diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16231 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2589073634204275,
+  "eval_steps": 211,
+  "global_step": 2650,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 0.8258,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 0.8862780928611755,
+      "eval_runtime": 1.1092,
+      "eval_samples_per_second": 18.03,
+      "eval_steps_per_second": 0.902,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_bench_accuracy_agieval": 0.288135593220339,
+      "eval_bench_accuracy_arc_challenge": 0.8703703703703703,
+      "eval_bench_accuracy_arc_easy": 0.9074074074074074,
+      "eval_bench_accuracy_bigbench": 0.32786885245901637,
+      "eval_bench_accuracy_boolq": 0.6296296296296297,
+      "eval_bench_accuracy_mmlu": 0.5,
+      "eval_bench_accuracy_openbookqa": 0.14814814814814814,
+      "eval_bench_accuracy_truthful_qa": 0.33962264150943394,
+      "eval_bench_accuracy_winogrande": 0.42592592592592593,
+      "eval_bench_average_accuracy": 0.49301206318558566,
+      "eval_bench_loss": 13.334516778049698,
+      "eval_bench_total_accuracy": 0.48893360160965793,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 0.7458,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 0.7332,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8e-07,
+      "loss": 0.739,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 0.7843,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.7191,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6e-07,
+      "loss": 0.7062,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 0.6866,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 0.6765,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.4e-07,
+      "loss": 0.7215,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.7392,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.6e-07,
+      "loss": 0.7096,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2e-07,
+      "loss": 0.6643,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.8e-07,
+      "loss": 0.6998,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 0.7114,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9e-07,
+      "loss": 0.7035,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 0.7194,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0200000000000002e-06,
+      "loss": 0.6801,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.08e-06,
+      "loss": 0.6534,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.14e-06,
+      "loss": 0.7099,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.7001,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.26e-06,
+      "loss": 0.6689,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.32e-06,
+      "loss": 0.7387,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3800000000000001e-06,
+      "loss": 0.592,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.44e-06,
+      "loss": 0.661,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5e-06,
+      "loss": 0.6808,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.56e-06,
+      "loss": 0.6572,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6200000000000002e-06,
+      "loss": 0.6625,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 0.6629,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7399999999999999e-06,
+      "loss": 0.6489,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-06,
+      "loss": 0.6547,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.86e-06,
+      "loss": 0.6465,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 0.6694,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.98e-06,
+      "loss": 0.6412,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0400000000000004e-06,
+      "loss": 0.628,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1e-06,
+      "loss": 0.6677,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.16e-06,
+      "loss": 0.5971,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.22e-06,
+      "loss": 0.6454,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.28e-06,
+      "loss": 0.6904,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.34e-06,
+      "loss": 0.6095,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.656,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4599999999999997e-06,
+      "loss": 0.5945,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.52e-06,
+      "loss": 0.6582,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.58e-06,
+      "loss": 0.5606,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64e-06,
+      "loss": 0.609,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7e-06,
+      "loss": 0.6496,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 0.6451,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.82e-06,
+      "loss": 0.5965,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88e-06,
+      "loss": 0.5845,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9400000000000002e-06,
+      "loss": 0.5782,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3e-06,
+      "loss": 0.6344,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.06e-06,
+      "loss": 0.6494,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.12e-06,
+      "loss": 0.5784,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.18e-06,
+      "loss": 0.6452,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.2400000000000003e-06,
+      "loss": 0.6249,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3e-06,
+      "loss": 0.6783,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 0.5776,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.42e-06,
+      "loss": 0.5939,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.4799999999999997e-06,
+      "loss": 0.6365,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.54e-06,
+      "loss": 0.5617,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.6e-06,
+      "loss": 0.6151,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.66e-06,
+      "loss": 0.6782,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.72e-06,
+      "loss": 0.6026,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.7800000000000002e-06,
+      "loss": 0.54,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 0.6509,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.9e-06,
+      "loss": 0.6074,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.96e-06,
+      "loss": 0.6264,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.0200000000000005e-06,
+      "loss": 0.6212,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.080000000000001e-06,
+      "loss": 0.6692,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.14e-06,
+      "loss": 0.5714,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.2e-06,
+      "loss": 0.6397,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.26e-06,
+      "loss": 0.5525,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.32e-06,
+      "loss": 0.5947,
+      "step": 73
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.38e-06,
+      "loss": 0.607,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.44e-06,
+      "loss": 0.5972,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.5e-06,
+      "loss": 0.5664,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.56e-06,
+      "loss": 0.6446,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.62e-06,
+      "loss": 0.5642,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.68e-06,
+      "loss": 0.6642,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.74e-06,
+      "loss": 0.6257,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.6318,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.86e-06,
+      "loss": 0.5848,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9199999999999995e-06,
+      "loss": 0.6009,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.98e-06,
+      "loss": 0.6102,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.04e-06,
+      "loss": 0.6489,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.1e-06,
+      "loss": 0.6646,
+      "step": 86
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.16e-06,
+      "loss": 0.6051,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.22e-06,
+      "loss": 0.6486,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.28e-06,
+      "loss": 0.5771,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.3400000000000005e-06,
+      "loss": 0.5989,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.4e-06,
+      "loss": 0.6167,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.46e-06,
+      "loss": 0.6193,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 0.6151,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.580000000000001e-06,
+      "loss": 0.5609,
+      "step": 94
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.64e-06,
+      "loss": 0.6174,
+      "step": 95
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.7e-06,
+      "loss": 0.568,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.76e-06,
+      "loss": 0.5576,
+      "step": 97
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.82e-06,
+      "loss": 0.5784,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 0.6191,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.94e-06,
+      "loss": 0.6025,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6e-06,
+      "loss": 0.5861,
+      "step": 101
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998540145985401e-06,
+      "loss": 0.6555,
+      "step": 102
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9970802919708034e-06,
+      "loss": 0.5911,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.995620437956205e-06,
+      "loss": 0.5763,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.994160583941606e-06,
+      "loss": 0.5508,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.992700729927007e-06,
+      "loss": 0.5964,
+      "step": 106
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.991240875912408e-06,
+      "loss": 0.5704,
+      "step": 107
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.98978102189781e-06,
+      "loss": 0.5885,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9883211678832116e-06,
+      "loss": 0.5386,
+      "step": 109
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.986861313868614e-06,
+      "loss": 0.5966,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.985401459854015e-06,
+      "loss": 0.6225,
+      "step": 111
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.983941605839416e-06,
+      "loss": 0.5943,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.982481751824818e-06,
+      "loss": 0.6132,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.981021897810219e-06,
+      "loss": 0.5506,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9795620437956205e-06,
+      "loss": 0.5938,
+      "step": 115
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.978102189781022e-06,
+      "loss": 0.5821,
+      "step": 116
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.976642335766423e-06,
+      "loss": 0.5915,
+      "step": 117
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.975182481751825e-06,
+      "loss": 0.5963,
+      "step": 118
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.973722627737226e-06,
+      "loss": 0.5535,
+      "step": 119
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9722627737226275e-06,
+      "loss": 0.5912,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9708029197080295e-06,
+      "loss": 0.5795,
+      "step": 121
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.969343065693431e-06,
+      "loss": 0.5859,
+      "step": 122
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.967883211678833e-06,
+      "loss": 0.5608,
+      "step": 123
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.966423357664234e-06,
+      "loss": 0.6171,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.964963503649635e-06,
+      "loss": 0.5332,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9635036496350364e-06,
+      "loss": 0.6077,
+      "step": 126
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.962043795620438e-06,
+      "loss": 0.5321,
+      "step": 127
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.96058394160584e-06,
+      "loss": 0.5517,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.959124087591241e-06,
+      "loss": 0.5624,
+      "step": 129
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.957664233576642e-06,
+      "loss": 0.4991,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.956204379562044e-06,
+      "loss": 0.5414,
+      "step": 131
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.954744525547445e-06,
+      "loss": 0.5593,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9532846715328475e-06,
+      "loss": 0.5985,
+      "step": 133
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.951824817518249e-06,
+      "loss": 0.5931,
+      "step": 134
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.95036496350365e-06,
+      "loss": 0.6686,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.948905109489051e-06,
+      "loss": 0.5791,
+      "step": 136
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.947445255474452e-06,
+      "loss": 0.5965,
+      "step": 137
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.945985401459854e-06,
+      "loss": 0.5237,
+      "step": 138
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.944525547445256e-06,
+      "loss": 0.565,
+      "step": 139
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.943065693430657e-06,
+      "loss": 0.5758,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.941605839416059e-06,
+      "loss": 0.5595,
+      "step": 141
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.94014598540146e-06,
+      "loss": 0.6073,
+      "step": 142
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.938686131386861e-06,
+      "loss": 0.6094,
+      "step": 143
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.937226277372263e-06,
+      "loss": 0.5735,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.935766423357664e-06,
+      "loss": 0.5639,
+      "step": 145
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.934306569343066e-06,
+      "loss": 0.5905,
+      "step": 146
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.932846715328467e-06,
+      "loss": 0.607,
+      "step": 147
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.931386861313868e-06,
+      "loss": 0.6186,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.92992700729927e-06,
+      "loss": 0.6587,
+      "step": 149
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9284671532846715e-06,
+      "loss": 0.536,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9270072992700736e-06,
+      "loss": 0.5784,
+      "step": 151
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.925547445255475e-06,
+      "loss": 0.5616,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.924087591240876e-06,
+      "loss": 0.5839,
+      "step": 153
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.922627737226278e-06,
+      "loss": 0.502,
+      "step": 154
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.921167883211678e-06,
+      "loss": 0.6171,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9197080291970805e-06,
+      "loss": 0.6209,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.918248175182482e-06,
+      "loss": 0.5856,
+      "step": 157
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.916788321167883e-06,
+      "loss": 0.5537,
+      "step": 158
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.915328467153285e-06,
+      "loss": 0.5255,
+      "step": 159
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.913868613138686e-06,
+      "loss": 0.5267,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.912408759124088e-06,
+      "loss": 0.5674,
+      "step": 161
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9109489051094895e-06,
+      "loss": 0.6378,
+      "step": 162
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.909489051094891e-06,
+      "loss": 0.5283,
+      "step": 163
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.908029197080293e-06,
+      "loss": 0.5958,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.906569343065693e-06,
+      "loss": 0.5736,
+      "step": 165
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.905109489051095e-06,
+      "loss": 0.5788,
+      "step": 166
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.903649635036496e-06,
+      "loss": 0.6097,
+      "step": 167
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.902189781021898e-06,
+      "loss": 0.5533,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9007299270073e-06,
+      "loss": 0.6474,
+      "step": 169
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.899270072992701e-06,
+      "loss": 0.5248,
+      "step": 170
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.897810218978102e-06,
+      "loss": 0.6149,
+      "step": 171
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.896350364963504e-06,
+      "loss": 0.6259,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.894890510948905e-06,
+      "loss": 0.5785,
+      "step": 173
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.893430656934307e-06,
+      "loss": 0.5466,
+      "step": 174
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.891970802919708e-06,
+      "loss": 0.6307,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.89051094890511e-06,
+      "loss": 0.5287,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.889051094890511e-06,
+      "loss": 0.5985,
+      "step": 177
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.887591240875912e-06,
+      "loss": 0.5386,
+      "step": 178
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.886131386861314e-06,
+      "loss": 0.5721,
+      "step": 179
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.8846715328467156e-06,
+      "loss": 0.5399,
+      "step": 180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.883211678832117e-06,
+      "loss": 0.5507,
+      "step": 181
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.881751824817519e-06,
+      "loss": 0.5543,
+      "step": 182
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.88029197080292e-06,
+      "loss": 0.556,
+      "step": 183
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.878832116788321e-06,
+      "loss": 0.5776,
+      "step": 184
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.8773722627737225e-06,
+      "loss": 0.5459,
+      "step": 185
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.875912408759124e-06,
+      "loss": 0.5779,
+      "step": 186
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.874452554744526e-06,
+      "loss": 0.5828,
+      "step": 187
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.872992700729927e-06,
+      "loss": 0.5353,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.871532846715329e-06,
+      "loss": 0.5767,
+      "step": 189
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.87007299270073e-06,
+      "loss": 0.5528,
+      "step": 190
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.8686131386861314e-06,
+      "loss": 0.5304,
+      "step": 191
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.8671532846715335e-06,
+      "loss": 0.5719,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.865693430656935e-06,
+      "loss": 0.5575,
+      "step": 193
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.864233576642336e-06,
+      "loss": 0.5698,
+      "step": 194
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.862773722627737e-06,
+      "loss": 0.4945,
+      "step": 195
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.861313868613138e-06,
+      "loss": 0.5665,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.8598540145985404e-06,
+      "loss": 0.5602,
+      "step": 197
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.858394160583942e-06,
+      "loss": 0.559,
+      "step": 198
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.856934306569343e-06,
+      "loss": 0.5634,
+      "step": 199
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.855474452554745e-06,
+      "loss": 0.5464,
+      "step": 200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.854014598540146e-06,
+      "loss": 0.516,
+      "step": 201
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.852554744525548e-06,
+      "loss": 0.5474,
+      "step": 202
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.851094890510949e-06,
+      "loss": 0.5225,
+      "step": 203
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.849635036496351e-06,
+      "loss": 0.5758,
+      "step": 204
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.848175182481752e-06,
+      "loss": 0.6115,
+      "step": 205
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.846715328467153e-06,
+      "loss": 0.59,
+      "step": 206
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.845255474452555e-06,
+      "loss": 0.5801,
+      "step": 207
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.843795620437956e-06,
+      "loss": 0.5606,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.8423357664233575e-06,
+      "loss": 0.578,
+      "step": 209
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.84087591240876e-06,
+      "loss": 0.5801,
+      "step": 210
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.839416058394161e-06,
+      "loss": 0.5806,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.598720371723175,
+      "eval_runtime": 1.1188,
+      "eval_samples_per_second": 17.877,
+      "eval_steps_per_second": 0.894,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "eval_bench_accuracy_agieval": 0.3220338983050847,
+      "eval_bench_accuracy_arc_challenge": 0.8518518518518519,
+      "eval_bench_accuracy_arc_easy": 0.8703703703703703,
+      "eval_bench_accuracy_bigbench": 0.4426229508196721,
+      "eval_bench_accuracy_boolq": 0.6851851851851852,
+      "eval_bench_accuracy_mmlu": 0.48148148148148145,
+      "eval_bench_accuracy_openbookqa": 0.18518518518518517,
+      "eval_bench_accuracy_truthful_qa": 0.2830188679245283,
+      "eval_bench_accuracy_winogrande": 0.3333333333333333,
+      "eval_bench_average_accuracy": 0.4950092360507436,
+      "eval_bench_loss": 15.735029473362198,
+      "eval_bench_total_accuracy": 0.49295774647887325,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.837956204379563e-06,
+      "loss": 0.5794,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.836496350364964e-06,
+      "loss": 0.5781,
+      "step": 213
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.8350364963503645e-06,
+      "loss": 0.6346,
+      "step": 214
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.8335766423357665e-06,
+      "loss": 0.5709,
+      "step": 215
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.832116788321168e-06,
+      "loss": 0.553,
+      "step": 216
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.83065693430657e-06,
+      "loss": 0.5894,
+      "step": 217
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.829197080291971e-06,
+      "loss": 0.6053,
+      "step": 218
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.827737226277372e-06,
+      "loss": 0.5744,
+      "step": 219
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.826277372262774e-06,
+      "loss": 0.5666,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.8248175182481755e-06,
+      "loss": 0.5227,
+      "step": 221
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.823357664233577e-06,
+      "loss": 0.5877,
+      "step": 222
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.821897810218978e-06,
+      "loss": 0.5836,
+      "step": 223
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.820437956204379e-06,
+      "loss": 0.5238,
+      "step": 224
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.818978102189781e-06,
+      "loss": 0.5815,
+      "step": 225
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.817518248175182e-06,
+      "loss": 0.5495,
+      "step": 226
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.816058394160584e-06,
+      "loss": 0.559,
+      "step": 227
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.814598540145986e-06,
+      "loss": 0.5496,
+      "step": 228
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.813138686131387e-06,
+      "loss": 0.6007,
+      "step": 229
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.811678832116789e-06,
+      "loss": 0.563,
+      "step": 230
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.81021897810219e-06,
+      "loss": 0.5872,
+      "step": 231
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.808759124087591e-06,
+      "loss": 0.5563,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.807299270072993e-06,
+      "loss": 0.5412,
+      "step": 233
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.805839416058394e-06,
+      "loss": 0.6444,
+      "step": 234
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.804379562043796e-06,
+      "loss": 0.5131,
+      "step": 235
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.802919708029197e-06,
+      "loss": 0.5632,
+      "step": 236
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.801459854014598e-06,
+      "loss": 0.594,
+      "step": 237
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.8e-06,
+      "loss": 0.5742,
+      "step": 238
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.798540145985402e-06,
+      "loss": 0.5738,
+      "step": 239
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.797080291970804e-06,
+      "loss": 0.6661,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.795620437956205e-06,
+      "loss": 0.5186,
+      "step": 241
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.794160583941606e-06,
+      "loss": 0.5425,
+      "step": 242
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.792700729927007e-06,
+      "loss": 0.6064,
+      "step": 243
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.7912408759124085e-06,
+      "loss": 0.5435,
+      "step": 244
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.7897810218978106e-06,
+      "loss": 0.4822,
+      "step": 245
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.788321167883212e-06,
+      "loss": 0.522,
+      "step": 246
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.786861313868613e-06,
+      "loss": 0.5341,
+      "step": 247
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.785401459854015e-06,
+      "loss": 0.6082,
+      "step": 248
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.783941605839416e-06,
+      "loss": 0.5747,
+      "step": 249
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.7824817518248175e-06,
+      "loss": 0.5825,
+      "step": 250
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.7810218978102195e-06,
+      "loss": 0.5917,
+      "step": 251
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.77956204379562e-06,
+      "loss": 0.6264,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.778102189781022e-06,
+      "loss": 0.5727,
+      "step": 253
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.776642335766423e-06,
+      "loss": 0.5368,
+      "step": 254
