310991957107e-05,
+      "loss": 0.0487,
+      "step": 1892
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.529490616621984e-05,
+      "loss": 0.0037,
+      "step": 1893
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.5281501340482577e-05,
+      "loss": 0.0512,
+      "step": 1894
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.526809651474531e-05,
+      "loss": 0.134,
+      "step": 1895
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.525469168900804e-05,
+      "loss": 0.3762,
+      "step": 1896
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.524128686327078e-05,
+      "loss": 0.0011,
+      "step": 1897
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.522788203753351e-05,
+      "loss": 0.0023,
+      "step": 1898
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.521447721179625e-05,
+      "loss": 0.0526,
+      "step": 1899
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.520107238605898e-05,
+      "loss": 0.0553,
+      "step": 1900
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.5187667560321716e-05,
+      "loss": 0.1773,
+      "step": 1901
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.517426273458445e-05,
+      "loss": 0.451,
+      "step": 1902
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.5160857908847186e-05,
+      "loss": 0.0217,
+      "step": 1903
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.514745308310992e-05,
+      "loss": 0.0728,
+      "step": 1904
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.5134048257372657e-05,
+      "loss": 0.0009,
+      "step": 1905
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.512064343163539e-05,
+      "loss": 0.1018,
+      "step": 1906
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.5107238605898127e-05,
+      "loss": 0.0012,
+      "step": 1907
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.509383378016086e-05,
+      "loss": 0.004,
+      "step": 1908
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.5080428954423597e-05,
+      "loss": 0.0012,
+      "step": 1909
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.506702412868633e-05,
+      "loss": 0.0128,
+      "step": 1910
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.505361930294906e-05,
+      "loss": 0.1116,
+      "step": 1911
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.50402144772118e-05,
+      "loss": 0.0011,
+      "step": 1912
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.502680965147453e-05,
+      "loss": 0.0011,
+      "step": 1913
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.501340482573727e-05,
+      "loss": 0.0897,
+      "step": 1914
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0014,
+      "step": 1915
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.4986595174262736e-05,
+      "loss": 0.0918,
+      "step": 1916
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.497319034852547e-05,
+      "loss": 0.0026,
+      "step": 1917
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.4959785522788203e-05,
+      "loss": 0.0225,
+      "step": 1918
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.494638069705094e-05,
+      "loss": 0.2655,
+      "step": 1919
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.4932975871313673e-05,
+      "loss": 0.0029,
+      "step": 1920
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.491957104557641e-05,
+      "loss": 0.0006,
+      "step": 1921
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.4906166219839144e-05,
+      "loss": 0.0008,
+      "step": 1922
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 2.4892761394101875e-05,
+      "loss": 0.0012,
+      "step": 1923
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 2.487935656836461e-05,
+      "loss": 0.0013,
+      "step": 1924
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 2.4865951742627346e-05,
+      "loss": 0.0524,
+      "step": 1925
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.485254691689008e-05,
+      "loss": 0.0059,
+      "step": 1926
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.4839142091152816e-05,
+      "loss": 0.0026,
+      "step": 1927
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.482573726541555e-05,
+      "loss": 0.0015,
+      "step": 1928
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.4812332439678286e-05,
+      "loss": 0.0073,
+      "step": 1929
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.479892761394102e-05,
+      "loss": 0.0008,
+      "step": 1930
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.4785522788203757e-05,
+      "loss": 0.1519,
+      "step": 1931
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.477211796246649e-05,
+      "loss": 0.008,
+      "step": 1932
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.4758713136729223e-05,
+      "loss": 0.0009,
+      "step": 1933
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 2.474530831099196e-05,
+      "loss": 0.0299,
+      "step": 1934
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 2.4731903485254694e-05,
+      "loss": 0.1637,
+      "step": 1935
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 2.471849865951743e-05,
+      "loss": 0.0369,
+      "step": 1936
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 2.4705093833780164e-05,
+      "loss": 0.0057,
+      "step": 1937
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.4691689008042896e-05,
+      "loss": 0.0035,
+      "step": 1938
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.467828418230563e-05,
+      "loss": 0.0011,
+      "step": 1939
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.4664879356568366e-05,
+      "loss": 0.0018,
+      "step": 1940
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.46514745308311e-05,
+      "loss": 0.0013,
+      "step": 1941
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.4638069705093836e-05,
+      "loss": 0.0779,
+      "step": 1942
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.4624664879356568e-05,
+      "loss": 0.1762,
+      "step": 1943
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.4611260053619303e-05,
+      "loss": 0.0006,
+      "step": 1944
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.4597855227882038e-05,
+      "loss": 0.0037,
+      "step": 1945
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.4584450402144773e-05,
+      "loss": 0.0005,
+      "step": 1946
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.457104557640751e-05,
+      "loss": 0.0397,
+      "step": 1947
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.4557640750670244e-05,
+      "loss": 0.0082,
+      "step": 1948
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.4544235924932975e-05,
+      "loss": 0.0008,
+      "step": 1949
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.453083109919571e-05,
+      "loss": 0.0219,
+      "step": 1950
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.4517426273458446e-05,
+      "loss": 0.3966,
+      "step": 1951
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.450402144772118e-05,
+      "loss": 0.0011,
+      "step": 1952
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.4490616621983916e-05,
+      "loss": 0.3447,
+      "step": 1953
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.4477211796246648e-05,
+      "loss": 0.0006,
+      "step": 1954
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.4463806970509383e-05,
+      "loss": 0.0011,
+      "step": 1955
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.4450402144772118e-05,
+      "loss": 0.0013,
+      "step": 1956
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.4436997319034853e-05,
+      "loss": 0.1495,
+      "step": 1957
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.4423592493297588e-05,
+      "loss": 0.0005,
+      "step": 1958
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.4410187667560323e-05,
+      "loss": 0.3345,
+      "step": 1959
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.439678284182306e-05,
+      "loss": 0.0048,
+      "step": 1960
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.4383378016085793e-05,
+      "loss": 0.001,
+      "step": 1961
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.436997319034853e-05,
+      "loss": 0.0025,
+      "step": 1962
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.4356568364611264e-05,
+      "loss": 0.3215,
+      "step": 1963
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.4343163538873995e-05,
+      "loss": 0.0197,
+      "step": 1964
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.432975871313673e-05,
+      "loss": 0.0018,
+      "step": 1965
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.4316353887399466e-05,
+      "loss": 0.1012,
+      "step": 1966
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.43029490616622e-05,
+      "loss": 0.0179,
+      "step": 1967
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.4289544235924936e-05,
+      "loss": 0.0032,
+      "step": 1968
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.4276139410187668e-05,
+      "loss": 0.0011,
+      "step": 1969
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.4262734584450403e-05,
+      "loss": 0.4875,
+      "step": 1970
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.4249329758713138e-05,
+      "loss": 0.2791,
+      "step": 1971
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.4235924932975873e-05,
+      "loss": 0.0011,
+      "step": 1972
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.4222520107238608e-05,
+      "loss": 0.0011,
+      "step": 1973
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.420911528150134e-05,
+      "loss": 0.0976,
+      "step": 1974
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.4195710455764075e-05,
+      "loss": 0.3669,
+      "step": 1975
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.418230563002681e-05,
+      "loss": 0.0022,
+      "step": 1976
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.4168900804289545e-05,
+      "loss": 0.0015,
+      "step": 1977
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.415549597855228e-05,
+      "loss": 0.0014,
+      "step": 1978
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 2.4142091152815012e-05,
+      "loss": 0.0433,
+      "step": 1979
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 2.4128686327077747e-05,
+      "loss": 0.0019,
+      "step": 1980
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 2.4115281501340483e-05,
+      "loss": 0.0007,
+      "step": 1981
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 2.4101876675603218e-05,
+      "loss": 0.0136,
+      "step": 1982
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.4088471849865953e-05,
+      "loss": 0.1744,
+      "step": 1983
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.4075067024128688e-05,
+      "loss": 0.1557,
+      "step": 1984
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.4061662198391423e-05,
+      "loss": 0.1192,
+      "step": 1985
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.4048257372654158e-05,
+      "loss": 0.0406,
+      "step": 1986
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.4034852546916893e-05,
+      "loss": 0.2243,