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.775182481751825e-06,
+      "loss": 0.5263,
+      "step": 255
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.7737226277372265e-06,
+      "loss": 0.6055,
+      "step": 256
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.772262773722628e-06,
+      "loss": 0.5725,
+      "step": 257
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.77080291970803e-06,
+      "loss": 0.5759,
+      "step": 258
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.769343065693431e-06,
+      "loss": 0.5609,
+      "step": 259
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.767883211678832e-06,
+      "loss": 0.5618,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.766423357664234e-06,
+      "loss": 0.5292,
+      "step": 261
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.764963503649635e-06,
+      "loss": 0.5699,
+      "step": 262
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.763503649635037e-06,
+      "loss": 0.5616,
+      "step": 263
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.762043795620438e-06,
+      "loss": 0.5206,
+      "step": 264
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.760583941605839e-06,
+      "loss": 0.5541,
+      "step": 265
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.759124087591241e-06,
+      "loss": 0.5601,
+      "step": 266
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.757664233576642e-06,
+      "loss": 0.5415,
+      "step": 267
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.756204379562044e-06,
+      "loss": 0.5594,
+      "step": 268
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.754744525547446e-06,
+      "loss": 0.5828,
+      "step": 269
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.753284671532847e-06,
+      "loss": 0.5527,
+      "step": 270
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.751824817518249e-06,
+      "loss": 0.5294,
+      "step": 271
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.750364963503649e-06,
+      "loss": 0.5383,
+      "step": 272
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.748905109489051e-06,
+      "loss": 0.5154,
+      "step": 273
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.7474452554744526e-06,
+      "loss": 0.6035,
+      "step": 274
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.745985401459854e-06,
+      "loss": 0.5459,
+      "step": 275
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.744525547445256e-06,
+      "loss": 0.557,
+      "step": 276
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.743065693430657e-06,
+      "loss": 0.6156,
+      "step": 277
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.741605839416058e-06,
+      "loss": 0.5384,
+      "step": 278
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.74014598540146e-06,
+      "loss": 0.5921,
+      "step": 279
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.7386861313868615e-06,
+      "loss": 0.5419,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.737226277372264e-06,
+      "loss": 0.5612,
+      "step": 281
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.735766423357664e-06,
+      "loss": 0.5558,
+      "step": 282
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.734306569343066e-06,
+      "loss": 0.5942,
+      "step": 283
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.732846715328467e-06,
+      "loss": 0.591,
+      "step": 284
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.7313868613138684e-06,
+      "loss": 0.5527,
+      "step": 285
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.7299270072992705e-06,
+      "loss": 0.5575,
+      "step": 286
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.728467153284672e-06,
+      "loss": 0.5371,
+      "step": 287
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.727007299270073e-06,
+      "loss": 0.527,
+      "step": 288
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.725547445255475e-06,
+      "loss": 0.5504,
+      "step": 289
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.724087591240876e-06,
+      "loss": 0.5871,
+      "step": 290
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.7226277372262774e-06,
+      "loss": 0.5569,
+      "step": 291
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.721167883211679e-06,
+      "loss": 0.6058,
+      "step": 292
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.71970802919708e-06,
+      "loss": 0.5512,
+      "step": 293
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.718248175182482e-06,
+      "loss": 0.5327,
+      "step": 294
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.716788321167883e-06,
+      "loss": 0.5528,
+      "step": 295
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.715328467153285e-06,
+      "loss": 0.6255,
+      "step": 296
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.713868613138686e-06,
+      "loss": 0.5423,
+      "step": 297
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.712408759124088e-06,
+      "loss": 0.5462,
+      "step": 298
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.71094890510949e-06,
+      "loss": 0.5092,
+      "step": 299
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.709489051094891e-06,
+      "loss": 0.5575,
+      "step": 300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.708029197080292e-06,
+      "loss": 0.4971,
+      "step": 301
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.706569343065693e-06,
+      "loss": 0.542,
+      "step": 302
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.7051094890510945e-06,
+      "loss": 0.5605,
+      "step": 303
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.703649635036497e-06,
+      "loss": 0.5493,
+      "step": 304
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.702189781021898e-06,
+      "loss": 0.5171,
+      "step": 305
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.700729927007299e-06,
+      "loss": 0.515,
+      "step": 306
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.699270072992701e-06,
+      "loss": 0.5662,
+      "step": 307
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.697810218978102e-06,
+      "loss": 0.5252,
+      "step": 308
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.696350364963504e-06,
+      "loss": 0.6074,
+      "step": 309
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.694890510948906e-06,
+      "loss": 0.5872,
+      "step": 310
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.693430656934307e-06,
+      "loss": 0.5219,
+      "step": 311
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.691970802919708e-06,
+      "loss": 0.5696,
+      "step": 312
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.690510948905109e-06,
+      "loss": 0.5705,
+      "step": 313
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.689051094890511e-06,
+      "loss": 0.584,
+      "step": 314
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.6875912408759125e-06,
+      "loss": 0.5392,
+      "step": 315
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.686131386861314e-06,
+      "loss": 0.5442,
+      "step": 316
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.684671532846716e-06,
+      "loss": 0.5517,
+      "step": 317
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.683211678832117e-06,
+      "loss": 0.556,
+      "step": 318
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.681751824817519e-06,
+      "loss": 0.5683,
+      "step": 319
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.68029197080292e-06,
+      "loss": 0.5172,
+      "step": 320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.678832116788321e-06,
+      "loss": 0.507,
+      "step": 321
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.677372262773723e-06,
+      "loss": 0.5764,
+      "step": 322
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.675912408759124e-06,
+      "loss": 0.5582,
+      "step": 323
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.674452554744526e-06,
+      "loss": 0.5145,
+      "step": 324
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.672992700729927e-06,
+      "loss": 0.5441,
+      "step": 325
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.671532846715328e-06,
+      "loss": 0.6081,
+      "step": 326
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.6700729927007305e-06,
+      "loss": 0.6022,
+      "step": 327
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.668613138686132e-06,
+      "loss": 0.5461,
+      "step": 328
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.667153284671533e-06,
+      "loss": 0.557,
+      "step": 329
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.665693430656934e-06,
+      "loss": 0.6166,
+      "step": 330
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.664233576642335e-06,
+      "loss": 0.487,
+      "step": 331
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.662773722627737e-06,
+      "loss": 0.559,
+      "step": 332
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.661313868613139e-06,
+      "loss": 0.4612,
+      "step": 333
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.659854014598541e-06,
+      "loss": 0.548,
+      "step": 334
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.658394160583942e-06,
+      "loss": 0.5731,
+      "step": 335
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.656934306569343e-06,
+      "loss": 0.5344,
+      "step": 336
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.655474452554745e-06,
+      "loss": 0.531,
+      "step": 337
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.654014598540146e-06,
+      "loss": 0.5314,
+      "step": 338
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.6525547445255476e-06,
+      "loss": 0.5447,
+      "step": 339
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.651094890510949e-06,
+      "loss": 0.4869,
+      "step": 340
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.64963503649635e-06,
+      "loss": 0.5389,
+      "step": 341
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.648175182481752e-06,
+      "loss": 0.5811,
+      "step": 342
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.646715328467153e-06,
+      "loss": 0.6479,
+      "step": 343
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.6452554744525545e-06,
+      "loss": 0.5307,
+      "step": 344
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.6437956204379565e-06,
+      "loss": 0.5683,
+      "step": 345
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.642335766423358e-06,
+      "loss": 0.5079,
+      "step": 346
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.64087591240876e-06,
+      "loss": 0.5026,
+      "step": 347
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.639416058394161e-06,
+      "loss": 0.5245,
+      "step": 348
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.637956204379562e-06,
+      "loss": 0.5369,
+      "step": 349
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.6364963503649635e-06,
+      "loss": 0.5423,
+      "step": 350
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.635036496350365e-06,
+      "loss": 0.5229,
+      "step": 351
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.633576642335767e-06,
+      "loss": 0.533,
+      "step": 352
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.632116788321168e-06,
+      "loss": 0.5572,
+      "step": 353
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.630656934306569e-06,
+      "loss": 0.5294,
+      "step": 354
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.629197080291971e-06,
+      "loss": 0.4568,
+      "step": 355
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.6277372262773724e-06,
+      "loss": 0.5216,
+      "step": 356
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.626277372262774e-06,
+      "loss": 0.5962,
+      "step": 357
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.624817518248176e-06,
+      "loss": 0.5743,
+      "step": 358
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.623357664233577e-06,
+      "loss": 0.5988,
+      "step": 359
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.621897810218978e-06,
+      "loss": 0.5797,
+      "step": 360
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.620437956204379e-06,
+      "loss": 0.5653,
+      "step": 361
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.618978102189781e-06,
+      "loss": 0.5281,
+      "step": 362
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.617518248175183e-06,
+      "loss": 0.5336,
+      "step": 363
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.616058394160584e-06,
+      "loss": 0.5379,
+      "step": 364
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.614598540145986e-06,
+      "loss": 0.521,
+      "step": 365
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.613138686131387e-06,
+      "loss": 0.5375,
+      "step": 366
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.611678832116788e-06,
+      "loss": 0.5544,
+      "step": 367
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.61021897810219e-06,
+      "loss": 0.5063,
+      "step": 368
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.608759124087591e-06,
+      "loss": 0.5093,
+      "step": 369
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.607299270072993e-06,
+      "loss": 0.5251,
+      "step": 370
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.605839416058394e-06,
+      "loss": 0.5538,
+      "step": 371
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.604379562043795e-06,
+      "loss": 0.5439,
+      "step": 372
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.602919708029197e-06,
+      "loss": 0.6022,
+      "step": 373
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.6014598540145985e-06,
+      "loss": 0.5279,
+      "step": 374
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.6297,
+      "step": 375
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.598540145985402e-06,
+      "loss": 0.5275,
+      "step": 376
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.597080291970803e-06,
+      "loss": 0.5221,
+      "step": 377
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.595620437956205e-06,
+      "loss": 0.5394,
+      "step": 378
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.5941605839416055e-06,
+      "loss": 0.5421,
+      "step": 379
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.5927007299270075e-06,
+      "loss": 0.5483,
+      "step": 380
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.591240875912409e-06,
+      "loss": 0.5385,
+      "step": 381
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.58978102189781e-06,
+      "loss": 0.6108,
+      "step": 382
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.588321167883212e-06,
+      "loss": 0.5814,
+      "step": 383
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.586861313868613e-06,
+      "loss": 0.5728,
+      "step": 384
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.5854014598540144e-06,
+      "loss": 0.5615,
+      "step": 385
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.5839416058394165e-06,
+      "loss": 0.5506,
+      "step": 386
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.582481751824818e-06,
+      "loss": 0.5904,
+      "step": 387
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.58102189781022e-06,
+      "loss": 0.5269,
+      "step": 388
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.57956204379562e-06,
+      "loss": 0.5447,
+      "step": 389
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.578102189781022e-06,
+      "loss": 0.5906,
+      "step": 390
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.576642335766423e-06,
+      "loss": 0.5295,
+      "step": 391
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.575182481751825e-06,
+      "loss": 0.5536,
+      "step": 392
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.573722627737227e-06,
+      "loss": 0.5818,
+      "step": 393
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.572262773722628e-06,
+      "loss": 0.5687,
+      "step": 394
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.570802919708029e-06,
+      "loss": 0.5062,
+      "step": 395
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.569343065693431e-06,
+      "loss": 0.5845,
+      "step": 396
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.567883211678832e-06,
+      "loss": 0.5187,
+      "step": 397
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.5664233576642344e-06,
+      "loss": 0.4979,
+      "step": 398
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.564963503649635e-06,
+      "loss": 0.5578,
+      "step": 399
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.563503649635036e-06,
+      "loss": 0.5948,
+      "step": 400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.562043795620438e-06,
+      "loss": 0.5468,
+      "step": 401
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.560583941605839e-06,
+      "loss": 0.5197,
+      "step": 402
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.559124087591241e-06,
+      "loss": 0.597,
+      "step": 403
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.557664233576643e-06,
+      "loss": 0.5662,
+      "step": 404
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.556204379562044e-06,
+      "loss": 0.5095,
+      "step": 405
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.554744525547446e-06,
+      "loss": 0.5091,
+      "step": 406
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.553284671532847e-06,
+      "loss": 0.5706,
+      "step": 407
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.551824817518248e-06,
+      "loss": 0.5146,
+      "step": 408
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.5503649635036495e-06,
+      "loss": 0.528,
+      "step": 409
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.548905109489051e-06,
+      "loss": 0.5584,
+      "step": 410
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.547445255474453e-06,
+      "loss": 0.588,
+      "step": 411
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.545985401459854e-06,
+      "loss": 0.5383,
+      "step": 412
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.544525547445256e-06,
+      "loss": 0.5404,
+      "step": 413
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.543065693430657e-06,
+      "loss": 0.5551,
+      "step": 414
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.5416058394160585e-06,
+      "loss": 0.5583,
+      "step": 415
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.5401459854014605e-06,
+      "loss": 0.5634,
+      "step": 416
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.538686131386862e-06,
+      "loss": 0.547,
+      "step": 417
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.537226277372263e-06,
+      "loss": 0.5686,
+      "step": 418
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.535766423357664e-06,
+      "loss": 0.5627,
+      "step": 419
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.534306569343065e-06,
+      "loss": 0.5116,
+      "step": 420
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.5328467153284675e-06,
+      "loss": 0.5556,
+      "step": 421
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.531386861313869e-06,
+      "loss": 0.5437,
+      "step": 422
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5592204928398132,
+      "eval_runtime": 1.1202,
+      "eval_samples_per_second": 17.854,
+      "eval_steps_per_second": 0.893,
+      "step": 422
+    },
+    {
+      "epoch": 0.2,
+      "eval_bench_accuracy_agieval": 0.3220338983050847,
+      "eval_bench_accuracy_arc_challenge": 0.8333333333333334,
+      "eval_bench_accuracy_arc_easy": 0.9074074074074074,
+      "eval_bench_accuracy_bigbench": 0.4426229508196721,
+      "eval_bench_accuracy_boolq": 0.7962962962962963,
+      "eval_bench_accuracy_mmlu": 0.4444444444444444,
+      "eval_bench_accuracy_openbookqa": 0.18518518518518517,
+      "eval_bench_accuracy_truthful_qa": 0.37735849056603776,
+      "eval_bench_accuracy_winogrande": 0.48148148148148145,
+      "eval_bench_average_accuracy": 0.5322403875376602,
+      "eval_bench_loss": 17.45224844691265,
+      "eval_bench_total_accuracy": 0.5291750503018109,
+      "step": 422
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.52992700729927e-06,
+      "loss": 0.5453,
+      "step": 423
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.528467153284672e-06,
+      "loss": 0.583,
+      "step": 424
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.527007299270073e-06,
+      "loss": 0.5119,
+      "step": 425
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.525547445255475e-06,
+      "loss": 0.5591,
+      "step": 426
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.5240875912408764e-06,
+      "loss": 0.5449,
+      "step": 427
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.522627737226278e-06,
+      "loss": 0.5282,
+      "step": 428
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.521167883211679e-06,
+      "loss": 0.5099,
+      "step": 429
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.51970802919708e-06,
+      "loss": 0.6226,
+      "step": 430
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.518248175182482e-06,
+      "loss": 0.5346,
+      "step": 431
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.516788321167883e-06,
+      "loss": 0.5394,
+      "step": 432
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.5153284671532846e-06,
+      "loss": 0.561,
+      "step": 433
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.513868613138687e-06,
+      "loss": 0.5881,
+      "step": 434
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.512408759124088e-06,
+      "loss": 0.5966,
+      "step": 435
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.510948905109489e-06,
+      "loss": 0.5551,
+      "step": 436
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.509489051094891e-06,
+      "loss": 0.5207,
+      "step": 437
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.5080291970802915e-06,
+      "loss": 0.5447,
+      "step": 438
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.5065693430656935e-06,
+      "loss": 0.584,
+      "step": 439
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.505109489051095e-06,
+      "loss": 0.487,
+      "step": 440
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.503649635036497e-06,
+      "loss": 0.5274,
+      "step": 441
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.502189781021898e-06,
+      "loss": 0.5079,
+      "step": 442
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.500729927007299e-06,
+      "loss": 0.5804,
+      "step": 443
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.499270072992701e-06,
+      "loss": 0.4931,
+      "step": 444
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.4978102189781025e-06,
+      "loss": 0.4635,
+      "step": 445
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.496350364963504e-06,
+      "loss": 0.496,
+      "step": 446
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.494890510948905e-06,
+      "loss": 0.5402,
+      "step": 447
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.493430656934306e-06,
+      "loss": 0.5074,
+      "step": 448
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.491970802919708e-06,
+      "loss": 0.5912,
+      "step": 449
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.4905109489051094e-06,
+      "loss": 0.6246,
+      "step": 450
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.489051094890511e-06,
+      "loss": 0.5922,
+      "step": 451
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.487591240875913e-06,
+      "loss": 0.6096,
+      "step": 452
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.486131386861314e-06,
+      "loss": 0.5493,
+      "step": 453
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.484671532846716e-06,
+      "loss": 0.5052,
+      "step": 454
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.483211678832117e-06,
+      "loss": 0.5312,
+      "step": 455
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.481751824817518e-06,
+      "loss": 0.5326,
+      "step": 456
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.48029197080292e-06,
+      "loss": 0.576,
+      "step": 457
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.478832116788321e-06,
+      "loss": 0.5712,
+      "step": 458
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.477372262773723e-06,
+      "loss": 0.5456,
+      "step": 459
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.475912408759124e-06,
+      "loss": 0.5573,
+      "step": 460
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.474452554744525e-06,
+      "loss": 0.564,
+      "step": 461
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.472992700729927e-06,
+      "loss": 0.5235,
+      "step": 462
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.471532846715329e-06,
+      "loss": 0.5069,
+      "step": 463
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.47007299270073e-06,
+      "loss": 0.521,
+      "step": 464
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.468613138686132e-06,
+      "loss": 0.5428,
+      "step": 465
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.467153284671533e-06,
+      "loss": 0.4962,
+      "step": 466
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.465693430656934e-06,
+      "loss": 0.5353,
+      "step": 467
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.4642335766423355e-06,
+      "loss": 0.5263,
+      "step": 468
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.462773722627738e-06,
+      "loss": 0.4951,
+      "step": 469
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.461313868613139e-06,
+      "loss": 0.5299,
+      "step": 470
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.45985401459854e-06,
+      "loss": 0.5015,
+      "step": 471
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.458394160583942e-06,
+      "loss": 0.5172,
+      "step": 472
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.456934306569343e-06,
+      "loss": 0.5598,
+      "step": 473
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.4554744525547445e-06,
+      "loss": 0.5284,
+      "step": 474
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.4540145985401466e-06,
+      "loss": 0.5945,
+      "step": 475
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.452554744525547e-06,
+      "loss": 0.5625,
+      "step": 476
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.451094890510949e-06,
+      "loss": 0.5739,
+      "step": 477
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.44963503649635e-06,
+      "loss": 0.4998,
+      "step": 478
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.4481751824817514e-06,
+      "loss": 0.5329,
+      "step": 479
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.4467153284671535e-06,
+      "loss": 0.5392,
+      "step": 480
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.445255474452555e-06,
+      "loss": 0.5167,
+      "step": 481
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.443795620437957e-06,
+      "loss": 0.4605,
+      "step": 482
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.442335766423358e-06,
+      "loss": 0.5302,
+      "step": 483
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.440875912408759e-06,
+      "loss": 0.5006,
+      "step": 484
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.439416058394161e-06,
+      "loss": 0.5007,
+      "step": 485
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.437956204379562e-06,
+      "loss": 0.5122,
+      "step": 486
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.436496350364964e-06,
+      "loss": 0.5064,
+      "step": 487
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.435036496350365e-06,
+      "loss": 0.5203,
+      "step": 488
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.433576642335766e-06,
+      "loss": 0.5429,
+      "step": 489
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.432116788321168e-06,
+      "loss": 0.461,
+      "step": 490
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.430656934306569e-06,
+      "loss": 0.5733,
+      "step": 491
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.4291970802919714e-06,
+      "loss": 0.5417,
+      "step": 492
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.427737226277373e-06,
+      "loss": 0.5268,
+      "step": 493
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.426277372262774e-06,
+      "loss": 0.5596,
+      "step": 494
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.424817518248176e-06,
+      "loss": 0.5505,
+      "step": 495
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.423357664233576e-06,
+      "loss": 0.562,
+      "step": 496
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.421897810218978e-06,
+      "loss": 0.4972,
+      "step": 497
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.42043795620438e-06,
+      "loss": 0.5383,
+      "step": 498
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.418978102189781e-06,
+      "loss": 0.5506,
+      "step": 499
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.417518248175183e-06,
+      "loss": 0.5212,
+      "step": 500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.416058394160584e-06,
+      "loss": 0.5145,
+      "step": 501
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.414598540145985e-06,
+      "loss": 0.5773,
+      "step": 502
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.413138686131387e-06,
+      "loss": 0.4584,
+      "step": 503
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.4116788321167886e-06,
+      "loss": 0.5085,
+      "step": 504
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.410218978102191e-06,
+      "loss": 0.5736,
+      "step": 505
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.408759124087591e-06,
+      "loss": 0.5683,
+      "step": 506
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.407299270072993e-06,
+      "loss": 0.4981,
+      "step": 507
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.405839416058394e-06,
+      "loss": 0.5202,
+      "step": 508
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.4043795620437955e-06,
+      "loss": 0.5181,
+      "step": 509
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.4029197080291975e-06,
+      "loss": 0.5083,
+      "step": 510
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.401459854014599e-06,
+      "loss": 0.5817,
+      "step": 511
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.4e-06,
+      "loss": 0.5129,
+      "step": 512
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.398540145985402e-06,
+      "loss": 0.595,
+      "step": 513
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.397080291970803e-06,
+      "loss": 0.5003,
+      "step": 514
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.3956204379562045e-06,
+      "loss": 0.5806,
+      "step": 515
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.394160583941606e-06,
+      "loss": 0.5659,
+      "step": 516
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.392700729927007e-06,
+      "loss": 0.5467,
+      "step": 517
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.391240875912409e-06,
+      "loss": 0.5024,
+      "step": 518
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.38978102189781e-06,
+      "loss": 0.4808,
+      "step": 519
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.388321167883212e-06,
+      "loss": 0.4965,
+      "step": 520
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.3868613138686134e-06,
+      "loss": 0.521,
+      "step": 521
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.385401459854015e-06,
+      "loss": 0.5814,
+      "step": 522
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.383941605839417e-06,
+      "loss": 0.5489,
+      "step": 523
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.382481751824818e-06,
+      "loss": 0.6684,
+      "step": 524
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.381021897810219e-06,
+      "loss": 0.5117,
+      "step": 525
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.37956204379562e-06,
+      "loss": 0.5198,
+      "step": 526
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.3781021897810216e-06,
+      "loss": 0.4783,
+      "step": 527
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.376642335766424e-06,
+      "loss": 0.539,
+      "step": 528
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.375182481751825e-06,
+      "loss": 0.564,
+      "step": 529
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.373722627737226e-06,
+      "loss": 0.5688,
+      "step": 530
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.372262773722628e-06,
+      "loss": 0.5058,
+      "step": 531
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.370802919708029e-06,
+      "loss": 0.5131,
+      "step": 532
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.369343065693431e-06,
+      "loss": 0.5356,
+      "step": 533
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.367883211678833e-06,
+      "loss": 0.5269,
+      "step": 534
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.366423357664234e-06,
+      "loss": 0.4997,
+      "step": 535
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.364963503649635e-06,
+      "loss": 0.5536,
+      "step": 536
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.363503649635036e-06,
+      "loss": 0.5496,
+      "step": 537
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.362043795620438e-06,
+      "loss": 0.5436,
+      "step": 538
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.3605839416058395e-06,
+      "loss": 0.5129,
+      "step": 539
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.359124087591241e-06,
+      "loss": 0.5709,
+      "step": 540
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.357664233576643e-06,
+      "loss": 0.4997,
+      "step": 541
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.356204379562044e-06,
+      "loss": 0.5007,
+      "step": 542
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.354744525547445e-06,
+      "loss": 0.5622,
+      "step": 543
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.353284671532847e-06,
+      "loss": 0.5099,
+      "step": 544
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.351824817518248e-06,
+      "loss": 0.5002,
+      "step": 545
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.35036496350365e-06,
+      "loss": 0.5068,
+      "step": 546
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.348905109489051e-06,
+      "loss": 0.5613,
+      "step": 547
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.347445255474453e-06,
+      "loss": 0.5398,
+      "step": 548
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.345985401459854e-06,
+      "loss": 0.5407,
+      "step": 549
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.344525547445255e-06,
+      "loss": 0.5745,
+      "step": 550
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.3430656934306575e-06,
+      "loss": 0.5051,
+      "step": 551
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.341605839416059e-06,
+      "loss": 0.5046,
+      "step": 552
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.34014598540146e-06,
+      "loss": 0.5212,
+      "step": 553
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.338686131386861e-06,
+      "loss": 0.5489,
+      "step": 554
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.337226277372262e-06,
+      "loss": 0.5565,
+      "step": 555
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.335766423357664e-06,
+      "loss": 0.5694,
+      "step": 556
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.334306569343066e-06,
+      "loss": 0.4989,
+      "step": 557
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.332846715328467e-06,
+      "loss": 0.5229,
+      "step": 558
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.331386861313869e-06,
+      "loss": 0.5231,
+      "step": 559
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.32992700729927e-06,
+      "loss": 0.5367,
+      "step": 560
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.328467153284672e-06,
+      "loss": 0.5275,
+      "step": 561
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.327007299270073e-06,
+      "loss": 0.5977,
+      "step": 562
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.325547445255475e-06,
+      "loss": 0.5085,
+      "step": 563
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.324087591240876e-06,
+      "loss": 0.5629,
+      "step": 564
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.322627737226277e-06,
+      "loss": 0.591,
+      "step": 565
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.321167883211679e-06,
+      "loss": 0.4828,
+      "step": 566
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.31970802919708e-06,
+      "loss": 0.4782,
+      "step": 567
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.3182481751824815e-06,
+      "loss": 0.5487,
+      "step": 568
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.3167883211678836e-06,
+      "loss": 0.5131,
+      "step": 569
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.315328467153285e-06,
+      "loss": 0.5396,
+      "step": 570
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.313868613138687e-06,
+      "loss": 0.5667,
+      "step": 571
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.312408759124088e-06,
+      "loss": 0.5793,
+      "step": 572
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.310948905109489e-06,
+      "loss": 0.5015,
+      "step": 573
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.3094890510948905e-06,
+      "loss": 0.5642,
+      "step": 574
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.308029197080292e-06,
+      "loss": 0.4769,
+      "step": 575
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.306569343065694e-06,
+      "loss": 0.5573,
+      "step": 576
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.305109489051095e-06,
+      "loss": 0.4985,
+      "step": 577
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.303649635036496e-06,
+      "loss": 0.5244,
+      "step": 578
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.302189781021898e-06,
+      "loss": 0.5283,
+      "step": 579
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.3007299270072995e-06,
+      "loss": 0.5459,
+      "step": 580
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.299270072992701e-06,
+      "loss": 0.5236,
+      "step": 581
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.297810218978103e-06,
+      "loss": 0.53,
+      "step": 582
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.296350364963504e-06,
+      "loss": 0.522,
+      "step": 583
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.294890510948905e-06,
+      "loss": 0.5493,
+      "step": 584
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.293430656934306e-06,
+      "loss": 0.5332,
+      "step": 585
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.2919708029197084e-06,
+      "loss": 0.6113,
+      "step": 586
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.29051094890511e-06,
+      "loss": 0.4756,
+      "step": 587
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.289051094890511e-06,
+      "loss": 0.5044,
+      "step": 588
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.287591240875913e-06,
+      "loss": 0.5609,
+      "step": 589
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.286131386861314e-06,
+      "loss": 0.5185,
+      "step": 590
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.284671532846715e-06,
+      "loss": 0.5406,
+      "step": 591
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.2832116788321174e-06,
+      "loss": 0.5789,
+      "step": 592
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.281751824817518e-06,
+      "loss": 0.5513,
+      "step": 593
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.28029197080292e-06,
+      "loss": 0.4989,
+      "step": 594
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.278832116788321e-06,
+      "loss": 0.5183,
+      "step": 595
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.277372262773722e-06,
+      "loss": 0.5863,
+      "step": 596
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.275912408759124e-06,
+      "loss": 0.5628,
+      "step": 597
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.2744525547445256e-06,
+      "loss": 0.527,
+      "step": 598
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.272992700729928e-06,
+      "loss": 0.548,
+      "step": 599
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.271532846715329e-06,
+      "loss": 0.5531,
+      "step": 600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.27007299270073e-06,
+      "loss": 0.4653,
+      "step": 601
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.268613138686132e-06,
+      "loss": 0.4928,
+      "step": 602
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.2671532846715325e-06,
+      "loss": 0.6501,
+      "step": 603
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.2656934306569345e-06,
+      "loss": 0.5115,
+      "step": 604
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.264233576642336e-06,
+      "loss": 0.4757,
+      "step": 605
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.262773722627737e-06,
+      "loss": 0.5231,
+      "step": 606
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.261313868613139e-06,
+      "loss": 0.5289,
+      "step": 607
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.25985401459854e-06,
+      "loss": 0.5758,
+      "step": 608
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.2583941605839415e-06,
+      "loss": 0.5778,
+      "step": 609
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.2569343065693435e-06,
+      "loss": 0.5334,
+      "step": 610
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.255474452554745e-06,
+      "loss": 0.4948,
+      "step": 611
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.254014598540147e-06,
+      "loss": 0.586,
+      "step": 612
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.252554744525547e-06,
+      "loss": 0.5082,
+      "step": 613
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.251094890510949e-06,
+      "loss": 0.5152,
+      "step": 614
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.2496350364963504e-06,
+      "loss": 0.4776,
+      "step": 615
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.248175182481752e-06,
+      "loss": 0.5285,
+      "step": 616
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.246715328467154e-06,
+      "loss": 0.5008,
+      "step": 617
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.245255474452555e-06,
+      "loss": 0.5487,
+      "step": 618
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.243795620437956e-06,
+      "loss": 0.5135,
+      "step": 619
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.242335766423358e-06,
+      "loss": 0.5256,
+      "step": 620
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.240875912408759e-06,
+      "loss": 0.563,
+      "step": 621
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.239416058394161e-06,
+      "loss": 0.5271,
+      "step": 622
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.237956204379562e-06,
+      "loss": 0.6072,
+      "step": 623
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.236496350364963e-06,
+      "loss": 0.5086,
+      "step": 624
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.235036496350365e-06,
+      "loss": 0.5328,
+      "step": 625
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.233576642335766e-06,
+      "loss": 0.6007,
+      "step": 626
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.232116788321168e-06,
+      "loss": 0.4774,
+      "step": 627
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.23065693430657e-06,
+      "loss": 0.5422,
+      "step": 628
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.229197080291971e-06,
+      "loss": 0.531,
+      "step": 629
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.227737226277373e-06,
+      "loss": 0.5413,
+      "step": 630
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.226277372262774e-06,
+      "loss": 0.5573,
+      "step": 631
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.224817518248175e-06,
+      "loss": 0.5392,
+      "step": 632
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.2233576642335765e-06,
+      "loss": 0.5471,
+      "step": 633
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5424632430076599,
+      "eval_runtime": 1.125,
+      "eval_samples_per_second": 17.778,
+      "eval_steps_per_second": 0.889,
+      "step": 633
+    },
+    {
+      "epoch": 0.3,
+      "eval_bench_accuracy_agieval": 0.3389830508474576,
+      "eval_bench_accuracy_arc_challenge": 0.8703703703703703,
+      "eval_bench_accuracy_arc_easy": 0.8888888888888888,
+      "eval_bench_accuracy_bigbench": 0.4918032786885246,
+      "eval_bench_accuracy_boolq": 0.7407407407407407,
+      "eval_bench_accuracy_mmlu": 0.46296296296296297,
+      "eval_bench_accuracy_openbookqa": 0.2222222222222222,
+      "eval_bench_accuracy_truthful_qa": 0.32075471698113206,
+      "eval_bench_accuracy_winogrande": 0.5,
+      "eval_bench_average_accuracy": 0.5374140257447,
+      "eval_bench_loss": 16.713373023343372,
+      "eval_bench_total_accuracy": 0.5352112676056338,
+      "step": 633
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.221897810218978e-06,
+      "loss": 0.552,
+      "step": 634
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.22043795620438e-06,
+      "loss": 0.5124,
+      "step": 635
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.218978102189781e-06,
+      "loss": 0.6002,
+      "step": 636
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.217518248175182e-06,
+      "loss": 0.5047,
+      "step": 637
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.216058394160584e-06,
+      "loss": 0.5469,
+      "step": 638
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.2145985401459855e-06,
+      "loss": 0.5658,
+      "step": 639
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.2131386861313876e-06,
+      "loss": 0.5638,
+      "step": 640
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.211678832116789e-06,
+      "loss": 0.53,
+      "step": 641
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.21021897810219e-06,
+      "loss": 0.4657,
+      "step": 642
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.208759124087591e-06,
+      "loss": 0.5076,
+      "step": 643
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.207299270072992e-06,
+      "loss": 0.5619,
+      "step": 644
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.2058394160583945e-06,
+      "loss": 0.5339,
+      "step": 645
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.204379562043796e-06,
+      "loss": 0.48,
+      "step": 646
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.202919708029197e-06,
+      "loss": 0.5131,
+      "step": 647
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.201459854014599e-06,
+      "loss": 0.4938,
+      "step": 648
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.2e-06,
+      "loss": 0.5198,
+      "step": 649
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.198540145985402e-06,
+      "loss": 0.5554,
+      "step": 650
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.1970802919708035e-06,
+      "loss": 0.5277,
+      "step": 651
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.195620437956204e-06,
+      "loss": 0.5315,
+      "step": 652
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.194160583941606e-06,
+      "loss": 0.5053,
+      "step": 653
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.192700729927007e-06,
+      "loss": 0.5428,
+      "step": 654
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.191240875912409e-06,
+      "loss": 0.4732,
+      "step": 655
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.18978102189781e-06,
+      "loss": 0.5067,
+      "step": 656
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.188321167883212e-06,
+      "loss": 0.5646,
+      "step": 657
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.186861313868614e-06,
+      "loss": 0.5052,
+      "step": 658
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.185401459854015e-06,
+      "loss": 0.5732,
+      "step": 659
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.183941605839416e-06,
+      "loss": 0.5934,
+      "step": 660
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.182481751824818e-06,
+      "loss": 0.4952,
+      "step": 661
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.1810218978102185e-06,
+      "loss": 0.5484,
+      "step": 662
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.1795620437956206e-06,
+      "loss": 0.5354,
+      "step": 663
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.178102189781022e-06,
+      "loss": 0.5366,
+      "step": 664
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.176642335766424e-06,
+      "loss": 0.5209,
+      "step": 665
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.175182481751825e-06,
+      "loss": 0.5949,
+      "step": 666
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.173722627737226e-06,
+      "loss": 0.5631,
+      "step": 667
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.172262773722628e-06,
+      "loss": 0.5497,
+      "step": 668
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.1708029197080296e-06,
+      "loss": 0.6135,
+      "step": 669
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.169343065693431e-06,
+      "loss": 0.5099,
+      "step": 670
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.167883211678832e-06,
+      "loss": 0.5655,
+      "step": 671
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.166423357664233e-06,
+      "loss": 0.5381,
+      "step": 672
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.164963503649635e-06,
+      "loss": 0.49,
+      "step": 673
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.1635036496350365e-06,
+      "loss": 0.5495,
+      "step": 674
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.162043795620438e-06,
+      "loss": 0.5558,
+      "step": 675
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.16058394160584e-06,
+      "loss": 0.5188,
+      "step": 676
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.159124087591241e-06,
+      "loss": 0.518,
+      "step": 677
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.157664233576643e-06,
+      "loss": 0.5162,
+      "step": 678
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.156204379562044e-06,
+      "loss": 0.5148,
+      "step": 679
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.1547445255474455e-06,
+      "loss": 0.5436,
+      "step": 680
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.153284671532847e-06,
+      "loss": 0.511,
+      "step": 681
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.151824817518248e-06,
+      "loss": 0.5008,
+      "step": 682
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.15036496350365e-06,
+      "loss": 0.5191,
+      "step": 683
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.148905109489051e-06,
+      "loss": 0.511,
+      "step": 684
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.147445255474452e-06,
+      "loss": 0.4978,
+      "step": 685
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.1459854014598544e-06,
+      "loss": 0.5188,
+      "step": 686
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.144525547445256e-06,
+      "loss": 0.5052,
+      "step": 687
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.143065693430657e-06,
+      "loss": 0.4945,
+      "step": 688
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.141605839416059e-06,
+      "loss": 0.5191,
+      "step": 689
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.14014598540146e-06,
+      "loss": 0.5894,
+      "step": 690
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.138686131386861e-06,
+      "loss": 0.5498,
+      "step": 691
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.1372262773722626e-06,
+      "loss": 0.4984,
+      "step": 692
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.135766423357665e-06,
+      "loss": 0.5035,
+      "step": 693
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.134306569343066e-06,
+      "loss": 0.5199,
+      "step": 694
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.132846715328467e-06,
+      "loss": 0.4868,
+      "step": 695
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.131386861313869e-06,
+      "loss": 0.4989,
+      "step": 696
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.12992700729927e-06,
+      "loss": 0.5176,
+      "step": 697
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.1284671532846715e-06,
+      "loss": 0.5165,
+      "step": 698
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.127007299270074e-06,
+      "loss": 0.518,
+      "step": 699
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.125547445255474e-06,
+      "loss": 0.5093,
+      "step": 700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.124087591240876e-06,
+      "loss": 0.5772,
+      "step": 701
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.122627737226277e-06,
+      "loss": 0.506,
+      "step": 702
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.1211678832116785e-06,
+      "loss": 0.5069,
+      "step": 703
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.1197080291970805e-06,
+      "loss": 0.5505,
+      "step": 704
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.118248175182482e-06,
+      "loss": 0.6013,
+      "step": 705
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.116788321167884e-06,
+      "loss": 0.5727,
+      "step": 706
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.115328467153285e-06,
+      "loss": 0.5109,
+      "step": 707
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.113868613138686e-06,