+      "step": 1987
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.402144772117963e-05,
+      "loss": 0.0021,
+      "step": 1988
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.400804289544236e-05,
+      "loss": 0.002,
+      "step": 1989
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.3994638069705095e-05,
+      "loss": 0.077,
+      "step": 1990
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 2.398123324396783e-05,
+      "loss": 0.0378,
+      "step": 1991
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 2.3967828418230566e-05,
+      "loss": 0.012,
+      "step": 1992
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 2.39544235924933e-05,
+      "loss": 0.1386,
+      "step": 1993
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.3941018766756032e-05,
+      "loss": 0.002,
+      "step": 1994
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.3927613941018768e-05,
+      "loss": 0.0008,
+      "step": 1995
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.3914209115281503e-05,
+      "loss": 0.0021,
+      "step": 1996
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.3900804289544238e-05,
+      "loss": 0.022,
+      "step": 1997
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 2.3887399463806973e-05,
+      "loss": 0.0015,
+      "step": 1998
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 2.3873994638069705e-05,
+      "loss": 0.1486,
+      "step": 1999
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 2.386058981233244e-05,
+      "loss": 0.2586,
+      "step": 2000
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 2.3847184986595175e-05,
+      "loss": 0.0088,
+      "step": 2001
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.383378016085791e-05,
+      "loss": 0.0044,
+      "step": 2002
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.3820375335120645e-05,
+      "loss": 0.0015,
+      "step": 2003
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.3806970509383377e-05,
+      "loss": 0.0008,
+      "step": 2004
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.3793565683646112e-05,
+      "loss": 0.212,
+      "step": 2005
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.3780160857908847e-05,
+      "loss": 0.0005,
+      "step": 2006
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.3766756032171582e-05,
+      "loss": 0.1511,
+      "step": 2007
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.3753351206434318e-05,
+      "loss": 0.0023,
+      "step": 2008
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.3739946380697053e-05,
+      "loss": 0.1544,
+      "step": 2009
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 2.3726541554959784e-05,
+      "loss": 0.0306,
+      "step": 2010
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 2.371313672922252e-05,
+      "loss": 0.1005,
+      "step": 2011
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 2.3699731903485255e-05,
+      "loss": 0.0744,
+      "step": 2012
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.368632707774799e-05,
+      "loss": 0.0622,
+      "step": 2013
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.3672922252010725e-05,
+      "loss": 0.0192,
+      "step": 2014
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.365951742627346e-05,
+      "loss": 0.016,
+      "step": 2015
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.3646112600536195e-05,
+      "loss": 0.293,
+      "step": 2016
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.363270777479893e-05,
+      "loss": 0.0114,
+      "step": 2017
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.3619302949061665e-05,
+      "loss": 0.1254,
+      "step": 2018
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.36058981233244e-05,
+      "loss": 0.0638,
+      "step": 2019
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.3592493297587132e-05,
+      "loss": 0.0192,
+      "step": 2020
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.3579088471849867e-05,
+      "loss": 0.0069,
+      "step": 2021
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.3565683646112603e-05,
+      "loss": 0.0573,
+      "step": 2022
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.3552278820375338e-05,
+      "loss": 0.0039,
+      "step": 2023
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.3538873994638073e-05,
+      "loss": 0.0079,
+      "step": 2024
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 2.3525469168900805e-05,
+      "loss": 0.0063,
+      "step": 2025
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 2.351206434316354e-05,
+      "loss": 0.0032,
+      "step": 2026
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 2.3498659517426275e-05,
+      "loss": 0.1018,
+      "step": 2027
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.348525469168901e-05,
+      "loss": 0.0017,
+      "step": 2028
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.3471849865951745e-05,
+      "loss": 0.0203,
+      "step": 2029
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.3458445040214477e-05,
+      "loss": 0.0008,
+      "step": 2030
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.3445040214477212e-05,
+      "loss": 0.3145,
+      "step": 2031
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.3431635388739947e-05,
+      "loss": 0.2476,
+      "step": 2032
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.3418230563002682e-05,
+      "loss": 0.0573,
+      "step": 2033
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.3404825737265417e-05,
+      "loss": 0.0008,
+      "step": 2034
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.339142091152815e-05,
+      "loss": 0.0636,
+      "step": 2035
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.3378016085790884e-05,
+      "loss": 0.001,
+      "step": 2036
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.336461126005362e-05,
+      "loss": 0.0079,
+      "step": 2037
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.3351206434316355e-05,
+      "loss": 0.0752,
+      "step": 2038
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.333780160857909e-05,
+      "loss": 0.0024,
+      "step": 2039
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.332439678284182e-05,
+      "loss": 0.0061,
+      "step": 2040
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.3310991957104557e-05,
+      "loss": 0.2882,
+      "step": 2041
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.329758713136729e-05,
+      "loss": 0.02,
+      "step": 2042
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.3284182305630027e-05,
+      "loss": 0.018,
+      "step": 2043
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.3270777479892762e-05,
+      "loss": 0.0125,
+      "step": 2044
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.3257372654155497e-05,
+      "loss": 0.0007,
+      "step": 2045
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.3243967828418232e-05,
+      "loss": 0.3849,
+      "step": 2046
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.3230563002680967e-05,
+      "loss": 0.0011,
+      "step": 2047
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.3217158176943702e-05,
+      "loss": 0.1235,
+      "step": 2048
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.3203753351206438e-05,
+      "loss": 0.0006,
+      "step": 2049
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.319034852546917e-05,
+      "loss": 0.0006,
+      "step": 2050
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.3176943699731904e-05,
+      "loss": 0.0011,
+      "step": 2051
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.316353887399464e-05,
+      "loss": 0.0011,
+      "step": 2052
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.3150134048257375e-05,
+      "loss": 0.0032,
+      "step": 2053
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.313672922252011e-05,
+      "loss": 0.0718,
+      "step": 2054
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.312332439678284e-05,
+      "loss": 0.0052,
+      "step": 2055
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.3109919571045577e-05,
+      "loss": 0.0026,
+      "step": 2056
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.3096514745308312e-05,
+      "loss": 0.1854,
+      "step": 2057
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.3083109919571047e-05,
+      "loss": 0.0008,
+      "step": 2058
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.3069705093833782e-05,
+      "loss": 0.2671,
+      "step": 2059
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.3056300268096514e-05,
+      "loss": 0.0064,
+      "step": 2060
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.304289544235925e-05,
+      "loss": 0.0012,
+      "step": 2061
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 2.3029490616621984e-05,
+      "loss": 0.0006,
+      "step": 2062
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 2.301608579088472e-05,
+      "loss": 0.0008,
+      "step": 2063
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 2.3002680965147454e-05,
+      "loss": 0.4998,
+      "step": 2064
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 2.2989276139410186e-05,
+      "loss": 0.0037,
+      "step": 2065
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.297587131367292e-05,
+      "loss": 0.0029,
+      "step": 2066
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.2962466487935656e-05,
+      "loss": 0.043,
+      "step": 2067
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.294906166219839e-05,
+      "loss": 0.0451,
+      "step": 2068
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.2935656836461127e-05,
+      "loss": 0.0009,
+      "step": 2069
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.2922252010723862e-05,
+      "loss": 0.002,
+      "step": 2070
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.2908847184986597e-05,
+      "loss": 0.2744,
+      "step": 2071
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.2895442359249332e-05,
+      "loss": 0.0146,
+      "step": 2072
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.2882037533512067e-05,
+      "loss": 0.0011,
+      "step": 2073
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.2868632707774802e-05,
+      "loss": 0.0421,
+      "step": 2074
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.2855227882037537e-05,
+      "loss": 0.1518,
+      "step": 2075
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.284182305630027e-05,
+      "loss": 0.0072,
+      "step": 2076
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.2828418230563004e-05,
+      "loss": 0.2781,
+      "step": 2077
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.281501340482574e-05,
+      "loss": 0.004,
+      "step": 2078
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.2801608579088475e-05,
+      "loss": 0.1029,
+      "step": 2079
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.278820375335121e-05,
+      "loss": 0.0526,
+      "step": 2080
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.277479892761394e-05,
+      "loss": 0.0011,
+      "step": 2081
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.2761394101876677e-05,