+      "loss": 0.4733,
+      "step": 708
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.112408759124088e-06,
+      "loss": 0.5339,
+      "step": 709
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.110948905109489e-06,
+      "loss": 0.5001,
+      "step": 710
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.109489051094891e-06,
+      "loss": 0.5373,
+      "step": 711
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.108029197080292e-06,
+      "loss": 0.5454,
+      "step": 712
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.106569343065693e-06,
+      "loss": 0.5428,
+      "step": 713
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.105109489051095e-06,
+      "loss": 0.5133,
+      "step": 714
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.103649635036496e-06,
+      "loss": 0.5568,
+      "step": 715
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.102189781021898e-06,
+      "loss": 0.4889,
+      "step": 716
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.1007299270073e-06,
+      "loss": 0.5389,
+      "step": 717
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.099270072992701e-06,
+      "loss": 0.5418,
+      "step": 718
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.097810218978103e-06,
+      "loss": 0.5522,
+      "step": 719
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.096350364963503e-06,
+      "loss": 0.5706,
+      "step": 720
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.094890510948905e-06,
+      "loss": 0.5042,
+      "step": 721
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.093430656934307e-06,
+      "loss": 0.5106,
+      "step": 722
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.091970802919708e-06,
+      "loss": 0.5468,
+      "step": 723
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.09051094890511e-06,
+      "loss": 0.5174,
+      "step": 724
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.089051094890511e-06,
+      "loss": 0.5867,
+      "step": 725
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.087591240875912e-06,
+      "loss": 0.5187,
+      "step": 726
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.086131386861314e-06,
+      "loss": 0.5208,
+      "step": 727
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.084671532846716e-06,
+      "loss": 0.5311,
+      "step": 728
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.083211678832118e-06,
+      "loss": 0.5366,
+      "step": 729
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.081751824817518e-06,
+      "loss": 0.5313,
+      "step": 730
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.080291970802919e-06,
+      "loss": 0.567,
+      "step": 731
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.078832116788321e-06,
+      "loss": 0.5044,
+      "step": 732
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.0773722627737225e-06,
+      "loss": 0.513,
+      "step": 733
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.0759124087591246e-06,
+      "loss": 0.5247,
+      "step": 734
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.074452554744526e-06,
+      "loss": 0.5158,
+      "step": 735
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.072992700729927e-06,
+      "loss": 0.4968,
+      "step": 736
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.071532846715329e-06,
+      "loss": 0.5315,
+      "step": 737
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.07007299270073e-06,
+      "loss": 0.5546,
+      "step": 738
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.0686131386861315e-06,
+      "loss": 0.4924,
+      "step": 739
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.067153284671533e-06,
+      "loss": 0.5267,
+      "step": 740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.065693430656934e-06,
+      "loss": 0.5249,
+      "step": 741
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.064233576642336e-06,
+      "loss": 0.5843,
+      "step": 742
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.062773722627737e-06,
+      "loss": 0.515,
+      "step": 743
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.061313868613139e-06,
+      "loss": 0.4329,
+      "step": 744
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.0598540145985405e-06,
+      "loss": 0.5492,
+      "step": 745
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.058394160583942e-06,
+      "loss": 0.5309,
+      "step": 746
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.056934306569344e-06,
+      "loss": 0.5446,
+      "step": 747
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.055474452554745e-06,
+      "loss": 0.4777,
+      "step": 748
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.054014598540146e-06,
+      "loss": 0.5137,
+      "step": 749
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.052554744525547e-06,
+      "loss": 0.5363,
+      "step": 750
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.051094890510949e-06,
+      "loss": 0.5353,
+      "step": 751
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.049635036496351e-06,
+      "loss": 0.5448,
+      "step": 752
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.048175182481752e-06,
+      "loss": 0.5075,
+      "step": 753
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.046715328467153e-06,
+      "loss": 0.5113,
+      "step": 754
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.045255474452555e-06,
+      "loss": 0.5904,
+      "step": 755
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.043795620437956e-06,
+      "loss": 0.5363,
+      "step": 756
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.042335766423358e-06,
+      "loss": 0.5162,
+      "step": 757
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.04087591240876e-06,
+      "loss": 0.4882,
+      "step": 758
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.03941605839416e-06,
+      "loss": 0.5593,
+      "step": 759
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.037956204379562e-06,
+      "loss": 0.5095,
+      "step": 760
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.036496350364963e-06,
+      "loss": 0.5196,
+      "step": 761
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.035036496350365e-06,
+      "loss": 0.5469,
+      "step": 762
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.0335766423357666e-06,
+      "loss": 0.5728,
+      "step": 763
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.032116788321168e-06,
+      "loss": 0.5499,
+      "step": 764
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.03065693430657e-06,
+      "loss": 0.5496,
+      "step": 765
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.029197080291971e-06,
+      "loss": 0.4857,
+      "step": 766
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.027737226277372e-06,
+      "loss": 0.5557,
+      "step": 767
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.026277372262774e-06,
+      "loss": 0.5236,
+      "step": 768
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.024817518248175e-06,
+      "loss": 0.5075,
+      "step": 769
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.023357664233577e-06,
+      "loss": 0.545,
+      "step": 770
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.021897810218978e-06,
+      "loss": 0.5244,
+      "step": 771
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.02043795620438e-06,
+      "loss": 0.467,
+      "step": 772
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.018978102189781e-06,
+      "loss": 0.5316,
+      "step": 773
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.0175182481751825e-06,
+      "loss": 0.5346,
+      "step": 774
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.0160583941605845e-06,
+      "loss": 0.4975,
+      "step": 775
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.014598540145986e-06,
+      "loss": 0.5197,
+      "step": 776
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.013138686131387e-06,
+      "loss": 0.5053,
+      "step": 777
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.011678832116788e-06,
+      "loss": 0.5306,
+      "step": 778
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.010218978102189e-06,
+      "loss": 0.5218,
+      "step": 779
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.0087591240875914e-06,
+      "loss": 0.48,
+      "step": 780
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.007299270072993e-06,
+      "loss": 0.4852,
+      "step": 781
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.005839416058394e-06,
+      "loss": 0.5901,
+      "step": 782
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.004379562043796e-06,
+      "loss": 0.5195,
+      "step": 783
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.002919708029197e-06,
+      "loss": 0.5052,
+      "step": 784
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.001459854014599e-06,
+      "loss": 0.5573,
+      "step": 785
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-06,
+      "loss": 0.5849,
+      "step": 786
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.998540145985402e-06,
+      "loss": 0.516,
+      "step": 787
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.997080291970803e-06,
+      "loss": 0.5731,
+      "step": 788
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.995620437956204e-06,
+      "loss": 0.5341,
+      "step": 789
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.994160583941606e-06,
+      "loss": 0.5232,
+      "step": 790
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.992700729927007e-06,
+      "loss": 0.492,
+      "step": 791
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.9912408759124085e-06,
+      "loss": 0.5274,
+      "step": 792
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.989781021897811e-06,
+      "loss": 0.5392,
+      "step": 793
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.988321167883212e-06,
+      "loss": 0.57,
+      "step": 794
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.986861313868613e-06,
+      "loss": 0.5384,
+      "step": 795
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.985401459854015e-06,
+      "loss": 0.5197,
+      "step": 796
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.983941605839416e-06,
+      "loss": 0.5136,
+      "step": 797
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.9824817518248175e-06,
+      "loss": 0.5441,
+      "step": 798
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.981021897810219e-06,
+      "loss": 0.5442,
+      "step": 799
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.979562043795621e-06,
+      "loss": 0.5185,
+      "step": 800
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.978102189781022e-06,
+      "loss": 0.5695,
+      "step": 801
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.976642335766423e-06,
+      "loss": 0.5176,
+      "step": 802
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.975182481751825e-06,
+      "loss": 0.5367,
+      "step": 803
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.9737226277372265e-06,
+      "loss": 0.5128,
+      "step": 804
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.972262773722628e-06,
+      "loss": 0.4656,
+      "step": 805
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.97080291970803e-06,
+      "loss": 0.5066,
+      "step": 806
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.969343065693431e-06,
+      "loss": 0.5275,
+      "step": 807
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.967883211678832e-06,
+      "loss": 0.5492,
+      "step": 808
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.966423357664233e-06,
+      "loss": 0.5149,
+      "step": 809
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.964963503649635e-06,
+      "loss": 0.494,
+      "step": 810
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.963503649635037e-06,
+      "loss": 0.5126,
+      "step": 811
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.962043795620438e-06,
+      "loss": 0.5332,
+      "step": 812
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.96058394160584e-06,
+      "loss": 0.5719,
+      "step": 813
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.959124087591241e-06,
+      "loss": 0.5342,
+      "step": 814
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.957664233576642e-06,
+      "loss": 0.5578,
+      "step": 815
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.9562043795620445e-06,
+      "loss": 0.5472,
+      "step": 816
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.954744525547445e-06,
+      "loss": 0.5289,
+      "step": 817
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.953284671532847e-06,
+      "loss": 0.4921,
+      "step": 818
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.951824817518248e-06,
+      "loss": 0.4952,
+      "step": 819
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.950364963503649e-06,
+      "loss": 0.4877,
+      "step": 820
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.948905109489051e-06,
+      "loss": 0.4727,
+      "step": 821
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.947445255474453e-06,
+      "loss": 0.5279,
+      "step": 822
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.945985401459855e-06,
+      "loss": 0.5616,
+      "step": 823
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.944525547445256e-06,
+      "loss": 0.561,
+      "step": 824
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.943065693430657e-06,
+      "loss": 0.5189,
+      "step": 825
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.941605839416059e-06,
+      "loss": 0.5639,
+      "step": 826
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.9401459854014595e-06,
+      "loss": 0.4805,
+      "step": 827
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.9386861313868616e-06,
+      "loss": 0.4877,
+      "step": 828
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.937226277372263e-06,
+      "loss": 0.5291,
+      "step": 829
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.935766423357664e-06,
+      "loss": 0.4921,
+      "step": 830
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.934306569343066e-06,
+      "loss": 0.5556,
+      "step": 831
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.932846715328467e-06,
+      "loss": 0.4926,
+      "step": 832
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.9313868613138685e-06,
+      "loss": 0.5229,
+      "step": 833
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.9299270072992705e-06,
+      "loss": 0.5369,
+      "step": 834
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.928467153284672e-06,
+      "loss": 0.5659,
+      "step": 835
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.927007299270074e-06,
+      "loss": 0.5128,
+      "step": 836
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.925547445255474e-06,
+      "loss": 0.5415,
+      "step": 837
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.924087591240875e-06,
+      "loss": 0.4864,
+      "step": 838
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.9226277372262775e-06,
+      "loss": 0.5657,
+      "step": 839
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.921167883211679e-06,
+      "loss": 0.5297,
+      "step": 840
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.919708029197081e-06,
+      "loss": 0.5253,
+      "step": 841
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.918248175182482e-06,
+      "loss": 0.545,
+      "step": 842
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.916788321167883e-06,
+      "loss": 0.5161,
+      "step": 843
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.915328467153285e-06,
+      "loss": 0.5334,
+      "step": 844
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5264854431152344,
+      "eval_runtime": 1.1186,
+      "eval_samples_per_second": 17.879,
+      "eval_steps_per_second": 0.894,
+      "step": 844
+    },
+    {
+      "epoch": 0.4,
+      "eval_bench_accuracy_agieval": 0.3220338983050847,
+      "eval_bench_accuracy_arc_challenge": 0.8148148148148148,
+      "eval_bench_accuracy_arc_easy": 0.8703703703703703,
+      "eval_bench_accuracy_bigbench": 0.5409836065573771,
+      "eval_bench_accuracy_boolq": 0.7592592592592593,
+      "eval_bench_accuracy_mmlu": 0.42592592592592593,
+      "eval_bench_accuracy_openbookqa": 0.25925925925925924,
+      "eval_bench_accuracy_truthful_qa": 0.22641509433962265,
+      "eval_bench_accuracy_winogrande": 0.5,
+      "eval_bench_average_accuracy": 0.5243402476479682,
+      "eval_bench_loss": 14.964627553181476,
+      "eval_bench_total_accuracy": 0.5231388329979879,
+      "step": 844
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.9138686131386864e-06,
+      "loss": 0.5166,
+      "step": 845
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.912408759124088e-06,
+      "loss": 0.5304,
+      "step": 846
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.910948905109489e-06,
+      "loss": 0.4453,
+      "step": 847
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.90948905109489e-06,
+      "loss": 0.541,
+      "step": 848
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.908029197080292e-06,
+      "loss": 0.509,
+      "step": 849
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.906569343065693e-06,
+      "loss": 0.5155,
+      "step": 850
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.905109489051095e-06,
+      "loss": 0.5263,
+      "step": 851
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.903649635036497e-06,
+      "loss": 0.5437,
+      "step": 852
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.902189781021898e-06,
+      "loss": 0.5398,
+      "step": 853
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.9007299270073e-06,
+      "loss": 0.5162,
+      "step": 854
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.899270072992701e-06,
+      "loss": 0.5545,
+      "step": 855
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.897810218978102e-06,
+      "loss": 0.4768,
+      "step": 856
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.8963503649635036e-06,
+      "loss": 0.5148,
+      "step": 857
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.894890510948905e-06,
+      "loss": 0.5191,
+      "step": 858
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.893430656934307e-06,
+      "loss": 0.5264,
+      "step": 859
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.891970802919708e-06,
+      "loss": 0.518,
+      "step": 860
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.890510948905109e-06,
+      "loss": 0.5344,
+      "step": 861
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.889051094890511e-06,
+      "loss": 0.4997,
+      "step": 862
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.8875912408759125e-06,
+      "loss": 0.5072,
+      "step": 863
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.886131386861315e-06,
+      "loss": 0.5579,
+      "step": 864
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.884671532846716e-06,
+      "loss": 0.5308,
+      "step": 865
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.883211678832117e-06,
+      "loss": 0.508,
+      "step": 866
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.881751824817518e-06,
+      "loss": 0.4822,
+      "step": 867
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.8802919708029195e-06,
+      "loss": 0.4854,
+      "step": 868
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.8788321167883215e-06,
+      "loss": 0.5695,
+      "step": 869
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.877372262773723e-06,
+      "loss": 0.5433,
+      "step": 870
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.875912408759124e-06,
+      "loss": 0.4957,
+      "step": 871
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.874452554744526e-06,
+      "loss": 0.5367,
+      "step": 872
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.872992700729927e-06,
+      "loss": 0.4614,
+      "step": 873
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8715328467153284e-06,
+      "loss": 0.5146,
+      "step": 874
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8700729927007305e-06,
+      "loss": 0.5478,
+      "step": 875
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.868613138686131e-06,
+      "loss": 0.4987,
+      "step": 876
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.867153284671533e-06,
+      "loss": 0.5383,
+      "step": 877
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.865693430656934e-06,
+      "loss": 0.565,
+      "step": 878
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.864233576642336e-06,
+      "loss": 0.4996,
+      "step": 879
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.862773722627737e-06,
+      "loss": 0.5533,
+      "step": 880
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.861313868613139e-06,
+      "loss": 0.5467,
+      "step": 881
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.859854014598541e-06,
+      "loss": 0.5251,
+      "step": 882
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.858394160583942e-06,
+      "loss": 0.4945,
+      "step": 883
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.856934306569343e-06,
+      "loss": 0.5389,
+      "step": 884
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.855474452554745e-06,
+      "loss": 0.5883,
+      "step": 885
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8540145985401455e-06,
+      "loss": 0.5336,
+      "step": 886
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.852554744525548e-06,
+      "loss": 0.4952,
+      "step": 887
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.851094890510949e-06,
+      "loss": 0.4709,
+      "step": 888
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.84963503649635e-06,
+      "loss": 0.4955,
+      "step": 889
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.848175182481752e-06,
+      "loss": 0.5323,
+      "step": 890
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.846715328467153e-06,
+      "loss": 0.5109,
+      "step": 891
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.845255474452555e-06,
+      "loss": 0.5434,
+      "step": 892
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.843795620437957e-06,
+      "loss": 0.5217,
+      "step": 893
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.842335766423358e-06,
+      "loss": 0.5137,
+      "step": 894
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.840875912408759e-06,
+      "loss": 0.4823,
+      "step": 895
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.83941605839416e-06,
+      "loss": 0.545,
+      "step": 896
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.837956204379562e-06,
+      "loss": 0.488,
+      "step": 897
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.8364963503649635e-06,
+      "loss": 0.5193,
+      "step": 898
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.835036496350365e-06,
+      "loss": 0.465,
+      "step": 899
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.833576642335767e-06,
+      "loss": 0.5373,
+      "step": 900
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.832116788321168e-06,
+      "loss": 0.5618,
+      "step": 901
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.83065693430657e-06,
+      "loss": 0.4963,
+      "step": 902
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.829197080291971e-06,
+      "loss": 0.5382,
+      "step": 903
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.8277372262773725e-06,
+      "loss": 0.5713,
+      "step": 904
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.826277372262774e-06,
+      "loss": 0.481,
+      "step": 905
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.824817518248175e-06,
+      "loss": 0.5402,
+      "step": 906
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.823357664233577e-06,
+      "loss": 0.5199,
+      "step": 907
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.821897810218978e-06,
+      "loss": 0.5652,
+      "step": 908
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.820437956204379e-06,
+      "loss": 0.4745,
+      "step": 909
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.8189781021897815e-06,
+      "loss": 0.5659,
+      "step": 910
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.817518248175183e-06,
+      "loss": 0.4606,
+      "step": 911
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.816058394160584e-06,
+      "loss": 0.4931,
+      "step": 912
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.814598540145986e-06,
+      "loss": 0.5674,
+      "step": 913
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.813138686131387e-06,
+      "loss": 0.511,
+      "step": 914
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.811678832116788e-06,
+      "loss": 0.4995,
+      "step": 915
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.81021897810219e-06,
+      "loss": 0.4969,
+      "step": 916
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.808759124087591e-06,
+      "loss": 0.4888,
+      "step": 917
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.807299270072993e-06,
+      "loss": 0.4936,
+      "step": 918
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.805839416058394e-06,
+      "loss": 0.55,
+      "step": 919
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.804379562043796e-06,
+      "loss": 0.5184,
+      "step": 920
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.802919708029197e-06,
+      "loss": 0.4978,
+      "step": 921
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.8014598540145986e-06,
+      "loss": 0.5589,
+      "step": 922
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.4712,
+      "step": 923
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.798540145985401e-06,
+      "loss": 0.4463,
+      "step": 924
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.797080291970803e-06,
+      "loss": 0.5611,
+      "step": 925
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.795620437956204e-06,
+      "loss": 0.5493,
+      "step": 926
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7941605839416055e-06,
+      "loss": 0.4838,
+      "step": 927
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7927007299270075e-06,
+      "loss": 0.475,
+      "step": 928
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.791240875912409e-06,
+      "loss": 0.4895,
+      "step": 929
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.789781021897811e-06,
+      "loss": 0.5481,
+      "step": 930
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.788321167883212e-06,
+      "loss": 0.4799,
+      "step": 931
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.786861313868613e-06,
+      "loss": 0.508,
+      "step": 932
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.785401459854015e-06,
+      "loss": 0.5577,
+      "step": 933
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.783941605839416e-06,
+      "loss": 0.5389,
+      "step": 934
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.782481751824818e-06,