+      "loss": 0.0139,
+      "step": 2082
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.274798927613941e-05,
+      "loss": 0.0509,
+      "step": 2083
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.2734584450402147e-05,
+      "loss": 0.0042,
+      "step": 2084
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 2.2721179624664882e-05,
+      "loss": 0.1964,
+      "step": 2085
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 2.2707774798927614e-05,
+      "loss": 0.0083,
+      "step": 2086
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 2.269436997319035e-05,
+      "loss": 0.101,
+      "step": 2087
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 2.2680965147453084e-05,
+      "loss": 0.0094,
+      "step": 2088
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.266756032171582e-05,
+      "loss": 0.1433,
+      "step": 2089
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.2654155495978554e-05,
+      "loss": 0.0091,
+      "step": 2090
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.2640750670241286e-05,
+      "loss": 0.0018,
+      "step": 2091
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.262734584450402e-05,
+      "loss": 0.001,
+      "step": 2092
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.2613941018766756e-05,
+      "loss": 0.3507,
+      "step": 2093
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.260053619302949e-05,
+      "loss": 0.0527,
+      "step": 2094
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.2587131367292226e-05,
+      "loss": 0.0015,
+      "step": 2095
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.2573726541554958e-05,
+      "loss": 0.0195,
+      "step": 2096
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.2560321715817693e-05,
+      "loss": 0.0007,
+      "step": 2097
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.254691689008043e-05,
+      "loss": 0.001,
+      "step": 2098
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.2533512064343164e-05,
+      "loss": 0.2777,
+      "step": 2099
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.25201072386059e-05,
+      "loss": 0.0657,
+      "step": 2100
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.2506702412868634e-05,
+      "loss": 0.159,
+      "step": 2101
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.249329758713137e-05,
+      "loss": 0.0342,
+      "step": 2102
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.2479892761394104e-05,
+      "loss": 0.3001,
+      "step": 2103
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.246648793565684e-05,
+      "loss": 0.0028,
+      "step": 2104
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.2453083109919574e-05,
+      "loss": 0.0191,
+      "step": 2105
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.2439678284182306e-05,
+      "loss": 0.0012,
+      "step": 2106
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.242627345844504e-05,
+      "loss": 0.2619,
+      "step": 2107
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.2412868632707776e-05,
+      "loss": 0.001,
+      "step": 2108
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.239946380697051e-05,
+      "loss": 0.094,
+      "step": 2109
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.2386058981233247e-05,
+      "loss": 0.003,
+      "step": 2110
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.237265415549598e-05,
+      "loss": 0.0528,
+      "step": 2111
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.2359249329758714e-05,
+      "loss": 0.1252,
+      "step": 2112
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.234584450402145e-05,
+      "loss": 0.0039,
+      "step": 2113
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.2332439678284184e-05,
+      "loss": 0.0913,
+      "step": 2114
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.231903485254692e-05,
+      "loss": 0.0023,
+      "step": 2115
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.230563002680965e-05,
+      "loss": 0.0047,
+      "step": 2116
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.2292225201072386e-05,
+      "loss": 0.0688,
+      "step": 2117
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.227882037533512e-05,
+      "loss": 0.0013,
+      "step": 2118
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.2265415549597856e-05,
+      "loss": 0.0012,
+      "step": 2119
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.225201072386059e-05,
+      "loss": 0.0048,
+      "step": 2120
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.2238605898123323e-05,
+      "loss": 0.0011,
+      "step": 2121
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.2225201072386058e-05,
+      "loss": 0.0009,
+      "step": 2122
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.2211796246648793e-05,
+      "loss": 0.0015,
+      "step": 2123
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.2198391420911528e-05,
+      "loss": 0.0077,
+      "step": 2124
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.2184986595174263e-05,
+      "loss": 0.0007,
+      "step": 2125
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.2171581769437e-05,
+      "loss": 0.0007,
+      "step": 2126
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.2158176943699734e-05,
+      "loss": 0.0506,
+      "step": 2127
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.214477211796247e-05,
+      "loss": 0.0016,
+      "step": 2128
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.2131367292225204e-05,
+      "loss": 0.0005,
+      "step": 2129
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.211796246648794e-05,
+      "loss": 0.0049,
+      "step": 2130
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.210455764075067e-05,
+      "loss": 0.0305,
+      "step": 2131
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.2091152815013406e-05,
+      "loss": 0.0448,
+      "step": 2132
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.207774798927614e-05,
+      "loss": 0.5391,
+      "step": 2133
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.2064343163538876e-05,
+      "loss": 0.0005,
+      "step": 2134
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.205093833780161e-05,
+      "loss": 0.0141,
+      "step": 2135
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.2037533512064346e-05,
+      "loss": 0.3613,
+      "step": 2136
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.2024128686327078e-05,
+      "loss": 0.0013,
+      "step": 2137
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 2.2010723860589813e-05,
+      "loss": 0.002,
+      "step": 2138
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 2.199731903485255e-05,
+      "loss": 0.0207,
+      "step": 2139
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 2.1983914209115284e-05,
+      "loss": 0.0004,
+      "step": 2140
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 2.197050938337802e-05,
+      "loss": 0.0022,
+      "step": 2141
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.195710455764075e-05,
+      "loss": 0.5076,
+      "step": 2142
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.1943699731903486e-05,
+      "loss": 0.0016,
+      "step": 2143
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.193029490616622e-05,
+      "loss": 0.0014,
+      "step": 2144
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.1916890080428956e-05,
+      "loss": 0.0101,
+      "step": 2145
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.190348525469169e-05,
+      "loss": 0.0048,
+      "step": 2146
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.1890080428954423e-05,
+      "loss": 0.001,
+      "step": 2147
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.1876675603217158e-05,
+      "loss": 0.0004,
+      "step": 2148
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 2.1863270777479893e-05,
+      "loss": 0.2627,
+      "step": 2149
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 2.1849865951742628e-05,
+      "loss": 0.0013,
+      "step": 2150
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 2.1836461126005363e-05,
+      "loss": 0.0074,
+      "step": 2151
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 2.1823056300268095e-05,
+      "loss": 0.0238,
+      "step": 2152
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.180965147453083e-05,
+      "loss": 0.0013,
+      "step": 2153
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.1796246648793565e-05,
+      "loss": 0.0005,
+      "step": 2154
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.17828418230563e-05,
+      "loss": 0.0006,
+      "step": 2155
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.1769436997319036e-05,
+      "loss": 0.0442,
+      "step": 2156
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.175603217158177e-05,
+      "loss": 0.0007,
+      "step": 2157
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.1742627345844506e-05,
+      "loss": 0.0032,
+      "step": 2158
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.172922252010724e-05,
+      "loss": 0.005,
+      "step": 2159
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.1715817694369976e-05,
+      "loss": 0.0005,
+      "step": 2160
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.170241286863271e-05,
+      "loss": 0.0005,
+      "step": 2161
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.1689008042895443e-05,
+      "loss": 0.2023,
+      "step": 2162
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.1675603217158178e-05,
+      "loss": 0.3146,
+      "step": 2163
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.1662198391420913e-05,
+      "loss": 0.0035,
+      "step": 2164
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 2.164879356568365e-05,
+      "loss": 0.141,
+      "step": 2165
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 2.1635388739946383e-05,
+      "loss": 0.0005,
+      "step": 2166
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 2.1621983914209115e-05,
+      "loss": 0.0013,
+      "step": 2167
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.160857908847185e-05,
+      "loss": 0.0484,
+      "step": 2168
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.1595174262734585e-05,
+      "loss": 0.0109,
+      "step": 2169
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.158176943699732e-05,
+      "loss": 0.3307,
+      "step": 2170
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.1568364611260056e-05,
+      "loss": 0.0013,
+      "step": 2171
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.1554959785522787e-05,
+      "loss": 0.0005,
+      "step": 2172
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.1541554959785523e-05,
+      "loss": 0.0004,
+      "step": 2173
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.1528150134048258e-05,
+      "loss": 0.0744,
+      "step": 2174
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.1514745308310993e-05,
+      "loss": 0.0008,
+      "step": 2175
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.1501340482573728e-05,
+      "loss": 0.0013,
+      "step": 2176
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.148793565683646e-05,
+      "loss": 0.0006,
+      "step": 2177
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.1474530831099195e-05,
+      "loss": 0.055,
+      "step": 2178
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.146112600536193e-05,
+      "loss": 0.0013,
+      "step": 2179
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.1447721179624665e-05,
+      "loss": 0.0022,
+      "step": 2180
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.14343163538874e-05,
+      "loss": 0.0009,
+      "step": 2181
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.1420911528150135e-05,
+      "loss": 0.0011,
+      "step": 2182
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.140750670241287e-05,
+      "loss": 0.0676,
+      "step": 2183
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.1394101876675606e-05,
+      "loss": 0.4583,
+      "step": 2184
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.138069705093834e-05,
+      "loss": 0.0009,
+      "step": 2185
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.1367292225201076e-05,
+      "loss": 0.002,
+      "step": 2186
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.1353887399463808e-05,
+      "loss": 0.7625,
+      "step": 2187
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.1340482573726543e-05,
+      "loss": 0.4912,
+      "step": 2188
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.1327077747989278e-05,
+      "loss": 0.0283,
+      "step": 2189
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.1313672922252013e-05,
+      "loss": 0.0011,
+      "step": 2190
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.1300268096514748e-05,
+      "loss": 0.0013,
+      "step": 2191
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.128686327077748e-05,
+      "loss": 0.3384,
+      "step": 2192
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.1273458445040215e-05,
+      "loss": 0.4533,
+      "step": 2193
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.126005361930295e-05,
+      "loss": 0.0039,
+      "step": 2194
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.1246648793565685e-05,
+      "loss": 0.0038,
+      "step": 2195
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.123324396782842e-05,
+      "loss": 0.0318,
+      "step": 2196
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.1219839142091156e-05,
+      "loss": 0.0045,
+      "step": 2197
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.1206434316353887e-05,
+      "loss": 0.3134,
+      "step": 2198
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 2.1193029490616622e-05,
+      "loss": 0.0011,
+      "step": 2199
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 2.1179624664879358e-05,
+      "loss": 0.0008,
+      "step": 2200
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 2.1166219839142093e-05,
+      "loss": 0.0009,
+      "step": 2201
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.1152815013404828e-05,
+      "loss": 0.0083,
+      "step": 2202
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.113941018766756e-05,
+      "loss": 0.0196,
+      "step": 2203
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.1126005361930295e-05,
+      "loss": 0.0063,
+      "step": 2204
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.111260053619303e-05,
+      "loss": 0.0064,
+      "step": 2205
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.1099195710455765e-05,
+      "loss": 0.0143,
+      "step": 2206
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.10857908847185e-05,
+      "loss": 0.0012,
+      "step": 2207
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.1072386058981232e-05,
+      "loss": 0.0033,
+      "step": 2208
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.1058981233243967e-05,
+      "loss": 0.0014,
+      "step": 2209
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.1045576407506702e-05,
+      "loss": 0.0219,
+      "step": 2210
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.1032171581769437e-05,
+      "loss": 0.3033,
+      "step": 2211
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.1018766756032172e-05,
+      "loss": 0.0711,
+      "step": 2212
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.1005361930294907e-05,
+      "loss": 0.0051,
+      "step": 2213
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.0991957104557643e-05,
+      "loss": 0.026,
+      "step": 2214
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.0978552278820378e-05,
+      "loss": 0.0024,
+      "step": 2215
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.0965147453083113e-05,
+      "loss": 0.3622,
+      "step": 2216
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.0951742627345848e-05,
+      "loss": 0.0009,
+      "step": 2217
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.093833780160858e-05,
+      "loss": 0.0019,
+      "step": 2218
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.0924932975871315e-05,
+      "loss": 0.3412,
+      "step": 2219
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.091152815013405e-05,
+      "loss": 0.0197,
+      "step": 2220
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.0898123324396785e-05,
+      "loss": 0.4229,
+      "step": 2221
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.088471849865952e-05,
+      "loss": 0.0014,
+      "step": 2222
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.0871313672922252e-05,
+      "loss": 0.0183,
+      "step": 2223
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.0857908847184987e-05,
+      "loss": 0.2005,
+      "step": 2224
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.0844504021447722e-05,
+      "loss": 0.0122,
+      "step": 2225
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.0831099195710457e-05,
+      "loss": 0.1178,
+      "step": 2226
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.0817694369973193e-05,
+      "loss": 0.0105,
+      "step": 2227
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.0804289544235924e-05,
+      "loss": 0.0328,
+      "step": 2228
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.079088471849866e-05,
+      "loss": 0.0087,
+      "step": 2229
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.0777479892761395e-05,
+      "loss": 0.0288,
+      "step": 2230
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.076407506702413e-05,
+      "loss": 0.0017,
+      "step": 2231
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.0750670241286865e-05,
+      "loss": 0.002,
+      "step": 2232
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 2.0737265415549597e-05,
+      "loss": 0.0024,
+      "step": 2233
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 2.072386058981233e-05,
+      "loss": 0.0008,
+      "step": 2234
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 2.0710455764075067e-05,
+      "loss": 0.002,
+      "step": 2235
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.0697050938337802e-05,
+      "loss": 0.0052,
+      "step": 2236
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.0683646112600537e-05,
+      "loss": 0.0058,
+      "step": 2237
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.0670241286863272e-05,
+      "loss": 0.0013,
+      "step": 2238
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.0656836461126007e-05,
+      "loss": 0.0158,
+      "step": 2239
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.0643431635388742e-05,
+      "loss": 0.229,
+      "step": 2240
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.0630026809651478e-05,
+      "loss": 0.1844,
+      "step": 2241
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.0616621983914213e-05,
+      "loss": 0.2905,
+      "step": 2242
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.0603217158176944e-05,
+      "loss": 0.0059,
+      "step": 2243
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.058981233243968e-05,
+      "loss": 0.0007,
+      "step": 2244
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.0576407506702415e-05,
+      "loss": 0.1638,
+      "step": 2245
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.056300268096515e-05,
+      "loss": 0.1195,
+      "step": 2246
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.0549597855227885e-05,
+      "loss": 0.0015,
+      "step": 2247
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.0536193029490617e-05,
+      "loss": 0.0013,
+      "step": 2248
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.0522788203753352e-05,
+      "loss": 0.5152,
+      "step": 2249
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.0509383378016087e-05,
+      "loss": 0.0315,
+      "step": 2250
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.0495978552278822e-05,
+      "loss": 0.1213,
+      "step": 2251
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.0482573726541557e-05,
+      "loss": 0.0006,
+      "step": 2252
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.046916890080429e-05,
+      "loss": 0.0011,
+      "step": 2253
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.0455764075067024e-05,
+      "loss": 0.2546,
+      "step": 2254
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.044235924932976e-05,
+      "loss": 0.1259,
+      "step": 2255
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.0428954423592494e-05,
+      "loss": 0.0179,
+      "step": 2256
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.041554959785523e-05,
+      "loss": 0.0257,
+      "step": 2257
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.0402144772117965e-05,
+      "loss": 0.0092,
+      "step": 2258
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.0388739946380696e-05,
+      "loss": 0.3231,
+      "step": 2259
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.037533512064343e-05,
+      "loss": 0.0084,
+      "step": 2260
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.0361930294906167e-05,
+      "loss": 0.0517,
+      "step": 2261
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.0348525469168902e-05,
+      "loss": 0.0012,
+      "step": 2262
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.0335120643431637e-05,
+      "loss": 0.0045,
+      "step": 2263
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.032171581769437e-05,
+      "loss": 0.0012,
+      "step": 2264
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.0308310991957104e-05,
+      "loss": 0.0032,
+      "step": 2265
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.029490616621984e-05,
+      "loss": 0.0038,
+      "step": 2266
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.0281501340482574e-05,
+      "loss": 0.2731,
+      "step": 2267
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.026809651474531e-05,
+      "loss": 0.3122,
+      "step": 2268
+    },
+    {
+      "epoch": 6.0,
+      "eval_f1": 0.7762762762762763,
+      "eval_loss": 1.0977023839950562,
+      "eval_runtime": 1.8624,
+      "eval_samples_per_second": 812.383,
+      "eval_steps_per_second": 51.009,
+      "step": 2268
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.0254691689008044e-05,
+      "loss": 0.0035,
+      "step": 2269
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.024128686327078e-05,