+      "loss": 0.5611,
+      "step": 935
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.781021897810219e-06,
+      "loss": 0.5356,
+      "step": 936
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.77956204379562e-06,
+      "loss": 0.5193,
+      "step": 937
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.778102189781022e-06,
+      "loss": 0.4746,
+      "step": 938
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7766423357664234e-06,
+      "loss": 0.4838,
+      "step": 939
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.775182481751825e-06,
+      "loss": 0.5102,
+      "step": 940
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.773722627737227e-06,
+      "loss": 0.4962,
+      "step": 941
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.772262773722628e-06,
+      "loss": 0.5301,
+      "step": 942
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.77080291970803e-06,
+      "loss": 0.4749,
+      "step": 943
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.76934306569343e-06,
+      "loss": 0.557,
+      "step": 944
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.767883211678832e-06,
+      "loss": 0.5293,
+      "step": 945
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.766423357664234e-06,
+      "loss": 0.4891,
+      "step": 946
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.764963503649635e-06,
+      "loss": 0.4732,
+      "step": 947
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.763503649635037e-06,
+      "loss": 0.5527,
+      "step": 948
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.762043795620438e-06,
+      "loss": 0.5302,
+      "step": 949
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.760583941605839e-06,
+      "loss": 0.4811,
+      "step": 950
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.759124087591241e-06,
+      "loss": 0.4882,
+      "step": 951
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.757664233576643e-06,
+      "loss": 0.5046,
+      "step": 952
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.756204379562044e-06,
+      "loss": 0.5081,
+      "step": 953
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.754744525547445e-06,
+      "loss": 0.5389,
+      "step": 954
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.753284671532846e-06,
+      "loss": 0.5437,
+      "step": 955
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.751824817518248e-06,
+      "loss": 0.4628,
+      "step": 956
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7503649635036495e-06,
+      "loss": 0.5357,
+      "step": 957
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.748905109489052e-06,
+      "loss": 0.4658,
+      "step": 958
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.747445255474453e-06,
+      "loss": 0.5376,
+      "step": 959
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.745985401459854e-06,
+      "loss": 0.4793,
+      "step": 960
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.744525547445256e-06,
+      "loss": 0.5281,
+      "step": 961
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.743065693430657e-06,
+      "loss": 0.5467,
+      "step": 962
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7416058394160585e-06,
+      "loss": 0.5169,
+      "step": 963
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.74014598540146e-06,
+      "loss": 0.5373,
+      "step": 964
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.738686131386861e-06,
+      "loss": 0.5148,
+      "step": 965
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.737226277372263e-06,
+      "loss": 0.5403,
+      "step": 966
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.735766423357664e-06,
+      "loss": 0.5379,
+      "step": 967
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7343065693430654e-06,
+      "loss": 0.4902,
+      "step": 968
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7328467153284675e-06,
+      "loss": 0.4787,
+      "step": 969
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.731386861313869e-06,
+      "loss": 0.5686,
+      "step": 970
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.729927007299271e-06,
+      "loss": 0.5215,
+      "step": 971
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.728467153284672e-06,
+      "loss": 0.5088,
+      "step": 972
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.727007299270073e-06,
+      "loss": 0.5425,
+      "step": 973
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.725547445255474e-06,
+      "loss": 0.4391,
+      "step": 974
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.724087591240876e-06,
+      "loss": 0.5201,
+      "step": 975
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.722627737226278e-06,
+      "loss": 0.4819,
+      "step": 976
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.721167883211679e-06,
+      "loss": 0.5379,
+      "step": 977
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.71970802919708e-06,
+      "loss": 0.5312,
+      "step": 978
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.718248175182482e-06,
+      "loss": 0.5242,
+      "step": 979
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.716788321167883e-06,
+      "loss": 0.5515,
+      "step": 980
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7153284671532855e-06,
+      "loss": 0.5041,
+      "step": 981
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.713868613138687e-06,
+      "loss": 0.5016,
+      "step": 982
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.712408759124087e-06,
+      "loss": 0.4898,
+      "step": 983
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.710948905109489e-06,
+      "loss": 0.4947,
+      "step": 984
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.70948905109489e-06,
+      "loss": 0.5417,
+      "step": 985
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.708029197080292e-06,
+      "loss": 0.5268,
+      "step": 986
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.706569343065694e-06,
+      "loss": 0.5072,
+      "step": 987
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.705109489051095e-06,
+      "loss": 0.488,
+      "step": 988
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.703649635036497e-06,
+      "loss": 0.5013,
+      "step": 989
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.702189781021898e-06,
+      "loss": 0.5512,
+      "step": 990
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.700729927007299e-06,
+      "loss": 0.4908,
+      "step": 991
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.699270072992701e-06,
+      "loss": 0.4731,
+      "step": 992
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.697810218978102e-06,
+      "loss": 0.5513,
+      "step": 993
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.696350364963504e-06,
+      "loss": 0.5199,
+      "step": 994
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.694890510948905e-06,
+      "loss": 0.4863,
+      "step": 995
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.693430656934306e-06,
+      "loss": 0.4756,
+      "step": 996
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.691970802919708e-06,
+      "loss": 0.5812,
+      "step": 997
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.6905109489051095e-06,
+      "loss": 0.5281,
+      "step": 998
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.6890510948905115e-06,
+      "loss": 0.5602,
+      "step": 999
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.687591240875913e-06,
+      "loss": 0.5544,
+      "step": 1000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.686131386861314e-06,
+      "loss": 0.5094,
+      "step": 1001
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.684671532846715e-06,
+      "loss": 0.5093,
+      "step": 1002
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.683211678832116e-06,
+      "loss": 0.5825,
+      "step": 1003
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6817518248175185e-06,
+      "loss": 0.5176,
+      "step": 1004
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.68029197080292e-06,
+      "loss": 0.5056,
+      "step": 1005
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.678832116788321e-06,
+      "loss": 0.5798,
+      "step": 1006
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.677372262773723e-06,
+      "loss": 0.5447,
+      "step": 1007
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.675912408759124e-06,
+      "loss": 0.5371,
+      "step": 1008
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.674452554744526e-06,
+      "loss": 0.5778,
+      "step": 1009
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6729927007299274e-06,
+      "loss": 0.5478,
+      "step": 1010
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.671532846715329e-06,
+      "loss": 0.488,
+      "step": 1011
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.67007299270073e-06,
+      "loss": 0.5794,
+      "step": 1012
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.668613138686131e-06,
+      "loss": 0.5087,
+      "step": 1013
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.667153284671533e-06,
+      "loss": 0.5062,
+      "step": 1014
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.665693430656934e-06,
+      "loss": 0.5124,
+      "step": 1015
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6642335766423356e-06,
+      "loss": 0.4801,
+      "step": 1016
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.662773722627738e-06,
+      "loss": 0.5204,
+      "step": 1017
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.661313868613139e-06,
+      "loss": 0.5612,
+      "step": 1018
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.65985401459854e-06,
+      "loss": 0.507,
+      "step": 1019
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.658394160583942e-06,
+      "loss": 0.4785,
+      "step": 1020
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.656934306569343e-06,
+      "loss": 0.4894,
+      "step": 1021
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.6554744525547445e-06,
+      "loss": 0.5522,
+      "step": 1022
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.654014598540146e-06,
+      "loss": 0.5049,
+      "step": 1023
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.652554744525548e-06,
+      "loss": 0.4875,
+      "step": 1024
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.651094890510949e-06,
+      "loss": 0.552,
+      "step": 1025
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.64963503649635e-06,
+      "loss": 0.5022,
+      "step": 1026
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.648175182481752e-06,
+      "loss": 0.4674,
+      "step": 1027
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.6467153284671535e-06,
+      "loss": 0.4984,
+      "step": 1028
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.645255474452555e-06,
+      "loss": 0.5538,
+      "step": 1029
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.643795620437957e-06,
+      "loss": 0.5154,
+      "step": 1030
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.642335766423358e-06,
+      "loss": 0.5279,
+      "step": 1031
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.640875912408759e-06,
+      "loss": 0.5202,
+      "step": 1032
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.6394160583941604e-06,
+      "loss": 0.538,
+      "step": 1033
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.637956204379562e-06,
+      "loss": 0.5164,
+      "step": 1034
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.636496350364964e-06,
+      "loss": 0.542,
+      "step": 1035
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.635036496350365e-06,
+      "loss": 0.5964,
+      "step": 1036
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.633576642335767e-06,
+      "loss": 0.4519,
+      "step": 1037
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.632116788321168e-06,
+      "loss": 0.4942,
+      "step": 1038
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.6306569343065694e-06,
+      "loss": 0.4875,
+      "step": 1039
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.6291970802919715e-06,
+      "loss": 0.486,
+      "step": 1040
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.627737226277372e-06,
+      "loss": 0.496,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.626277372262774e-06,
+      "loss": 0.5729,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.624817518248175e-06,
+      "loss": 0.4955,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.623357664233576e-06,
+      "loss": 0.5351,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.621897810218978e-06,
+      "loss": 0.5151,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.62043795620438e-06,
+      "loss": 0.4705,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.618978102189781e-06,
+      "loss": 0.4846,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.617518248175183e-06,
+      "loss": 0.4662,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.616058394160584e-06,
+      "loss": 0.4338,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.614598540145986e-06,
+      "loss": 0.5011,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6131386861313865e-06,
+      "loss": 0.5102,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.611678832116789e-06,
+      "loss": 0.4788,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.61021897810219e-06,
+      "loss": 0.5049,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.608759124087591e-06,
+      "loss": 0.5088,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.607299270072993e-06,
+      "loss": 0.5317,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5172545313835144,
+      "eval_runtime": 1.1301,
+      "eval_samples_per_second": 17.698,
+      "eval_steps_per_second": 0.885,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5,
+      "eval_bench_accuracy_agieval": 0.3220338983050847,
+      "eval_bench_accuracy_arc_challenge": 0.8518518518518519,
+      "eval_bench_accuracy_arc_easy": 0.9074074074074074,
+      "eval_bench_accuracy_bigbench": 0.5081967213114754,
+      "eval_bench_accuracy_boolq": 0.7777777777777778,
+      "eval_bench_accuracy_mmlu": 0.46296296296296297,
+      "eval_bench_accuracy_openbookqa": 0.24074074074074073,
+      "eval_bench_accuracy_truthful_qa": 0.3018867924528302,
+      "eval_bench_accuracy_winogrande": 0.46296296296296297,
+      "eval_bench_average_accuracy": 0.5373134573081216,
+      "eval_bench_loss": 15.405342561652862,
+      "eval_bench_total_accuracy": 0.5352112676056338,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.605839416058394e-06,
+      "loss": 0.5167,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6043795620437955e-06,
+      "loss": 0.4841,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6029197080291976e-06,
+      "loss": 0.509,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.601459854014599e-06,
+      "loss": 0.5137,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.5226,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.598540145985401e-06,
+      "loss": 0.5168,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.5970802919708024e-06,
+      "loss": 0.492,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.5956204379562045e-06,
+      "loss": 0.5148,
+      "step": 1063
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.594160583941606e-06,
+      "loss": 0.494,
+      "step": 1064
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.592700729927008e-06,
+      "loss": 0.4941,
+      "step": 1065
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.591240875912409e-06,
+      "loss": 0.5275,
+      "step": 1066
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.58978102189781e-06,
+      "loss": 0.557,
+      "step": 1067
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.588321167883212e-06,
+      "loss": 0.5311,
+      "step": 1068
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.5868613138686135e-06,
+      "loss": 0.5052,
+      "step": 1069
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.585401459854015e-06,
+      "loss": 0.5098,
+      "step": 1070
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.583941605839416e-06,
+      "loss": 0.535,
+      "step": 1071
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.582481751824817e-06,
+      "loss": 0.541,
+      "step": 1072
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.581021897810219e-06,
+      "loss": 0.479,
+      "step": 1073
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.57956204379562e-06,
+      "loss": 0.4895,
+      "step": 1074
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.578102189781022e-06,
+      "loss": 0.5406,
+      "step": 1075
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.576642335766424e-06,
+      "loss": 0.5258,
+      "step": 1076
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.575182481751825e-06,
+      "loss": 0.5411,
+      "step": 1077
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.573722627737227e-06,
+      "loss": 0.5091,
+      "step": 1078
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.572262773722628e-06,
+      "loss": 0.5039,
+      "step": 1079
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.570802919708029e-06,
+      "loss": 0.5825,
+      "step": 1080
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.569343065693431e-06,
+      "loss": 0.5077,
+      "step": 1081
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.567883211678832e-06,
+      "loss": 0.5207,
+      "step": 1082
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.566423357664234e-06,
+      "loss": 0.4913,
+      "step": 1083
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.564963503649635e-06,
+      "loss": 0.4821,
+      "step": 1084
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.563503649635036e-06,
+      "loss": 0.4647,
+      "step": 1085
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.562043795620438e-06,
+      "loss": 0.5312,
+      "step": 1086
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.5605839416058396e-06,
+      "loss": 0.5655,
+      "step": 1087
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.559124087591242e-06,
+      "loss": 0.4291,
+      "step": 1088
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.557664233576643e-06,
+      "loss": 0.4561,
+      "step": 1089
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.556204379562043e-06,
+      "loss": 0.4999,
+      "step": 1090
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.554744525547445e-06,
+      "loss": 0.5463,
+      "step": 1091
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.5532846715328465e-06,
+      "loss": 0.5698,
+      "step": 1092
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.5518248175182485e-06,
+      "loss": 0.5453,
+      "step": 1093
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.55036496350365e-06,
+      "loss": 0.5419,
+      "step": 1094
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.548905109489051e-06,
+      "loss": 0.57,
+      "step": 1095
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.547445255474453e-06,
+      "loss": 0.4904,
+      "step": 1096
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.545985401459854e-06,
+      "loss": 0.4671,
+      "step": 1097
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.5445255474452555e-06,
+      "loss": 0.4755,
+      "step": 1098
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.5430656934306575e-06,
+      "loss": 0.4713,
+      "step": 1099
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.541605839416058e-06,
+      "loss": 0.4544,
+      "step": 1100
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.54014598540146e-06,
+      "loss": 0.5393,
+      "step": 1101
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.538686131386861e-06,
+      "loss": 0.5188,
+      "step": 1102
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.537226277372263e-06,
+      "loss": 0.501,
+      "step": 1103
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.5357664233576644e-06,
+      "loss": 0.5162,
+      "step": 1104
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.534306569343066e-06,
+      "loss": 0.4936,
+      "step": 1105
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.532846715328468e-06,
+      "loss": 0.4947,
+      "step": 1106
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.531386861313869e-06,
+      "loss": 0.4385,
+      "step": 1107
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.52992700729927e-06,
+      "loss": 0.5676,
+      "step": 1108
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.528467153284672e-06,
+      "loss": 0.4793,
+      "step": 1109
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.5270072992700726e-06,
+      "loss": 0.533,
+      "step": 1110
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.525547445255475e-06,
+      "loss": 0.467,
+      "step": 1111
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.524087591240876e-06,
+      "loss": 0.531,
+      "step": 1112
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.522627737226277e-06,
+      "loss": 0.5132,
+      "step": 1113
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.521167883211679e-06,
+      "loss": 0.5271,
+      "step": 1114
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.51970802919708e-06,
+      "loss": 0.4927,
+      "step": 1115
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.518248175182482e-06,
+      "loss": 0.5444,
+      "step": 1116
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.516788321167884e-06,
+      "loss": 0.5158,
+      "step": 1117
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.515328467153285e-06,
+      "loss": 0.5801,
+      "step": 1118
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.513868613138686e-06,
+      "loss": 0.5264,
+      "step": 1119
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.512408759124087e-06,
+      "loss": 0.5088,
+      "step": 1120
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.510948905109489e-06,
+      "loss": 0.4818,
+      "step": 1121
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.5094890510948905e-06,
+      "loss": 0.4416,
+      "step": 1122
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.508029197080292e-06,
+      "loss": 0.5372,
+      "step": 1123
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.506569343065694e-06,
+      "loss": 0.4884,
+      "step": 1124
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.505109489051095e-06,
+      "loss": 0.4341,
+      "step": 1125
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.503649635036496e-06,
+      "loss": 0.517,
+      "step": 1126
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.502189781021898e-06,
+      "loss": 0.5792,
+      "step": 1127
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.5007299270072995e-06,
+      "loss": 0.4908,
+      "step": 1128
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.499270072992701e-06,
+      "loss": 0.4995,
+      "step": 1129
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.497810218978102e-06,
+      "loss": 0.5195,
+      "step": 1130
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.496350364963504e-06,
+      "loss": 0.5273,
+      "step": 1131
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.494890510948905e-06,
+      "loss": 0.4995,
+      "step": 1132
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.4934306569343064e-06,
+      "loss": 0.5269,
+      "step": 1133
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.4919708029197085e-06,
+      "loss": 0.5182,
+      "step": 1134
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.49051094890511e-06,
+      "loss": 0.5846,
+      "step": 1135
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.489051094890511e-06,
+      "loss": 0.5201,
+      "step": 1136
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.487591240875913e-06,
+      "loss": 0.5272,
+      "step": 1137
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.486131386861314e-06,
+      "loss": 0.58,
+      "step": 1138
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.484671532846715e-06,
+      "loss": 0.4768,
+      "step": 1139
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.483211678832117e-06,
+      "loss": 0.5043,
+      "step": 1140
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.481751824817518e-06,
+      "loss": 0.5129,
+      "step": 1141
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.48029197080292e-06,
+      "loss": 0.452,
+      "step": 1142
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.478832116788321e-06,
+      "loss": 0.5472,
+      "step": 1143
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.477372262773723e-06,
+      "loss": 0.4901,
+      "step": 1144
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.475912408759124e-06,
+      "loss": 0.5677,
+      "step": 1145
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.474452554744526e-06,
+      "loss": 0.4631,
+      "step": 1146
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.472992700729928e-06,
+      "loss": 0.5599,
+      "step": 1147
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.471532846715328e-06,
+      "loss": 0.5278,
+      "step": 1148
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.47007299270073e-06,
+      "loss": 0.4774,
+      "step": 1149
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.468613138686131e-06,
+      "loss": 0.6065,
+      "step": 1150
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4671532846715325e-06,
+      "loss": 0.4698,
+      "step": 1151
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4656934306569346e-06,
+      "loss": 0.5429,
+      "step": 1152
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.464233576642336e-06,
+      "loss": 0.5024,
+      "step": 1153
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.462773722627737e-06,
+      "loss": 0.5228,
+      "step": 1154
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.461313868613139e-06,
+      "loss": 0.5259,
+      "step": 1155
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.45985401459854e-06,
+      "loss": 0.4944,
+      "step": 1156
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.458394160583942e-06,
+      "loss": 0.4766,
+      "step": 1157
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.456934306569343e-06,
+      "loss": 0.4696,
+      "step": 1158
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.455474452554745e-06,
+      "loss": 0.4732,
+      "step": 1159
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.454014598540146e-06,
+      "loss": 0.4987,
+      "step": 1160
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.452554744525547e-06,
+      "loss": 0.4902,
+      "step": 1161
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.451094890510949e-06,
+      "loss": 0.5077,
+      "step": 1162
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4496350364963505e-06,
+      "loss": 0.5798,
+      "step": 1163
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.448175182481752e-06,
+      "loss": 0.5086,
+      "step": 1164