+      "loss": 0.0709,
+      "step": 2270
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.0227882037533515e-05,
+      "loss": 0.1548,
+      "step": 2271
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.021447721179625e-05,
+      "loss": 0.0035,
+      "step": 2272
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.0201072386058985e-05,
+      "loss": 0.0009,
+      "step": 2273
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.0187667560321717e-05,
+      "loss": 0.0061,
+      "step": 2274
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.0174262734584452e-05,
+      "loss": 0.3586,
+      "step": 2275
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.0160857908847187e-05,
+      "loss": 0.0019,
+      "step": 2276
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.0147453083109922e-05,
+      "loss": 0.0077,
+      "step": 2277
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.0134048257372657e-05,
+      "loss": 0.0022,
+      "step": 2278
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.012064343163539e-05,
+      "loss": 0.0021,
+      "step": 2279
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.0107238605898124e-05,
+      "loss": 0.0022,
+      "step": 2280
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.009383378016086e-05,
+      "loss": 0.1387,
+      "step": 2281
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.0080428954423594e-05,
+      "loss": 0.0034,
+      "step": 2282
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.006702412868633e-05,
+      "loss": 0.0128,
+      "step": 2283
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.005361930294906e-05,
+      "loss": 0.0201,
+      "step": 2284
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.0040214477211796e-05,
+      "loss": 0.0072,
+      "step": 2285
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 2.002680965147453e-05,
+      "loss": 0.0031,
+      "step": 2286
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 2.0013404825737267e-05,
+      "loss": 0.0273,
+      "step": 2287
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 2288
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.9986595174262733e-05,
+      "loss": 0.0012,
+      "step": 2289
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.997319034852547e-05,
+      "loss": 0.0045,
+      "step": 2290
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.9959785522788204e-05,
+      "loss": 0.0289,
+      "step": 2291
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.994638069705094e-05,
+      "loss": 0.1421,
+      "step": 2292
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.9932975871313674e-05,
+      "loss": 0.0007,
+      "step": 2293
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.9919571045576406e-05,
+      "loss": 0.0037,
+      "step": 2294
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.990616621983914e-05,
+      "loss": 0.1566,
+      "step": 2295
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.9892761394101876e-05,
+      "loss": 0.0008,
+      "step": 2296
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1.987935656836461e-05,
+      "loss": 0.0008,
+      "step": 2297
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1.9865951742627346e-05,
+      "loss": 0.1319,
+      "step": 2298
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1.985254691689008e-05,
+      "loss": 0.0007,
+      "step": 2299
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1.9839142091152816e-05,
+      "loss": 0.0048,
+      "step": 2300
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1.982573726541555e-05,
+      "loss": 0.223,
+      "step": 2301
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1.9812332439678287e-05,
+      "loss": 0.2188,
+      "step": 2302
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1.9798927613941022e-05,
+      "loss": 0.0209,
+      "step": 2303
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1.9785522788203754e-05,
+      "loss": 0.001,
+      "step": 2304
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1.977211796246649e-05,
+      "loss": 0.0006,
+      "step": 2305
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1.9758713136729224e-05,
+      "loss": 0.0011,
+      "step": 2306
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1.974530831099196e-05,
+      "loss": 0.0008,
+      "step": 2307
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.9731903485254694e-05,
+      "loss": 0.0015,
+      "step": 2308
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.9718498659517426e-05,
+      "loss": 0.0024,
+      "step": 2309
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.970509383378016e-05,
+      "loss": 0.0006,
+      "step": 2310
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.9691689008042896e-05,
+      "loss": 0.0007,
+      "step": 2311
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.967828418230563e-05,
+      "loss": 0.0026,
+      "step": 2312
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.9664879356568366e-05,
+      "loss": 0.0019,
+      "step": 2313
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.9651474530831098e-05,
+      "loss": 0.0055,
+      "step": 2314
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.9638069705093833e-05,
+      "loss": 0.0023,
+      "step": 2315
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.962466487935657e-05,
+      "loss": 0.0011,
+      "step": 2316
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.9611260053619303e-05,
+      "loss": 0.0407,
+      "step": 2317
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.959785522788204e-05,
+      "loss": 0.046,
+      "step": 2318
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.958445040214477e-05,
+      "loss": 0.0062,
+      "step": 2319
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1.9571045576407505e-05,
+      "loss": 0.0007,
+      "step": 2320
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1.955764075067024e-05,
+      "loss": 0.3377,
+      "step": 2321
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1.9544235924932976e-05,
+      "loss": 0.4294,
+      "step": 2322
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.953083109919571e-05,
+      "loss": 0.0016,
+      "step": 2323
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.9517426273458446e-05,
+      "loss": 0.0283,
+      "step": 2324
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.950402144772118e-05,
+      "loss": 0.0005,
+      "step": 2325
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.9490616621983916e-05,
+      "loss": 0.0011,
+      "step": 2326
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.947721179624665e-05,
+      "loss": 0.0237,
+      "step": 2327
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.9463806970509387e-05,
+      "loss": 0.0581,
+      "step": 2328
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.945040214477212e-05,
+      "loss": 0.0905,
+      "step": 2329
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.9436997319034853e-05,
+      "loss": 0.0012,
+      "step": 2330
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1.942359249329759e-05,
+      "loss": 0.0053,
+      "step": 2331
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1.9410187667560324e-05,
+      "loss": 0.0225,
+      "step": 2332
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1.939678284182306e-05,
+      "loss": 0.0374,
+      "step": 2333
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1.9383378016085794e-05,
+      "loss": 0.0006,
+      "step": 2334
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1.9369973190348526e-05,
+      "loss": 0.0008,
+      "step": 2335
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1.935656836461126e-05,
+      "loss": 0.0006,
+      "step": 2336
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1.9343163538873996e-05,
+      "loss": 0.0011,
+      "step": 2337
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1.932975871313673e-05,
+      "loss": 0.0014,
+      "step": 2338
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1.9316353887399466e-05,
+      "loss": 0.0006,
+      "step": 2339
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1.9302949061662198e-05,
+      "loss": 0.0504,
+      "step": 2340
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1.9289544235924933e-05,
+      "loss": 0.005,
+      "step": 2341
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.9276139410187668e-05,
+      "loss": 0.2673,
+      "step": 2342
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.9262734584450403e-05,
+      "loss": 0.2173,
+      "step": 2343
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.924932975871314e-05,
+      "loss": 0.0235,
+      "step": 2344
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.923592493297587e-05,
+      "loss": 0.0092,
+      "step": 2345
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1.9222520107238605e-05,
+      "loss": 0.0088,
+      "step": 2346
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1.920911528150134e-05,
+      "loss": 0.0506,
+      "step": 2347
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1.9195710455764076e-05,
+      "loss": 0.0033,
+      "step": 2348
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1.918230563002681e-05,
+      "loss": 0.0006,
+      "step": 2349
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.9168900804289542e-05,
+      "loss": 0.0021,
+      "step": 2350
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.9155495978552278e-05,
+      "loss": 0.0006,
+      "step": 2351
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.9142091152815013e-05,
+      "loss": 0.0041,
+      "step": 2352
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.9128686327077748e-05,
+      "loss": 0.0209,
+      "step": 2353
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.9115281501340483e-05,
+      "loss": 0.0145,
+      "step": 2354
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.9101876675603218e-05,
+      "loss": 0.0303,
+      "step": 2355
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.9088471849865953e-05,
+      "loss": 0.0951,
+      "step": 2356
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1.907506702412869e-05,
+      "loss": 0.042,
+      "step": 2357
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1.9061662198391424e-05,
+      "loss": 0.0009,
+      "step": 2358
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1.904825737265416e-05,
+      "loss": 0.0006,
+      "step": 2359
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1.903485254691689e-05,
+      "loss": 0.0057,
+      "step": 2360
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1.9021447721179626e-05,
+      "loss": 0.0578,
+      "step": 2361
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1.900804289544236e-05,
+      "loss": 0.3295,
+      "step": 2362
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1.8994638069705096e-05,
+      "loss": 0.0005,
+      "step": 2363
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1.898123324396783e-05,
+      "loss": 0.0009,
+      "step": 2364
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.8967828418230563e-05,
+      "loss": 0.0036,
+      "step": 2365
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.8954423592493298e-05,
+      "loss": 0.0029,
+      "step": 2366
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.8941018766756033e-05,
+      "loss": 0.1468,
+      "step": 2367
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.8927613941018768e-05,
+      "loss": 0.3163,
+      "step": 2368
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1.8914209115281503e-05,
+      "loss": 0.169,
+      "step": 2369
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1.8900804289544235e-05,
+      "loss": 0.0493,
+      "step": 2370
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1.888739946380697e-05,
+      "loss": 0.0003,
+      "step": 2371
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.8873994638069705e-05,
+      "loss": 0.0015,
+      "step": 2372
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.886058981233244e-05,
+      "loss": 0.258,
+      "step": 2373
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.8847184986595175e-05,
+      "loss": 0.002,
+      "step": 2374
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.8833780160857907e-05,
+      "loss": 0.0008,
+      "step": 2375
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.8820375335120642e-05,
+      "loss": 0.0003,
+      "step": 2376
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.8806970509383377e-05,
+      "loss": 0.0014,
+      "step": 2377
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.8793565683646113e-05,
+      "loss": 0.0068,
+      "step": 2378
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.8780160857908848e-05,
+      "loss": 0.39,
+      "step": 2379
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1.8766756032171583e-05,
+      "loss": 0.0046,
+      "step": 2380
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1.8753351206434318e-05,
+      "loss": 0.0008,
+      "step": 2381
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1.8739946380697053e-05,
+      "loss": 0.0007,
+      "step": 2382
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1.8726541554959788e-05,
+      "loss": 0.0013,
+      "step": 2383
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.8713136729222523e-05,
+      "loss": 0.0055,
+      "step": 2384
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.869973190348526e-05,
+      "loss": 0.0014,
+      "step": 2385
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.868632707774799e-05,
+      "loss": 0.0039,
+      "step": 2386
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.8672922252010725e-05,
+      "loss": 0.0995,
+      "step": 2387
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1.865951742627346e-05,
+      "loss": 0.0015,
+      "step": 2388
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1.8646112600536196e-05,
+      "loss": 0.083,
+      "step": 2389
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1.863270777479893e-05,
+      "loss": 0.0044,
+      "step": 2390
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1.8619302949061662e-05,
+      "loss": 0.0105,
+      "step": 2391
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1.8605898123324398e-05,
+      "loss": 0.0005,
+      "step": 2392
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1.8592493297587133e-05,
+      "loss": 0.0014,
+      "step": 2393
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1.8579088471849868e-05,
+      "loss": 0.0114,
+      "step": 2394
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.8565683646112603e-05,
+      "loss": 0.0334,
+      "step": 2395
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.8552278820375335e-05,
+      "loss": 0.006,
+      "step": 2396
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.853887399463807e-05,
+      "loss": 0.3124,
+      "step": 2397
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.8525469168900805e-05,
+      "loss": 0.2324,
+      "step": 2398
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1.851206434316354e-05,
+      "loss": 0.0889,
+      "step": 2399
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1.8498659517426275e-05,
+      "loss": 0.0705,
+      "step": 2400
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1.8485254691689007e-05,
+      "loss": 0.0012,
+      "step": 2401
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1.8471849865951742e-05,
+      "loss": 0.0033,
+      "step": 2402
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1.8458445040214477e-05,
+      "loss": 0.0021,
+      "step": 2403
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1.8445040214477212e-05,
+      "loss": 0.0005,
+      "step": 2404
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1.8431635388739948e-05,
+      "loss": 0.0011,
+      "step": 2405
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.841823056300268e-05,
+      "loss": 0.0006,
+      "step": 2406
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.8404825737265414e-05,
+      "loss": 0.0726,
+      "step": 2407
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.839142091152815e-05,
+      "loss": 0.0009,
+      "step": 2408
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.8378016085790885e-05,
+      "loss": 0.0007,
+      "step": 2409
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.836461126005362e-05,
+      "loss": 0.0859,
+      "step": 2410
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.8351206434316355e-05,
+      "loss": 0.0011,
+      "step": 2411
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.833780160857909e-05,
+      "loss": 0.6542,
+      "step": 2412
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.8324396782841825e-05,
+      "loss": 0.2733,
+      "step": 2413
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1.831099195710456e-05,
+      "loss": 0.2825,
+      "step": 2414
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1.8297587131367295e-05,
+      "loss": 0.0012,
+      "step": 2415
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1.8284182305630027e-05,
+      "loss": 0.1404,
+      "step": 2416
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1.8270777479892762e-05,
+      "loss": 0.0006,
+      "step": 2417
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.8257372654155497e-05,
+      "loss": 0.0007,
+      "step": 2418
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.8243967828418233e-05,
+      "loss": 0.1429,
+      "step": 2419
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.8230563002680968e-05,
+      "loss": 0.0008,
+      "step": 2420
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.82171581769437e-05,
+      "loss": 0.0062,
+      "step": 2421
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.8203753351206435e-05,
+      "loss": 0.0071,
+      "step": 2422
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.819034852546917e-05,
+      "loss": 0.0017,
+      "step": 2423
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.8176943699731905e-05,
+      "loss": 0.084,
+      "step": 2424
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.816353887399464e-05,
+      "loss": 0.0011,
+      "step": 2425
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.8150134048257372e-05,
+      "loss": 0.0255,
+      "step": 2426
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.8136729222520107e-05,
+      "loss": 0.0009,
+      "step": 2427
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.8123324396782842e-05,
+      "loss": 0.3105,
+      "step": 2428
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.8109919571045577e-05,
+      "loss": 0.0046,
+      "step": 2429
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.8096514745308312e-05,
+      "loss": 0.0089,
+      "step": 2430
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.8083109919571044e-05,
+      "loss": 0.1176,
+      "step": 2431
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.806970509383378e-05,
+      "loss": 0.0235,
+      "step": 2432
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.8056300268096514e-05,
+      "loss": 0.029,
+      "step": 2433
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.804289544235925e-05,
+      "loss": 0.0013,
+      "step": 2434
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.8029490616621985e-05,
+      "loss": 0.0075,
+      "step": 2435
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.801608579088472e-05,
+      "loss": 0.1744,
+      "step": 2436
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.8002680965147455e-05,
+      "loss": 0.0017,
+      "step": 2437
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.798927613941019e-05,
+      "loss": 0.0188,
+      "step": 2438
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.7975871313672925e-05,
+      "loss": 0.0232,
+      "step": 2439
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.796246648793566e-05,
+      "loss": 0.1459,
+      "step": 2440
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.7949061662198392e-05,
+      "loss": 0.0007,
+      "step": 2441
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.7935656836461127e-05,
+      "loss": 0.0005,
+      "step": 2442
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.7922252010723862e-05,
+      "loss": 0.0012,
+      "step": 2443
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.7908847184986597e-05,
+      "loss": 0.0041,
+      "step": 2444
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.7895442359249332e-05,
+      "loss": 0.4884,
+      "step": 2445
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.7882037533512068e-05,
+      "loss": 0.0017,
+      "step": 2446
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.78686327077748e-05,
+      "loss": 0.0566,
+      "step": 2447
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.7855227882037534e-05,
+      "loss": 0.012,
+      "step": 2448
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.784182305630027e-05,
+      "loss": 0.001,
+      "step": 2449
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.7828418230563005e-05,
+      "loss": 0.0028,
+      "step": 2450
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.781501340482574e-05,
+      "loss": 0.4622,
+      "step": 2451
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.780160857908847e-05,
+      "loss": 0.0042,
+      "step": 2452
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.7788203753351207e-05,
+      "loss": 0.0176,
+      "step": 2453
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.7774798927613942e-05,
+      "loss": 0.0012,
+      "step": 2454
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.7761394101876677e-05,
+      "loss": 0.0344,
+      "step": 2455
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1.7747989276139412e-05,
+      "loss": 0.1278,
+      "step": 2456
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1.7734584450402144e-05,
+      "loss": 0.0017,
+      "step": 2457
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1.772117962466488e-05,
+      "loss": 0.0044,
+      "step": 2458
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.7707774798927614e-05,
+      "loss": 0.0016,
+      "step": 2459
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.769436997319035e-05,