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.446715328467154e-06,
+      "loss": 0.5356,
+      "step": 1165
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.445255474452555e-06,
+      "loss": 0.5184,
+      "step": 1166
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.443795620437957e-06,
+      "loss": 0.4697,
+      "step": 1167
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.442335766423357e-06,
+      "loss": 0.5071,
+      "step": 1168
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.440875912408759e-06,
+      "loss": 0.5368,
+      "step": 1169
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.439416058394161e-06,
+      "loss": 0.4632,
+      "step": 1170
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.437956204379562e-06,
+      "loss": 0.4913,
+      "step": 1171
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.436496350364964e-06,
+      "loss": 0.5211,
+      "step": 1172
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.435036496350365e-06,
+      "loss": 0.5438,
+      "step": 1173
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.433576642335766e-06,
+      "loss": 0.5609,
+      "step": 1174
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4321167883211684e-06,
+      "loss": 0.5061,
+      "step": 1175
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.43065693430657e-06,
+      "loss": 0.4765,
+      "step": 1176
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.429197080291971e-06,
+      "loss": 0.5354,
+      "step": 1177
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.427737226277372e-06,
+      "loss": 0.4844,
+      "step": 1178
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.426277372262773e-06,
+      "loss": 0.5488,
+      "step": 1179
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.424817518248175e-06,
+      "loss": 0.4843,
+      "step": 1180
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4233576642335766e-06,
+      "loss": 0.464,
+      "step": 1181
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.421897810218979e-06,
+      "loss": 0.507,
+      "step": 1182
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.42043795620438e-06,
+      "loss": 0.5329,
+      "step": 1183
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.418978102189781e-06,
+      "loss": 0.5659,
+      "step": 1184
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.417518248175183e-06,
+      "loss": 0.4787,
+      "step": 1185
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.416058394160584e-06,
+      "loss": 0.4949,
+      "step": 1186
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4145985401459855e-06,
+      "loss": 0.5524,
+      "step": 1187
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.413138686131387e-06,
+      "loss": 0.5036,
+      "step": 1188
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.411678832116788e-06,
+      "loss": 0.5407,
+      "step": 1189
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.41021897810219e-06,
+      "loss": 0.5235,
+      "step": 1190
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.408759124087591e-06,
+      "loss": 0.4636,
+      "step": 1191
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.4072992700729925e-06,
+      "loss": 0.5074,
+      "step": 1192
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.4058394160583945e-06,
+      "loss": 0.5345,
+      "step": 1193
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.404379562043796e-06,
+      "loss": 0.4749,
+      "step": 1194
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.402919708029198e-06,
+      "loss": 0.4573,
+      "step": 1195
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.401459854014599e-06,
+      "loss": 0.5206,
+      "step": 1196
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.399999999999999e-06,
+      "loss": 0.5411,
+      "step": 1197
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.3985401459854014e-06,
+      "loss": 0.475,
+      "step": 1198
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.397080291970803e-06,
+      "loss": 0.5676,
+      "step": 1199
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.395620437956205e-06,
+      "loss": 0.4533,
+      "step": 1200
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.394160583941606e-06,
+      "loss": 0.4798,
+      "step": 1201
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.392700729927007e-06,
+      "loss": 0.5219,
+      "step": 1202
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.391240875912409e-06,
+      "loss": 0.505,
+      "step": 1203
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.38978102189781e-06,
+      "loss": 0.5103,
+      "step": 1204
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.388321167883212e-06,
+      "loss": 0.6069,
+      "step": 1205
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.386861313868614e-06,
+      "loss": 0.5335,
+      "step": 1206
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.385401459854014e-06,
+      "loss": 0.5257,
+      "step": 1207
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.383941605839416e-06,
+      "loss": 0.4993,
+      "step": 1208
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.382481751824817e-06,
+      "loss": 0.4845,
+      "step": 1209
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.381021897810219e-06,
+      "loss": 0.5058,
+      "step": 1210
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.379562043795621e-06,
+      "loss": 0.5385,
+      "step": 1211
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.378102189781022e-06,
+      "loss": 0.5367,
+      "step": 1212
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.376642335766424e-06,
+      "loss": 0.4729,
+      "step": 1213
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.375182481751825e-06,
+      "loss": 0.5255,
+      "step": 1214
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.373722627737226e-06,
+      "loss": 0.5329,
+      "step": 1215
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.372262773722628e-06,
+      "loss": 0.5641,
+      "step": 1216
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.370802919708029e-06,
+      "loss": 0.4989,
+      "step": 1217
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.369343065693431e-06,
+      "loss": 0.429,
+      "step": 1218
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.367883211678832e-06,
+      "loss": 0.5405,
+      "step": 1219
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.366423357664233e-06,
+      "loss": 0.5387,
+      "step": 1220
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.364963503649635e-06,
+      "loss": 0.4691,
+      "step": 1221
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.3635036496350365e-06,
+      "loss": 0.5115,
+      "step": 1222
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.3620437956204386e-06,
+      "loss": 0.4972,
+      "step": 1223
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.36058394160584e-06,
+      "loss": 0.516,
+      "step": 1224
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.359124087591241e-06,
+      "loss": 0.544,
+      "step": 1225
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.357664233576642e-06,
+      "loss": 0.4952,
+      "step": 1226
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.3562043795620434e-06,
+      "loss": 0.4698,
+      "step": 1227
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.3547445255474455e-06,
+      "loss": 0.5143,
+      "step": 1228
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.353284671532847e-06,
+      "loss": 0.4795,
+      "step": 1229
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.351824817518248e-06,
+      "loss": 0.5503,
+      "step": 1230
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.35036496350365e-06,
+      "loss": 0.5026,
+      "step": 1231
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.348905109489051e-06,
+      "loss": 0.5013,
+      "step": 1232
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.347445255474452e-06,
+      "loss": 0.5427,
+      "step": 1233
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.3459854014598545e-06,
+      "loss": 0.5048,
+      "step": 1234
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.344525547445256e-06,
+      "loss": 0.5332,
+      "step": 1235
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.343065693430657e-06,
+      "loss": 0.5777,
+      "step": 1236
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.341605839416058e-06,
+      "loss": 0.5147,
+      "step": 1237
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.34014598540146e-06,
+      "loss": 0.4939,
+      "step": 1238
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.338686131386861e-06,
+      "loss": 0.4698,
+      "step": 1239
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.337226277372263e-06,
+      "loss": 0.4798,
+      "step": 1240
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.335766423357665e-06,
+      "loss": 0.5203,
+      "step": 1241
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.334306569343066e-06,
+      "loss": 0.5304,
+      "step": 1242
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.332846715328467e-06,
+      "loss": 0.4648,
+      "step": 1243
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.331386861313869e-06,
+      "loss": 0.4963,
+      "step": 1244
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.32992700729927e-06,
+      "loss": 0.4986,
+      "step": 1245
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.328467153284672e-06,
+      "loss": 0.4953,
+      "step": 1246
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.327007299270073e-06,
+      "loss": 0.4911,
+      "step": 1247
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.325547445255474e-06,
+      "loss": 0.4821,
+      "step": 1248
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.324087591240876e-06,
+      "loss": 0.5103,
+      "step": 1249
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.322627737226277e-06,
+      "loss": 0.4819,
+      "step": 1250
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.321167883211679e-06,
+      "loss": 0.4816,
+      "step": 1251
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.3197080291970806e-06,
+      "loss": 0.5005,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.318248175182482e-06,
+      "loss": 0.5384,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.316788321167884e-06,
+      "loss": 0.4843,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.315328467153285e-06,
+      "loss": 0.4706,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.313868613138686e-06,
+      "loss": 0.5,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.3124087591240875e-06,
+      "loss": 0.5332,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.310948905109489e-06,
+      "loss": 0.4639,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.309489051094891e-06,
+      "loss": 0.5419,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.308029197080292e-06,
+      "loss": 0.5047,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.306569343065694e-06,
+      "loss": 0.4874,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.305109489051095e-06,
+      "loss": 0.4717,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.3036496350364965e-06,
+      "loss": 0.4631,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.3021897810218985e-06,
+      "loss": 0.5081,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.300729927007299e-06,
+      "loss": 0.5198,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.299270072992701e-06,
+      "loss": 0.4617,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.5042092204093933,
+      "eval_runtime": 1.1244,
+      "eval_samples_per_second": 17.787,
+      "eval_steps_per_second": 0.889,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6,
+      "eval_bench_accuracy_agieval": 0.3898305084745763,
+      "eval_bench_accuracy_arc_challenge": 0.8148148148148148,
+      "eval_bench_accuracy_arc_easy": 0.8518518518518519,
+      "eval_bench_accuracy_bigbench": 0.45901639344262296,
+      "eval_bench_accuracy_boolq": 0.6851851851851852,
+      "eval_bench_accuracy_mmlu": 0.48148148148148145,
+      "eval_bench_accuracy_openbookqa": 0.2037037037037037,
+      "eval_bench_accuracy_truthful_qa": 0.2641509433962264,
+      "eval_bench_accuracy_winogrande": 0.46296296296296297,
+      "eval_bench_average_accuracy": 0.5125553161459362,
+      "eval_bench_loss": 15.669148272778614,
+      "eval_bench_total_accuracy": 0.5110663983903421,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.297810218978102e-06,
+      "loss": 0.4302,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.296350364963503e-06,
+      "loss": 0.4982,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.2948905109489054e-06,
+      "loss": 0.5132,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.293430656934307e-06,
+      "loss": 0.594,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.291970802919708e-06,
+      "loss": 0.5258,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.29051094890511e-06,
+      "loss": 0.5703,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.289051094890511e-06,
+      "loss": 0.5296,
+      "step": 1273
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.287591240875913e-06,
+      "loss": 0.448,
+      "step": 1274
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.2861313868613136e-06,
+      "loss": 0.4978,
+      "step": 1275
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.284671532846715e-06,
+      "loss": 0.5194,
+      "step": 1276
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.283211678832117e-06,
+      "loss": 0.5738,
+      "step": 1277
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.281751824817518e-06,
+      "loss": 0.4953,
+      "step": 1278
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.28029197080292e-06,
+      "loss": 0.5176,
+      "step": 1279
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.278832116788321e-06,
+      "loss": 0.527,
+      "step": 1280
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.2773722627737225e-06,
+      "loss": 0.5205,
+      "step": 1281
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.275912408759125e-06,
+      "loss": 0.494,
+      "step": 1282
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.274452554744526e-06,
+      "loss": 0.4914,
+      "step": 1283
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.272992700729927e-06,
+      "loss": 0.5026,
+      "step": 1284
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.271532846715328e-06,
+      "loss": 0.5011,
+      "step": 1285
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.2700729927007295e-06,
+      "loss": 0.4959,
+      "step": 1286
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.2686131386861315e-06,
+      "loss": 0.4963,
+      "step": 1287
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.267153284671533e-06,
+      "loss": 0.4997,
+      "step": 1288
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.265693430656935e-06,
+      "loss": 0.589,
+      "step": 1289
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.264233576642336e-06,
+      "loss": 0.4941,
+      "step": 1290
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.262773722627737e-06,
+      "loss": 0.5327,
+      "step": 1291
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.261313868613139e-06,
+      "loss": 0.575,
+      "step": 1292
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.2598540145985405e-06,
+      "loss": 0.5068,
+      "step": 1293
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.258394160583942e-06,
+      "loss": 0.4955,
+      "step": 1294
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.256934306569343e-06,
+      "loss": 0.4405,
+      "step": 1295
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.255474452554744e-06,
+      "loss": 0.53,
+      "step": 1296
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.254014598540146e-06,
+      "loss": 0.5433,
+      "step": 1297
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.252554744525547e-06,
+      "loss": 0.4955,
+      "step": 1298
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.251094890510949e-06,
+      "loss": 0.4943,
+      "step": 1299
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.249635036496351e-06,
+      "loss": 0.475,
+      "step": 1300
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.248175182481752e-06,
+      "loss": 0.5878,
+      "step": 1301
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.246715328467154e-06,
+      "loss": 0.4648,
+      "step": 1302
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.245255474452555e-06,
+      "loss": 0.4863,
+      "step": 1303
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.243795620437956e-06,
+      "loss": 0.4528,
+      "step": 1304
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.242335766423358e-06,
+      "loss": 0.5639,
+      "step": 1305
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.240875912408759e-06,
+      "loss": 0.4301,
+      "step": 1306
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.239416058394161e-06,
+      "loss": 0.5062,
+      "step": 1307
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.237956204379562e-06,
+      "loss": 0.4779,
+      "step": 1308
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.236496350364963e-06,
+      "loss": 0.4761,
+      "step": 1309
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.235036496350365e-06,
+      "loss": 0.5198,
+      "step": 1310
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.233576642335767e-06,
+      "loss": 0.474,
+      "step": 1311
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.232116788321168e-06,
+      "loss": 0.4736,
+      "step": 1312
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.23065693430657e-06,
+      "loss": 0.547,
+      "step": 1313
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.22919708029197e-06,
+      "loss": 0.4826,
+      "step": 1314
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.227737226277372e-06,
+      "loss": 0.4878,
+      "step": 1315
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2262773722627735e-06,
+      "loss": 0.5284,
+      "step": 1316
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2248175182481756e-06,
+      "loss": 0.4658,
+      "step": 1317
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.223357664233577e-06,
+      "loss": 0.5425,
+      "step": 1318
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.221897810218978e-06,
+      "loss": 0.6104,
+      "step": 1319
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.22043795620438e-06,
+      "loss": 0.5149,
+      "step": 1320
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.218978102189781e-06,
+      "loss": 0.5289,
+      "step": 1321
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2175182481751825e-06,
+      "loss": 0.5287,
+      "step": 1322
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2160583941605845e-06,
+      "loss": 0.4913,
+      "step": 1323
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.214598540145985e-06,
+      "loss": 0.5391,
+      "step": 1324
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.213138686131387e-06,
+      "loss": 0.4664,
+      "step": 1325
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.211678832116788e-06,
+      "loss": 0.5182,
+      "step": 1326
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.210218978102189e-06,
+      "loss": 0.5177,
+      "step": 1327
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2087591240875915e-06,
+      "loss": 0.484,
+      "step": 1328
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.207299270072993e-06,
+      "loss": 0.531,
+      "step": 1329
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.205839416058395e-06,
+      "loss": 0.4526,
+      "step": 1330
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.204379562043796e-06,
+      "loss": 0.5185,
+      "step": 1331
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.202919708029197e-06,
+      "loss": 0.4593,
+      "step": 1332
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.201459854014599e-06,
+      "loss": 0.5028,
+      "step": 1333
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.2e-06,
+      "loss": 0.5376,
+      "step": 1334
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.198540145985402e-06,
+      "loss": 0.5391,
+      "step": 1335
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.197080291970803e-06,
+      "loss": 0.5055,
+      "step": 1336
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.195620437956204e-06,
+      "loss": 0.4786,
+      "step": 1337
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.194160583941606e-06,
+      "loss": 0.5135,
+      "step": 1338
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.192700729927007e-06,
+      "loss": 0.5426,
+      "step": 1339
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.1912408759124094e-06,
+      "loss": 0.4942,
+      "step": 1340
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.189781021897811e-06,
+      "loss": 0.5743,
+      "step": 1341
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.188321167883212e-06,
+      "loss": 0.5116,
+      "step": 1342
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.186861313868613e-06,
+      "loss": 0.4822,
+      "step": 1343
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.185401459854014e-06,
+      "loss": 0.4949,
+      "step": 1344
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.183941605839416e-06,
+      "loss": 0.4873,
+      "step": 1345
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.1824817518248176e-06,
+      "loss": 0.5252,
+      "step": 1346
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.181021897810219e-06,
+      "loss": 0.5302,
+      "step": 1347
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.179562043795621e-06,
+      "loss": 0.5185,
+      "step": 1348
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.178102189781022e-06,
+      "loss": 0.5157,
+      "step": 1349
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.176642335766423e-06,
+      "loss": 0.506,
+      "step": 1350
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.175182481751825e-06,
+      "loss": 0.5323,
+      "step": 1351
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.1737226277372265e-06,
+      "loss": 0.5149,
+      "step": 1352
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.172262773722628e-06,
+      "loss": 0.4719,
+      "step": 1353
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.170802919708029e-06,
+      "loss": 0.465,
+      "step": 1354
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.16934306569343e-06,
+      "loss": 0.4613,
+      "step": 1355
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.167883211678832e-06,
+      "loss": 0.5971,
+      "step": 1356
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.1664233576642335e-06,
+      "loss": 0.5114,
+      "step": 1357
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.1649635036496355e-06,
+      "loss": 0.5011,
+      "step": 1358
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.163503649635037e-06,
+      "loss": 0.5496,
+      "step": 1359
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.162043795620438e-06,
+      "loss": 0.4673,
+      "step": 1360
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.16058394160584e-06,
+      "loss": 0.5004,
+      "step": 1361
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.159124087591241e-06,
+      "loss": 0.4933,
+      "step": 1362
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.1576642335766424e-06,
+      "loss": 0.4564,
+      "step": 1363
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.156204379562044e-06,
+      "loss": 0.5167,
+      "step": 1364
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.154744525547445e-06,
+      "loss": 0.4929,
+      "step": 1365
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.153284671532847e-06,
+      "loss": 0.4794,
+      "step": 1366
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.151824817518248e-06,
+      "loss": 0.5117,
+      "step": 1367
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.15036496350365e-06,
+      "loss": 0.488,
+      "step": 1368
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.148905109489051e-06,
+      "loss": 0.5032,
+      "step": 1369
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.147445255474453e-06,
+      "loss": 0.4629,
+      "step": 1370
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.145985401459855e-06,
+      "loss": 0.4621,
+      "step": 1371
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.144525547445255e-06,
+      "loss": 0.5253,
+      "step": 1372
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.143065693430657e-06,
+      "loss": 0.5203,
+      "step": 1373
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.141605839416058e-06,
+      "loss": 0.5107,
+      "step": 1374
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.1401459854014595e-06,
+      "loss": 0.529,
+      "step": 1375
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.138686131386862e-06,
+      "loss": 0.4795,
+      "step": 1376
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.137226277372263e-06,
+      "loss": 0.4748,
+      "step": 1377
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.135766423357664e-06,
+      "loss": 0.4929,
+      "step": 1378
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.134306569343066e-06,
+      "loss": 0.5266,
+      "step": 1379
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.132846715328467e-06,
+      "loss": 0.5556,
+      "step": 1380
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.131386861313869e-06,
+      "loss": 0.4817,
+      "step": 1381
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.12992700729927e-06,
+      "loss": 0.5115,
+      "step": 1382
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.128467153284672e-06,
+      "loss": 0.4865,
+      "step": 1383
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.127007299270073e-06,
+      "loss": 0.5178,
+      "step": 1384
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.125547445255474e-06,
+      "loss": 0.5361,
+      "step": 1385
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.124087591240876e-06,
+      "loss": 0.4251,
+      "step": 1386
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.1226277372262775e-06,
+      "loss": 0.5026,
+      "step": 1387
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.121167883211679e-06,
+      "loss": 0.5721,
+      "step": 1388
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.119708029197081e-06,
+      "loss": 0.4937,
+      "step": 1389
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.118248175182482e-06,
+      "loss": 0.5187,
+      "step": 1390
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.116788321167883e-06,
+      "loss": 0.4705,
+      "step": 1391
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.115328467153284e-06,
+      "loss": 0.512,
+      "step": 1392
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.113868613138686e-06,