+      "loss": 0.0799,
+      "step": 2460
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.7680965147453084e-05,
+      "loss": 0.0066,
+      "step": 2461
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.7667560321715816e-05,
+      "loss": 0.1607,
+      "step": 2462
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.765415549597855e-05,
+      "loss": 0.0742,
+      "step": 2463
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.7640750670241286e-05,
+      "loss": 0.0005,
+      "step": 2464
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.762734584450402e-05,
+      "loss": 0.0006,
+      "step": 2465
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.7613941018766757e-05,
+      "loss": 0.0019,
+      "step": 2466
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.7600536193029492e-05,
+      "loss": 0.0009,
+      "step": 2467
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.7587131367292227e-05,
+      "loss": 0.0023,
+      "step": 2468
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.7573726541554962e-05,
+      "loss": 0.0202,
+      "step": 2469
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.7560321715817697e-05,
+      "loss": 0.0223,
+      "step": 2470
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.7546916890080432e-05,
+      "loss": 0.0009,
+      "step": 2471
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.7533512064343164e-05,
+      "loss": 0.0014,
+      "step": 2472
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.75201072386059e-05,
+      "loss": 0.0514,
+      "step": 2473
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.7506702412868634e-05,
+      "loss": 0.0013,
+      "step": 2474
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1.749329758713137e-05,
+      "loss": 0.0087,
+      "step": 2475
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1.7479892761394105e-05,
+      "loss": 0.0035,
+      "step": 2476
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1.7466487935656836e-05,
+      "loss": 0.0397,
+      "step": 2477
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.745308310991957e-05,
+      "loss": 0.0021,
+      "step": 2478
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.7439678284182307e-05,
+      "loss": 0.052,
+      "step": 2479
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.742627345844504e-05,
+      "loss": 0.0027,
+      "step": 2480
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.7412868632707777e-05,
+      "loss": 0.001,
+      "step": 2481
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.739946380697051e-05,
+      "loss": 0.2899,
+      "step": 2482
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.7386058981233244e-05,
+      "loss": 0.0007,
+      "step": 2483
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.737265415549598e-05,
+      "loss": 0.0704,
+      "step": 2484
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.7359249329758714e-05,
+      "loss": 0.001,
+      "step": 2485
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.734584450402145e-05,
+      "loss": 0.057,
+      "step": 2486
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.733243967828418e-05,
+      "loss": 0.0002,
+      "step": 2487
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.7319034852546916e-05,
+      "loss": 0.0064,
+      "step": 2488
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.730563002680965e-05,
+      "loss": 0.0638,
+      "step": 2489
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 1.7292225201072386e-05,
+      "loss": 0.0006,
+      "step": 2490
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 1.727882037533512e-05,
+      "loss": 0.0142,
+      "step": 2491
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 1.7265415549597856e-05,
+      "loss": 0.0015,
+      "step": 2492
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.725201072386059e-05,
+      "loss": 0.0949,
+      "step": 2493
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.7238605898123327e-05,
+      "loss": 0.0004,
+      "step": 2494
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.7225201072386062e-05,
+      "loss": 0.0111,
+      "step": 2495
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.7211796246648797e-05,
+      "loss": 0.0215,
+      "step": 2496
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.719839142091153e-05,
+      "loss": 0.266,
+      "step": 2497
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.7184986595174264e-05,
+      "loss": 0.4487,
+      "step": 2498
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.7171581769437e-05,
+      "loss": 0.0021,
+      "step": 2499
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.7158176943699734e-05,
+      "loss": 0.0004,
+      "step": 2500
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1.714477211796247e-05,
+      "loss": 0.0004,
+      "step": 2501
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1.71313672922252e-05,
+      "loss": 0.0011,
+      "step": 2502
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1.7117962466487936e-05,
+      "loss": 0.0006,
+      "step": 2503
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1.710455764075067e-05,
+      "loss": 0.1005,
+      "step": 2504
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.7091152815013406e-05,
+      "loss": 0.0472,
+      "step": 2505
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.707774798927614e-05,
+      "loss": 0.0004,
+      "step": 2506
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.7064343163538877e-05,
+      "loss": 0.0162,
+      "step": 2507
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.705093833780161e-05,
+      "loss": 0.004,
+      "step": 2508
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 1.7037533512064344e-05,
+      "loss": 0.0007,
+      "step": 2509
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 1.702412868632708e-05,
+      "loss": 0.1447,
+      "step": 2510
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 1.7010723860589814e-05,
+      "loss": 0.0006,
+      "step": 2511
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.699731903485255e-05,
+      "loss": 0.0002,
+      "step": 2512
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.698391420911528e-05,
+      "loss": 0.0004,
+      "step": 2513
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.6970509383378016e-05,
+      "loss": 0.0017,
+      "step": 2514
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.695710455764075e-05,
+      "loss": 0.4581,
+      "step": 2515
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.6943699731903486e-05,
+      "loss": 0.0005,
+      "step": 2516
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.693029490616622e-05,
+      "loss": 0.0043,
+      "step": 2517
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.6916890080428953e-05,
+      "loss": 0.0005,
+      "step": 2518
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.6903485254691688e-05,
+      "loss": 0.0002,
+      "step": 2519
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.6890080428954423e-05,
+      "loss": 0.0005,
+      "step": 2520
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.687667560321716e-05,
+      "loss": 0.0037,
+      "step": 2521
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.6863270777479893e-05,
+      "loss": 0.0003,
+      "step": 2522
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.684986595174263e-05,
+      "loss": 0.0019,
+      "step": 2523
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.6836461126005364e-05,
+      "loss": 0.0023,
+      "step": 2524
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.68230563002681e-05,
+      "loss": 0.0004,
+      "step": 2525
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.6809651474530834e-05,
+      "loss": 0.3317,
+      "step": 2526
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.679624664879357e-05,
+      "loss": 0.0004,
+      "step": 2527
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.67828418230563e-05,
+      "loss": 0.002,
+      "step": 2528
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.6769436997319036e-05,
+      "loss": 0.0003,
+      "step": 2529
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.675603217158177e-05,
+      "loss": 0.0007,
+      "step": 2530
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1.6742627345844506e-05,
+      "loss": 0.0012,
+      "step": 2531
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1.672922252010724e-05,
+      "loss": 0.024,
+      "step": 2532
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1.6715817694369973e-05,
+      "loss": 0.0041,
+      "step": 2533
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1.6702412868632708e-05,
+      "loss": 0.1821,
+      "step": 2534
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.6689008042895443e-05,
+      "loss": 0.0004,
+      "step": 2535
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.667560321715818e-05,
+      "loss": 0.0293,
+      "step": 2536
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.6662198391420914e-05,
+      "loss": 0.0005,
+      "step": 2537
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.6648793565683645e-05,
+      "loss": 0.0043,
+      "step": 2538
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1.663538873994638e-05,
+      "loss": 0.0112,
+      "step": 2539
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1.6621983914209116e-05,
+      "loss": 0.0003,
+      "step": 2540
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1.660857908847185e-05,
+      "loss": 0.0005,
+      "step": 2541
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1.6595174262734586e-05,
+      "loss": 0.4617,
+      "step": 2542
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1.6581769436997318e-05,
+      "loss": 0.0004,
+      "step": 2543
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1.6568364611260053e-05,
+      "loss": 0.1932,
+      "step": 2544
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1.6554959785522788e-05,
+      "loss": 0.0004,
+      "step": 2545
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.6541554959785523e-05,
+      "loss": 0.0785,
+      "step": 2546
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.6528150134048258e-05,
+      "loss": 0.0882,
+      "step": 2547
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.651474530831099e-05,
+      "loss": 0.3937,
+      "step": 2548
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.6501340482573725e-05,
+      "loss": 0.3401,
+      "step": 2549
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1.648793565683646e-05,
+      "loss": 0.026,
+      "step": 2550
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1.6474530831099195e-05,
+      "loss": 0.1959,
+      "step": 2551
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1.646112600536193e-05,
+      "loss": 0.0022,
+      "step": 2552
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1.6447721179624666e-05,
+      "loss": 0.0012,
+      "step": 2553
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.64343163538874e-05,
+      "loss": 0.0064,
+      "step": 2554
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.6420911528150136e-05,
+      "loss": 0.0105,
+      "step": 2555
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.640750670241287e-05,
+      "loss": 0.0008,
+      "step": 2556
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.6394101876675606e-05,
+      "loss": 0.0339,
+      "step": 2557
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1.6380697050938338e-05,
+      "loss": 0.1458,
+      "step": 2558
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1.6367292225201073e-05,
+      "loss": 0.2526,
+      "step": 2559
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1.6353887399463808e-05,
+      "loss": 0.038,
+      "step": 2560
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.6340482573726543e-05,
+      "loss": 0.174,
+      "step": 2561
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.632707774798928e-05,
+      "loss": 0.1936,
+      "step": 2562
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.631367292225201e-05,
+      "loss": 0.0014,
+      "step": 2563
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.6300268096514745e-05,
+      "loss": 0.0008,
+      "step": 2564
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.628686327077748e-05,
+      "loss": 0.0616,
+      "step": 2565
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.6273458445040215e-05,
+      "loss": 0.0029,
+      "step": 2566
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.626005361930295e-05,
+      "loss": 0.1301,
+      "step": 2567
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.6246648793565686e-05,
+      "loss": 0.0076,
+      "step": 2568
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.6233243967828417e-05,
+      "loss": 0.038,
+      "step": 2569
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.6219839142091153e-05,
+      "loss": 0.0376,
+      "step": 2570
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.6206434316353888e-05,
+      "loss": 0.0007,
+      "step": 2571
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.6193029490616623e-05,
+      "loss": 0.016,
+      "step": 2572
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.6179624664879358e-05,
+      "loss": 0.0005,
+      "step": 2573
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.616621983914209e-05,
+      "loss": 0.0024,
+      "step": 2574
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.6152815013404825e-05,
+      "loss": 0.001,
+      "step": 2575
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.613941018766756e-05,
+      "loss": 0.0009,
+      "step": 2576
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1.6126005361930295e-05,
+      "loss": 0.1889,
+      "step": 2577
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1.611260053619303e-05,
+      "loss": 0.5094,
+      "step": 2578
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1.6099195710455765e-05,
+      "loss": 0.017,
+      "step": 2579
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.60857908847185e-05,
+      "loss": 0.029,
+      "step": 2580
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.6072386058981236e-05,
+      "loss": 0.1249,
+      "step": 2581
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.605898123324397e-05,
+      "loss": 0.2531,
+      "step": 2582
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.6045576407506706e-05,
+      "loss": 0.0006,
+      "step": 2583
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.6032171581769438e-05,
+      "loss": 0.0624,
+      "step": 2584
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.6018766756032173e-05,
+      "loss": 0.0254,
+      "step": 2585
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.6005361930294908e-05,
+      "loss": 0.0034,
+      "step": 2586
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.5991957104557643e-05,
+      "loss": 0.0204,
+      "step": 2587
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.5978552278820378e-05,
+      "loss": 0.002,
+      "step": 2588
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.596514745308311e-05,
+      "loss": 0.0015,
+      "step": 2589
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.5951742627345845e-05,
+      "loss": 0.0465,
+      "step": 2590
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.593833780160858e-05,
+      "loss": 0.1892,
+      "step": 2591
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.5924932975871315e-05,
+      "loss": 0.0932,
+      "step": 2592
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.591152815013405e-05,
+      "loss": 0.0015,
+      "step": 2593
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.5898123324396782e-05,
+      "loss": 0.0062,
+      "step": 2594
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1.5884718498659517e-05,
+      "loss": 0.0731,
+      "step": 2595
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1.5871313672922252e-05,
+      "loss": 0.002,
+      "step": 2596
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1.5857908847184988e-05,
+      "loss": 0.0484,
+      "step": 2597
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1.5844504021447723e-05,
+      "loss": 0.0082,
+      "step": 2598
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.5831099195710454e-05,
+      "loss": 0.0213,
+      "step": 2599
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.581769436997319e-05,
+      "loss": 0.1612,
+      "step": 2600
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.5804289544235925e-05,
+      "loss": 0.184,
+      "step": 2601
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.579088471849866e-05,
+      "loss": 0.1413,
+      "step": 2602
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.5777479892761395e-05,
+      "loss": 0.0019,
+      "step": 2603
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.5764075067024127e-05,
+      "loss": 0.0047,
+      "step": 2604
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.5750670241286862e-05,
+      "loss": 0.0409,
+      "step": 2605
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.5737265415549597e-05,
+      "loss": 0.0379,
+      "step": 2606
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.5723860589812332e-05,
+      "loss": 0.0005,
+      "step": 2607
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.5710455764075067e-05,
+      "loss": 0.0332,
+      "step": 2608
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.5697050938337802e-05,
+      "loss": 0.0543,
+      "step": 2609
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.5683646112600538e-05,
+      "loss": 0.0009,
+      "step": 2610
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.5670241286863273e-05,
+      "loss": 0.016,
+      "step": 2611
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.5656836461126008e-05,
+      "loss": 0.0035,
+      "step": 2612
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.5643431635388743e-05,
+      "loss": 0.0713,
+      "step": 2613
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.5630026809651475e-05,
+      "loss": 0.0022,
+      "step": 2614
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.561662198391421e-05,
+      "loss": 0.0005,
+      "step": 2615
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.5603217158176945e-05,
+      "loss": 0.0009,
+      "step": 2616
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.558981233243968e-05,
+      "loss": 0.0016,
+      "step": 2617
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.5576407506702415e-05,
+      "loss": 0.0017,
+      "step": 2618
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.5563002680965147e-05,
+      "loss": 0.0094,
+      "step": 2619
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.5549597855227882e-05,
+      "loss": 0.016,
+      "step": 2620
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.5536193029490617e-05,
+      "loss": 0.0005,
+      "step": 2621
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.5522788203753352e-05,
+      "loss": 0.0549,
+      "step": 2622
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.5509383378016087e-05,
+      "loss": 0.3791,
+      "step": 2623
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.549597855227882e-05,
+      "loss": 0.0003,
+      "step": 2624
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.5482573726541554e-05,
+      "loss": 0.0774,
+      "step": 2625
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1.546916890080429e-05,
+      "loss": 0.0879,
+      "step": 2626
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1.5455764075067025e-05,
+      "loss": 0.0007,
+      "step": 2627
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1.544235924932976e-05,
+      "loss": 0.0047,
+      "step": 2628
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.542895442359249e-05,
+      "loss": 0.0011,
+      "step": 2629
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.5415549597855227e-05,
+      "loss": 0.0004,
+      "step": 2630
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.5402144772117962e-05,
+      "loss": 0.4962,
+      "step": 2631
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.5388739946380697e-05,
+      "loss": 0.1182,
+      "step": 2632
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.5375335120643432e-05,
+      "loss": 0.0269,
+      "step": 2633
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.5361930294906167e-05,
+      "loss": 0.0157,
+      "step": 2634
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.5348525469168902e-05,
+      "loss": 0.0022,
+      "step": 2635
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.5335120643431637e-05,
+      "loss": 0.3299,
+      "step": 2636
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1.5321715817694372e-05,
+      "loss": 0.0529,
+      "step": 2637
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1.5308310991957108e-05,
+      "loss": 0.1396,
+      "step": 2638
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1.5294906166219843e-05,
+      "loss": 0.0008,
+      "step": 2639
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1.5281501340482574e-05,
+      "loss": 0.0086,
+      "step": 2640
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.526809651474531e-05,
+      "loss": 0.0036,
+      "step": 2641
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.5254691689008043e-05,
+      "loss": 0.0149,
+      "step": 2642
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.5241286863270778e-05,
+      "loss": 0.0011,
+      "step": 2643
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.5227882037533513e-05,
+      "loss": 0.0003,
+      "step": 2644
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.5214477211796247e-05,
+      "loss": 0.0064,
+      "step": 2645
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.5201072386058982e-05,
+      "loss": 0.0281,
+      "step": 2646
+    },
+    {
+      "epoch": 7.0,
+      "eval_f1": 0.7856000000000002,
+      "eval_loss": 1.1071351766586304,
+      "eval_runtime": 1.8613,
+      "eval_samples_per_second": 812.89,
+      "eval_steps_per_second": 51.041,
+      "step": 2646
+    }
+  ],
+  "max_steps": 3780,
+  "num_train_epochs": 10,
+  "total_flos": 678005402777280.0,
+  "trial_name": null,
+  "trial_params": null
+}