+      "loss": 0.5939,
+      "step": 1393
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.112408759124088e-06,
+      "loss": 0.5048,
+      "step": 1394
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.110948905109489e-06,
+      "loss": 0.5053,
+      "step": 1395
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.109489051094891e-06,
+      "loss": 0.5258,
+      "step": 1396
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.108029197080292e-06,
+      "loss": 0.5706,
+      "step": 1397
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.106569343065693e-06,
+      "loss": 0.5246,
+      "step": 1398
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.1051094890510955e-06,
+      "loss": 0.4682,
+      "step": 1399
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.103649635036497e-06,
+      "loss": 0.5086,
+      "step": 1400
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.102189781021898e-06,
+      "loss": 0.5118,
+      "step": 1401
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.100729927007299e-06,
+      "loss": 0.4609,
+      "step": 1402
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.0992700729927e-06,
+      "loss": 0.508,
+      "step": 1403
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.097810218978102e-06,
+      "loss": 0.5583,
+      "step": 1404
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.096350364963504e-06,
+      "loss": 0.4866,
+      "step": 1405
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.094890510948905e-06,
+      "loss": 0.518,
+      "step": 1406
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.093430656934307e-06,
+      "loss": 0.5698,
+      "step": 1407
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.091970802919708e-06,
+      "loss": 0.5045,
+      "step": 1408
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.09051094890511e-06,
+      "loss": 0.5268,
+      "step": 1409
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.089051094890511e-06,
+      "loss": 0.5064,
+      "step": 1410
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.0875912408759126e-06,
+      "loss": 0.5607,
+      "step": 1411
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.086131386861314e-06,
+      "loss": 0.5166,
+      "step": 1412
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.084671532846715e-06,
+      "loss": 0.5341,
+      "step": 1413
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.083211678832117e-06,
+      "loss": 0.4936,
+      "step": 1414
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.081751824817518e-06,
+      "loss": 0.5219,
+      "step": 1415
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.0802919708029195e-06,
+      "loss": 0.4553,
+      "step": 1416
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.0788321167883215e-06,
+      "loss": 0.4311,
+      "step": 1417
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.077372262773723e-06,
+      "loss": 0.5453,
+      "step": 1418
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.075912408759125e-06,
+      "loss": 0.5214,
+      "step": 1419
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.074452554744526e-06,
+      "loss": 0.5184,
+      "step": 1420
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.072992700729926e-06,
+      "loss": 0.4699,
+      "step": 1421
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0715328467153285e-06,
+      "loss": 0.5582,
+      "step": 1422
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.07007299270073e-06,
+      "loss": 0.5839,
+      "step": 1423
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.068613138686132e-06,
+      "loss": 0.4663,
+      "step": 1424
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.067153284671533e-06,
+      "loss": 0.5403,
+      "step": 1425
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.065693430656934e-06,
+      "loss": 0.5641,
+      "step": 1426
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.064233576642336e-06,
+      "loss": 0.5365,
+      "step": 1427
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0627737226277374e-06,
+      "loss": 0.5091,
+      "step": 1428
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.061313868613139e-06,
+      "loss": 0.4881,
+      "step": 1429
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.059854014598541e-06,
+      "loss": 0.4917,
+      "step": 1430
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.058394160583941e-06,
+      "loss": 0.5195,
+      "step": 1431
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.056934306569343e-06,
+      "loss": 0.4748,
+      "step": 1432
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.055474452554744e-06,
+      "loss": 0.5241,
+      "step": 1433
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.054014598540146e-06,
+      "loss": 0.516,
+      "step": 1434
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.052554744525548e-06,
+      "loss": 0.4759,
+      "step": 1435
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.051094890510949e-06,
+      "loss": 0.4792,
+      "step": 1436
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.049635036496351e-06,
+      "loss": 0.4934,
+      "step": 1437
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.048175182481752e-06,
+      "loss": 0.5418,
+      "step": 1438
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.046715328467153e-06,
+      "loss": 0.4849,
+      "step": 1439
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.045255474452555e-06,
+      "loss": 0.5503,
+      "step": 1440
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.043795620437956e-06,
+      "loss": 0.5142,
+      "step": 1441
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.042335766423358e-06,
+      "loss": 0.4356,
+      "step": 1442
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.040875912408759e-06,
+      "loss": 0.4788,
+      "step": 1443
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.03941605839416e-06,
+      "loss": 0.4697,
+      "step": 1444
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.037956204379562e-06,
+      "loss": 0.513,
+      "step": 1445
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0364963503649635e-06,
+      "loss": 0.496,
+      "step": 1446
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.035036496350366e-06,
+      "loss": 0.4884,
+      "step": 1447
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.033576642335767e-06,
+      "loss": 0.5199,
+      "step": 1448
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.032116788321168e-06,
+      "loss": 0.4886,
+      "step": 1449
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.030656934306569e-06,
+      "loss": 0.491,
+      "step": 1450
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0291970802919705e-06,
+      "loss": 0.5051,
+      "step": 1451
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0277372262773725e-06,
+      "loss": 0.4844,
+      "step": 1452
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.026277372262774e-06,
+      "loss": 0.4767,
+      "step": 1453
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.024817518248175e-06,
+      "loss": 0.5122,
+      "step": 1454
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.023357664233577e-06,
+      "loss": 0.5457,
+      "step": 1455
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.021897810218978e-06,
+      "loss": 0.5342,
+      "step": 1456
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0204379562043794e-06,
+      "loss": 0.5231,
+      "step": 1457
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0189781021897815e-06,
+      "loss": 0.551,
+      "step": 1458
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.017518248175183e-06,
+      "loss": 0.4734,
+      "step": 1459
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.016058394160584e-06,
+      "loss": 0.4697,
+      "step": 1460
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.014598540145985e-06,
+      "loss": 0.5278,
+      "step": 1461
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.013138686131387e-06,
+      "loss": 0.5622,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.011678832116788e-06,
+      "loss": 0.4523,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.01021897810219e-06,
+      "loss": 0.4917,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.008759124087592e-06,
+      "loss": 0.5243,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.007299270072993e-06,
+      "loss": 0.5379,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.005839416058394e-06,
+      "loss": 0.5854,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.004379562043796e-06,
+      "loss": 0.5805,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.002919708029197e-06,
+      "loss": 0.4772,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.001459854014599e-06,
+      "loss": 0.5359,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4e-06,
+      "loss": 0.5394,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.998540145985401e-06,
+      "loss": 0.5533,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.997080291970803e-06,
+      "loss": 0.4614,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.995620437956204e-06,
+      "loss": 0.4919,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.994160583941606e-06,
+      "loss": 0.4779,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.992700729927008e-06,
+      "loss": 0.4703,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.991240875912409e-06,
+      "loss": 0.456,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.49469125270843506,
+      "eval_runtime": 1.123,
+      "eval_samples_per_second": 17.809,
+      "eval_steps_per_second": 0.89,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7,
+      "eval_bench_accuracy_agieval": 0.3220338983050847,
+      "eval_bench_accuracy_arc_challenge": 0.8148148148148148,
+      "eval_bench_accuracy_arc_easy": 0.8518518518518519,
+      "eval_bench_accuracy_bigbench": 0.45901639344262296,
+      "eval_bench_accuracy_boolq": 0.7222222222222222,
+      "eval_bench_accuracy_mmlu": 0.48148148148148145,
+      "eval_bench_accuracy_openbookqa": 0.2222222222222222,
+      "eval_bench_accuracy_truthful_qa": 0.3018867924528302,
+      "eval_bench_accuracy_winogrande": 0.5,
+      "eval_bench_average_accuracy": 0.519503297421459,
+      "eval_bench_loss": 15.187939747270331,
+      "eval_bench_total_accuracy": 0.5171026156941649,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.989781021897811e-06,
+      "loss": 0.4871,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.988321167883212e-06,
+      "loss": 0.4911,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.986861313868613e-06,
+      "loss": 0.466,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.9854014598540145e-06,
+      "loss": 0.5064,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.983941605839416e-06,
+      "loss": 0.5657,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.982481751824818e-06,
+      "loss": 0.4878,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.981021897810219e-06,
+      "loss": 0.5058,
+      "step": 1484
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.97956204379562e-06,
+      "loss": 0.4874,
+      "step": 1485
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.978102189781022e-06,
+      "loss": 0.4903,
+      "step": 1486
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9766423357664235e-06,
+      "loss": 0.5081,
+      "step": 1487
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9751824817518255e-06,
+      "loss": 0.4505,
+      "step": 1488
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.973722627737226e-06,
+      "loss": 0.5583,
+      "step": 1489
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.972262773722628e-06,
+      "loss": 0.4853,
+      "step": 1490
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.970802919708029e-06,
+      "loss": 0.4762,
+      "step": 1491
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.96934306569343e-06,
+      "loss": 0.5747,
+      "step": 1492
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9678832116788325e-06,
+      "loss": 0.5001,
+      "step": 1493
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.966423357664234e-06,
+      "loss": 0.4888,
+      "step": 1494
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.964963503649635e-06,
+      "loss": 0.5002,
+      "step": 1495
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.963503649635037e-06,
+      "loss": 0.5483,
+      "step": 1496
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.962043795620438e-06,
+      "loss": 0.5222,
+      "step": 1497
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.96058394160584e-06,
+      "loss": 0.4868,
+      "step": 1498
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.959124087591241e-06,
+      "loss": 0.5403,
+      "step": 1499
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.957664233576642e-06,
+      "loss": 0.5036,
+      "step": 1500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.956204379562044e-06,
+      "loss": 0.5574,
+      "step": 1501
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.954744525547445e-06,
+      "loss": 0.5584,
+      "step": 1502
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.953284671532847e-06,
+      "loss": 0.5107,
+      "step": 1503
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.951824817518248e-06,
+      "loss": 0.4826,
+      "step": 1504
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9503649635036496e-06,
+      "loss": 0.5403,
+      "step": 1505
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.948905109489052e-06,
+      "loss": 0.4915,
+      "step": 1506
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.947445255474453e-06,
+      "loss": 0.4791,
+      "step": 1507
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.945985401459854e-06,
+      "loss": 0.5526,
+      "step": 1508
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.944525547445255e-06,
+      "loss": 0.4899,
+      "step": 1509
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9430656934306565e-06,
+      "loss": 0.4693,
+      "step": 1510
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9416058394160585e-06,
+      "loss": 0.4142,
+      "step": 1511
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.94014598540146e-06,
+      "loss": 0.5199,
+      "step": 1512
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.938686131386861e-06,
+      "loss": 0.5303,
+      "step": 1513
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.937226277372263e-06,
+      "loss": 0.4536,
+      "step": 1514
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.935766423357664e-06,
+      "loss": 0.5276,
+      "step": 1515
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.934306569343066e-06,
+      "loss": 0.4862,
+      "step": 1516
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9328467153284675e-06,
+      "loss": 0.4656,
+      "step": 1517
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.931386861313869e-06,
+      "loss": 0.4855,
+      "step": 1518
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.92992700729927e-06,
+      "loss": 0.4765,
+      "step": 1519
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.928467153284671e-06,
+      "loss": 0.5213,
+      "step": 1520
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.927007299270073e-06,
+      "loss": 0.4933,
+      "step": 1521
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9255474452554744e-06,
+      "loss": 0.5066,
+      "step": 1522
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.924087591240876e-06,
+      "loss": 0.4783,
+      "step": 1523
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.922627737226278e-06,
+      "loss": 0.4979,
+      "step": 1524
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.921167883211679e-06,
+      "loss": 0.4714,
+      "step": 1525
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.919708029197081e-06,
+      "loss": 0.4456,
+      "step": 1526
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.918248175182482e-06,
+      "loss": 0.4884,
+      "step": 1527
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.916788321167883e-06,
+      "loss": 0.4726,
+      "step": 1528
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.915328467153285e-06,
+      "loss": 0.4854,
+      "step": 1529
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.913868613138686e-06,
+      "loss": 0.4726,
+      "step": 1530
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.912408759124088e-06,
+      "loss": 0.472,
+      "step": 1531
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.910948905109489e-06,
+      "loss": 0.5532,
+      "step": 1532
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.90948905109489e-06,
+      "loss": 0.4756,
+      "step": 1533
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.908029197080292e-06,
+      "loss": 0.5138,
+      "step": 1534
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.906569343065694e-06,
+      "loss": 0.5369,
+      "step": 1535
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.905109489051095e-06,
+      "loss": 0.5237,
+      "step": 1536
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.903649635036497e-06,
+      "loss": 0.5115,
+      "step": 1537
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.902189781021897e-06,
+      "loss": 0.5341,
+      "step": 1538
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.900729927007299e-06,
+      "loss": 0.4896,
+      "step": 1539
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8992700729927005e-06,
+      "loss": 0.4945,
+      "step": 1540
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.897810218978103e-06,
+      "loss": 0.5346,
+      "step": 1541
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.896350364963504e-06,
+      "loss": 0.4881,
+      "step": 1542
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.894890510948905e-06,
+      "loss": 0.5293,
+      "step": 1543
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.893430656934307e-06,
+      "loss": 0.5168,
+      "step": 1544
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.891970802919708e-06,
+      "loss": 0.5054,
+      "step": 1545
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8905109489051095e-06,
+      "loss": 0.5305,
+      "step": 1546
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.889051094890512e-06,
+      "loss": 0.5409,
+      "step": 1547
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.887591240875912e-06,
+      "loss": 0.4737,
+      "step": 1548
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.886131386861314e-06,
+      "loss": 0.5242,
+      "step": 1549
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.884671532846715e-06,
+      "loss": 0.4883,
+      "step": 1550
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.8832116788321164e-06,
+      "loss": 0.5205,
+      "step": 1551
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.8817518248175185e-06,
+      "loss": 0.5296,
+      "step": 1552
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.88029197080292e-06,
+      "loss": 0.5037,
+      "step": 1553
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.878832116788322e-06,
+      "loss": 0.545,
+      "step": 1554
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.877372262773723e-06,
+      "loss": 0.4909,
+      "step": 1555
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.875912408759124e-06,
+      "loss": 0.4911,
+      "step": 1556
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.874452554744526e-06,
+      "loss": 0.5587,
+      "step": 1557
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.872992700729927e-06,
+      "loss": 0.4861,
+      "step": 1558
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.871532846715329e-06,
+      "loss": 0.4906,
+      "step": 1559
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.87007299270073e-06,
+      "loss": 0.4773,
+      "step": 1560
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.868613138686131e-06,
+      "loss": 0.5042,
+      "step": 1561
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.867153284671533e-06,
+      "loss": 0.5011,
+      "step": 1562
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.865693430656934e-06,
+      "loss": 0.4492,
+      "step": 1563
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.864233576642336e-06,
+      "loss": 0.4992,
+      "step": 1564
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.862773722627738e-06,
+      "loss": 0.5513,
+      "step": 1565
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.861313868613139e-06,
+      "loss": 0.4851,
+      "step": 1566
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.85985401459854e-06,
+      "loss": 0.5221,
+      "step": 1567
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.858394160583941e-06,
+      "loss": 0.4156,
+      "step": 1568
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.856934306569343e-06,
+      "loss": 0.5132,
+      "step": 1569
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.855474452554745e-06,
+      "loss": 0.4704,
+      "step": 1570
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.854014598540146e-06,
+      "loss": 0.4884,
+      "step": 1571
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.852554744525548e-06,
+      "loss": 0.4961,
+      "step": 1572
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.851094890510949e-06,
+      "loss": 0.4722,
+      "step": 1573
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.84963503649635e-06,
+      "loss": 0.4846,
+      "step": 1574
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.848175182481752e-06,
+      "loss": 0.5042,
+      "step": 1575
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8467153284671536e-06,
+      "loss": 0.5277,
+      "step": 1576
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.845255474452555e-06,
+      "loss": 0.5015,
+      "step": 1577
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.843795620437956e-06,
+      "loss": 0.5335,
+      "step": 1578
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.842335766423357e-06,
+      "loss": 0.4627,
+      "step": 1579
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.840875912408759e-06,
+      "loss": 0.5612,
+      "step": 1580
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8394160583941605e-06,
+      "loss": 0.5262,
+      "step": 1581
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8379562043795625e-06,
+      "loss": 0.5276,
+      "step": 1582
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.836496350364964e-06,
+      "loss": 0.4445,
+      "step": 1583
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.835036496350365e-06,
+      "loss": 0.473,
+      "step": 1584
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.833576642335767e-06,
+      "loss": 0.499,
+      "step": 1585
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.832116788321168e-06,
+      "loss": 0.5018,
+      "step": 1586
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8306569343065695e-06,
+      "loss": 0.476,
+      "step": 1587
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.829197080291971e-06,
+      "loss": 0.4797,
+      "step": 1588
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.827737226277372e-06,
+      "loss": 0.5536,
+      "step": 1589
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.826277372262774e-06,
+      "loss": 0.5392,
+      "step": 1590
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.824817518248175e-06,
+      "loss": 0.4611,
+      "step": 1591
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.823357664233576e-06,
+      "loss": 0.4882,
+      "step": 1592
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8218978102189784e-06,
+      "loss": 0.516,
+      "step": 1593
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.82043795620438e-06,
+      "loss": 0.5273,
+      "step": 1594
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.818978102189782e-06,
+      "loss": 0.5093,
+      "step": 1595
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.817518248175182e-06,
+      "loss": 0.4608,
+      "step": 1596
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.816058394160584e-06,
+      "loss": 0.5154,
+      "step": 1597
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8145985401459854e-06,
+      "loss": 0.4659,
+      "step": 1598
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.813138686131387e-06,
+      "loss": 0.4872,
+      "step": 1599
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.811678832116788e-06,
+      "loss": 0.4831,
+      "step": 1600
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.81021897810219e-06,
+      "loss": 0.5343,
+      "step": 1601
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8087591240875915e-06,
+      "loss": 0.5476,
+      "step": 1602
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.807299270072993e-06,
+      "loss": 0.465,
+      "step": 1603
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8058394160583943e-06,
+      "loss": 0.5225,
+      "step": 1604
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.804379562043796e-06,
+      "loss": 0.5862,
+      "step": 1605
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8029197080291968e-06,
+      "loss": 0.5365,
+      "step": 1606
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8014598540145984e-06,
+      "loss": 0.5259,
+      "step": 1607
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8e-06,
+      "loss": 0.5152,
+      "step": 1608
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7985401459854013e-06,
+      "loss": 0.4839,
+      "step": 1609
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.797080291970803e-06,
+      "loss": 0.482,
+      "step": 1610
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7956204379562045e-06,
+      "loss": 0.514,
+      "step": 1611
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.794160583941606e-06,
+      "loss": 0.4487,
+      "step": 1612
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7927007299270074e-06,
+      "loss": 0.4651,
+      "step": 1613
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.791240875912409e-06,
+      "loss": 0.4838,
+      "step": 1614
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7897810218978107e-06,
+      "loss": 0.529,
+      "step": 1615
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7883211678832114e-06,
+      "loss": 0.4483,
+      "step": 1616
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.786861313868613e-06,
+      "loss": 0.5693,
+      "step": 1617
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7854014598540147e-06,
+      "loss": 0.45,
+      "step": 1618
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.783941605839416e-06,
+      "loss": 0.4916,
+      "step": 1619
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7824817518248176e-06,
+      "loss": 0.4826,
+      "step": 1620
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7810218978102192e-06,
+      "loss": 0.5214,
+      "step": 1621
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.779562043795621e-06,
+      "loss": 0.5101,
+      "step": 1622
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.778102189781022e-06,
+      "loss": 0.5635,
+      "step": 1623
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7766423357664237e-06,
+      "loss": 0.5011,
+      "step": 1624
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7751824817518253e-06,
+      "loss": 0.5062,
+      "step": 1625
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.773722627737226e-06,
+      "loss": 0.4941,
+      "step": 1626
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7722627737226278e-06,
+      "loss": 0.5158,
+      "step": 1627
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.770802919708029e-06,
+      "loss": 0.5108,
+      "step": 1628
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7693430656934306e-06,
+      "loss": 0.4835,
+      "step": 1629
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7678832116788323e-06,
+      "loss": 0.4273,
+      "step": 1630
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.766423357664234e-06,
+      "loss": 0.4992,
+      "step": 1631
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.764963503649635e-06,
+      "loss": 0.5051,
+      "step": 1632
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7635036496350367e-06,
+      "loss": 0.4654,
+      "step": 1633
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7620437956204384e-06,
+      "loss": 0.5057,
+      "step": 1634
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.760583941605839e-06,
+      "loss": 0.4967,
+      "step": 1635
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.759124087591241e-06,
+      "loss": 0.5224,
+      "step": 1636
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7576642335766425e-06,
+      "loss": 0.595,
+      "step": 1637
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7562043795620437e-06,
+      "loss": 0.5007,
+      "step": 1638
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7547445255474453e-06,
+      "loss": 0.5127,
+      "step": 1639
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.753284671532847e-06,
+      "loss": 0.5196,
+      "step": 1640
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.751824817518248e-06,
+      "loss": 0.4748,
+      "step": 1641
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.75036496350365e-06,
+      "loss": 0.5366,
+      "step": 1642
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7489051094890514e-06,
+      "loss": 0.5007,
+      "step": 1643
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.747445255474453e-06,
+      "loss": 0.517,
+      "step": 1644
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.745985401459854e-06,
+      "loss": 0.5149,
+      "step": 1645
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7445255474452555e-06,
+      "loss": 0.4709,
+      "step": 1646
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7430656934306567e-06,
+      "loss": 0.5056,
+      "step": 1647
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7416058394160583e-06,
+      "loss": 0.4588,
+      "step": 1648
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.74014598540146e-06,
+      "loss": 0.4303,
+      "step": 1649
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7386861313868616e-06,
+      "loss": 0.4556,
+      "step": 1650
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.737226277372263e-06,
+      "loss": 0.522,
+      "step": 1651
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7357664233576645e-06,
+      "loss": 0.5321,
+      "step": 1652
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.734306569343066e-06,
+      "loss": 0.4843,
+      "step": 1653
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7328467153284677e-06,
+      "loss": 0.527,
+      "step": 1654
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7313868613138685e-06,
+      "loss": 0.4751,
+      "step": 1655
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7299270072992698e-06,
+      "loss": 0.5323,
+      "step": 1656
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7284671532846714e-06,
+      "loss": 0.5197,
+      "step": 1657
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.727007299270073e-06,
+      "loss": 0.5176,
+      "step": 1658
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7255474452554747e-06,
+      "loss": 0.4693,
+      "step": 1659
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.724087591240876e-06,
+      "loss": 0.5314,
+      "step": 1660
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7226277372262775e-06,
+      "loss": 0.5065,
+      "step": 1661
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.721167883211679e-06,
+      "loss": 0.5259,
+      "step": 1662
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.719708029197081e-06,
+      "loss": 0.5249,
+      "step": 1663
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.718248175182482e-06,
+      "loss": 0.5319,
+      "step": 1664
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7167883211678832e-06,
+      "loss": 0.5199,
+      "step": 1665
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7153284671532844e-06,
+      "loss": 0.5123,
+      "step": 1666
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.713868613138686e-06,
+      "loss": 0.4969,
+      "step": 1667
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7124087591240877e-06,
+      "loss": 0.4685,
+      "step": 1668
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7109489051094893e-06,
+      "loss": 0.4801,
+      "step": 1669
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7094890510948906e-06,
+      "loss": 0.5478,
+      "step": 1670
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.708029197080292e-06,
+      "loss": 0.4843,
+      "step": 1671
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.706569343065694e-06,
+      "loss": 0.4645,
+      "step": 1672
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.705109489051095e-06,
+      "loss": 0.5418,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.7036496350364963e-06,
+      "loss": 0.5073,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.7021897810218975e-06,
+      "loss": 0.4914,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.700729927007299e-06,
+      "loss": 0.4938,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6992700729927008e-06,
+      "loss": 0.526,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6978102189781024e-06,
+      "loss": 0.4789,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6963503649635036e-06,
+      "loss": 0.524,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6948905109489052e-06,
+      "loss": 0.476,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.693430656934307e-06,
+      "loss": 0.5061,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6919708029197085e-06,
+      "loss": 0.4795,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6905109489051097e-06,
+      "loss": 0.5297,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6890510948905105e-06,
+      "loss": 0.4765,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.687591240875912e-06,
+      "loss": 0.4699,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.686131386861314e-06,
+      "loss": 0.4511,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6846715328467154e-06,
+      "loss": 0.5005,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6832116788321167e-06,
+      "loss": 0.5494,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.48713812232017517,
+      "eval_runtime": 1.1205,
+      "eval_samples_per_second": 17.849,
+      "eval_steps_per_second": 0.892,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8,
+      "eval_bench_accuracy_agieval": 0.3898305084745763,
+      "eval_bench_accuracy_arc_challenge": 0.8703703703703703,
+      "eval_bench_accuracy_arc_easy": 0.8888888888888888,
+      "eval_bench_accuracy_bigbench": 0.5901639344262295,
+      "eval_bench_accuracy_boolq": 0.7407407407407407,
+      "eval_bench_accuracy_mmlu": 0.46296296296296297,
+      "eval_bench_accuracy_openbookqa": 0.2222222222222222,
+      "eval_bench_accuracy_truthful_qa": 0.24528301886792453,
+      "eval_bench_accuracy_winogrande": 0.5,
+      "eval_bench_average_accuracy": 0.5456069607726572,
+      "eval_bench_loss": 15.527333454913403,
+      "eval_bench_total_accuracy": 0.545271629778672,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6817518248175183e-06,
+      "loss": 0.5148,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.68029197080292e-06,
+      "loss": 0.4513,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6788321167883216e-06,
+      "loss": 0.4576,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6773722627737228e-06,
+      "loss": 0.4735,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6759124087591244e-06,
+      "loss": 0.5156,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.674452554744525e-06,
+      "loss": 0.5529,
+      "step": 1694
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.672992700729927e-06,
+      "loss": 0.5027,
+      "step": 1695
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6715328467153285e-06,
+      "loss": 0.4883,
+      "step": 1696
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.67007299270073e-06,
+      "loss": 0.4681,
+      "step": 1697
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6686131386861313e-06,
+      "loss": 0.4852,
+      "step": 1698
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.667153284671533e-06,
+      "loss": 0.5149,
+      "step": 1699
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6656934306569346e-06,
+      "loss": 0.5176,
+      "step": 1700
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6642335766423362e-06,
+      "loss": 0.5341,
+      "step": 1701
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6627737226277375e-06,
+      "loss": 0.4691,
+      "step": 1702
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.661313868613139e-06,
+      "loss": 0.4745,
+      "step": 1703
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.65985401459854e-06,
+      "loss": 0.5195,
+      "step": 1704
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6583941605839415e-06,
+      "loss": 0.5186,
+      "step": 1705
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.656934306569343e-06,
+      "loss": 0.4721,
+      "step": 1706
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6554744525547444e-06,
+      "loss": 0.5151,
+      "step": 1707
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.654014598540146e-06,
+      "loss": 0.4604,
+      "step": 1708
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6525547445255477e-06,
+      "loss": 0.451,
+      "step": 1709
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6510948905109493e-06,
+      "loss": 0.5283,
+      "step": 1710
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6496350364963505e-06,
+      "loss": 0.51,
+      "step": 1711
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.648175182481752e-06,
+      "loss": 0.5555,
+      "step": 1712
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.646715328467153e-06,
+      "loss": 0.5826,
+      "step": 1713
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6452554744525546e-06,
+      "loss": 0.4323,
+      "step": 1714
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6437956204379562e-06,
+      "loss": 0.4803,
+      "step": 1715
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6423357664233574e-06,
+      "loss": 0.5336,
+      "step": 1716
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.640875912408759e-06,
+      "loss": 0.4961,
+      "step": 1717
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6394160583941607e-06,
+      "loss": 0.537,
+      "step": 1718
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6379562043795623e-06,
+      "loss": 0.5406,
+      "step": 1719
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6364963503649636e-06,
+      "loss": 0.4533,
+      "step": 1720
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.635036496350365e-06,
+      "loss": 0.4952,
+      "step": 1721
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.633576642335767e-06,
+      "loss": 0.4922,
+      "step": 1722
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6321167883211676e-06,
+      "loss": 0.5325,
+      "step": 1723
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6306569343065693e-06,
+      "loss": 0.5016,
+      "step": 1724
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.629197080291971e-06,
+      "loss": 0.4718,
+      "step": 1725
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.627737226277372e-06,
+      "loss": 0.4655,
+      "step": 1726
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6262773722627737e-06,
+      "loss": 0.5043,
+      "step": 1727
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6248175182481754e-06,
+      "loss": 0.4922,
+      "step": 1728
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.623357664233577e-06,
+      "loss": 0.4842,
+      "step": 1729
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6218978102189782e-06,
+      "loss": 0.5125,
+      "step": 1730
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.62043795620438e-06,
+      "loss": 0.5313,
+      "step": 1731
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6189781021897815e-06,
+      "loss": 0.5082,
+      "step": 1732
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6175182481751823e-06,
+      "loss": 0.4605,
+      "step": 1733
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.616058394160584e-06,
+      "loss": 0.4513,
+      "step": 1734
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.614598540145985e-06,
+      "loss": 0.4991,
+      "step": 1735
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.613138686131387e-06,
+      "loss": 0.5108,
+      "step": 1736
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6116788321167884e-06,
+      "loss": 0.4861,
+      "step": 1737
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.61021897810219e-06,
+      "loss": 0.5114,
+      "step": 1738
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6087591240875913e-06,
+      "loss": 0.519,
+      "step": 1739
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.607299270072993e-06,
+      "loss": 0.4652,
+      "step": 1740
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6058394160583946e-06,
+      "loss": 0.4692,
+      "step": 1741
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.604379562043796e-06,
+      "loss": 0.4816,
+      "step": 1742
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.602919708029197e-06,
+      "loss": 0.5023,
+      "step": 1743
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6014598540145986e-06,
+      "loss": 0.5045,
+      "step": 1744
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6e-06,
+      "loss": 0.5549,
+      "step": 1745
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5985401459854015e-06,
+      "loss": 0.534,
+      "step": 1746
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.597080291970803e-06,
+      "loss": 0.4738,
+      "step": 1747
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5956204379562048e-06,
+      "loss": 0.5163,
+      "step": 1748
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.594160583941606e-06,
+      "loss": 0.4574,
+      "step": 1749
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5927007299270076e-06,
+      "loss": 0.5163,
+      "step": 1750
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5912408759124092e-06,
+      "loss": 0.4699,
+      "step": 1751
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.58978102189781e-06,
+      "loss": 0.4727,
+      "step": 1752
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5883211678832117e-06,
+      "loss": 0.4833,
+      "step": 1753
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.586861313868613e-06,
+      "loss": 0.5337,
+      "step": 1754
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.5854014598540145e-06,
+      "loss": 0.4611,
+      "step": 1755
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.583941605839416e-06,
+      "loss": 0.4869,
+      "step": 1756
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.582481751824818e-06,
+      "loss": 0.5108,
+      "step": 1757
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.581021897810219e-06,
+      "loss": 0.461,
+      "step": 1758
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5795620437956206e-06,
+      "loss": 0.4721,
+      "step": 1759
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5781021897810223e-06,
+      "loss": 0.4838,
+      "step": 1760
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.576642335766424e-06,
+      "loss": 0.4653,
+      "step": 1761
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5751824817518247e-06,
+      "loss": 0.4616,
+      "step": 1762
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.573722627737226e-06,
+      "loss": 0.4597,
+      "step": 1763
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5722627737226276e-06,
+      "loss": 0.4648,
+      "step": 1764
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.570802919708029e-06,
+      "loss": 0.4966,
+      "step": 1765
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.569343065693431e-06,
+      "loss": 0.4497,
+      "step": 1766
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.567883211678832e-06,
+      "loss": 0.5154,
+      "step": 1767
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5664233576642337e-06,
+      "loss": 0.4647,
+      "step": 1768
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5649635036496353e-06,
+      "loss": 0.4897,
+      "step": 1769
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.563503649635037e-06,
+      "loss": 0.5815,
+      "step": 1770
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.562043795620438e-06,
+      "loss": 0.4909,
+      "step": 1771
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5605839416058394e-06,
+      "loss": 0.5032,
+      "step": 1772
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5591240875912406e-06,
+      "loss": 0.4719,
+      "step": 1773
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5576642335766422e-06,
+      "loss": 0.4828,
+      "step": 1774
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.556204379562044e-06,
+      "loss": 0.4827,
+      "step": 1775
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5547445255474455e-06,
+      "loss": 0.4979,
+      "step": 1776
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5532846715328467e-06,
+      "loss": 0.5206,
+      "step": 1777
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.5518248175182484e-06,
+      "loss": 0.5303,
+      "step": 1778
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.55036496350365e-06,
+      "loss": 0.4847,
+      "step": 1779
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5489051094890516e-06,
+      "loss": 0.5023,
+      "step": 1780
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.547445255474453e-06,
+      "loss": 0.5104,
+      "step": 1781
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5459854014598537e-06,
+      "loss": 0.5309,
+      "step": 1782
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5445255474452553e-06,
+      "loss": 0.4944,
+      "step": 1783
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.543065693430657e-06,
+      "loss": 0.486,
+      "step": 1784
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5416058394160586e-06,
+      "loss": 0.4843,
+      "step": 1785
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5401459854014598e-06,
+      "loss": 0.5213,
+      "step": 1786
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5386861313868614e-06,
+      "loss": 0.466,
+      "step": 1787
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.537226277372263e-06,
+      "loss": 0.5613,
+      "step": 1788
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5357664233576647e-06,
+      "loss": 0.613,
+      "step": 1789
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.534306569343066e-06,
+      "loss": 0.5229,
+      "step": 1790
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.532846715328467e-06,
+      "loss": 0.5264,
+      "step": 1791
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5313868613138683e-06,
+      "loss": 0.5009,
+      "step": 1792
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.52992700729927e-06,
+      "loss": 0.5006,
+      "step": 1793
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5284671532846716e-06,
+      "loss": 0.4933,
+      "step": 1794
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.527007299270073e-06,
+      "loss": 0.4554,
+      "step": 1795
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5255474452554745e-06,
+      "loss": 0.4625,
+      "step": 1796
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.524087591240876e-06,
+      "loss": 0.5469,
+      "step": 1797
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.5226277372262777e-06,
+      "loss": 0.4993,
+      "step": 1798
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.521167883211679e-06,
+      "loss": 0.5151,
+      "step": 1799
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5197080291970806e-06,
+      "loss": 0.5042,
+      "step": 1800
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5182481751824814e-06,
+      "loss": 0.4486,
+      "step": 1801
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.516788321167883e-06,
+      "loss": 0.4985,
+      "step": 1802
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5153284671532847e-06,
+      "loss": 0.5057,
+      "step": 1803
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5138686131386863e-06,
+      "loss": 0.5118,
+      "step": 1804
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5124087591240875e-06,
+      "loss": 0.4536,
+      "step": 1805
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.510948905109489e-06,
+      "loss": 0.5212,
+      "step": 1806
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5094890510948908e-06,
+      "loss": 0.4654,
+      "step": 1807
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5080291970802924e-06,
+      "loss": 0.5178,
+      "step": 1808
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5065693430656936e-06,
+      "loss": 0.4938,
+      "step": 1809
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5051094890510953e-06,
+      "loss": 0.5168,
+      "step": 1810
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.503649635036496e-06,
+      "loss": 0.5287,
+      "step": 1811
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5021897810218977e-06,
+      "loss": 0.5342,
+      "step": 1812
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5007299270072993e-06,
+      "loss": 0.4644,
+      "step": 1813
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.4992700729927006e-06,
+      "loss": 0.4684,
+      "step": 1814
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.497810218978102e-06,
+      "loss": 0.4635,
+      "step": 1815
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.496350364963504e-06,
+      "loss": 0.4851,
+      "step": 1816
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.4948905109489055e-06,
+      "loss": 0.4709,
+      "step": 1817
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.4934306569343067e-06,
+      "loss": 0.4775,
+      "step": 1818
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.4919708029197083e-06,
+      "loss": 0.532,
+      "step": 1819
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.490510948905109e-06,
+      "loss": 0.5008,
+      "step": 1820
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4890510948905107e-06,
+      "loss": 0.4989,
+      "step": 1821
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4875912408759124e-06,
+      "loss": 0.4915,
+      "step": 1822
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.486131386861314e-06,
+      "loss": 0.5271,
+      "step": 1823
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4846715328467152e-06,
+      "loss": 0.521,
+      "step": 1824
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.483211678832117e-06,
+      "loss": 0.517,
+      "step": 1825
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4817518248175185e-06,
+      "loss": 0.4809,
+      "step": 1826
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.48029197080292e-06,
+      "loss": 0.4567,
+      "step": 1827
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4788321167883214e-06,
+      "loss": 0.5614,
+      "step": 1828
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.477372262773723e-06,
+      "loss": 0.5491,
+      "step": 1829
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.475912408759124e-06,
+      "loss": 0.4591,
+      "step": 1830
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4744525547445254e-06,
+      "loss": 0.5318,
+      "step": 1831
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.472992700729927e-06,
+      "loss": 0.4958,
+      "step": 1832
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4715328467153283e-06,
+      "loss": 0.5161,
+      "step": 1833
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.47007299270073e-06,
+      "loss": 0.5104,
+      "step": 1834
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4686131386861316e-06,
+      "loss": 0.4562,
+      "step": 1835
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.467153284671533e-06,
+      "loss": 0.4492,
+      "step": 1836
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4656934306569344e-06,
+      "loss": 0.4745,
+      "step": 1837
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.464233576642336e-06,
+      "loss": 0.4868,
+      "step": 1838
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4627737226277377e-06,
+      "loss": 0.4762,
+      "step": 1839
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.4613138686131385e-06,
+      "loss": 0.5049,
+      "step": 1840
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.45985401459854e-06,
+      "loss": 0.4637,
+      "step": 1841
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4583941605839413e-06,
+      "loss": 0.5793,
+      "step": 1842
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.456934306569343e-06,
+      "loss": 0.4565,
+      "step": 1843
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4554744525547446e-06,
+      "loss": 0.4794,
+      "step": 1844
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4540145985401462e-06,
+      "loss": 0.5475,
+      "step": 1845
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4525547445255475e-06,
+      "loss": 0.496,
+      "step": 1846
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.451094890510949e-06,
+      "loss": 0.4623,
+      "step": 1847
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4496350364963507e-06,
+      "loss": 0.4768,
+      "step": 1848
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4481751824817524e-06,
+      "loss": 0.4837,
+      "step": 1849
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.446715328467153e-06,
+      "loss": 0.5268,
+      "step": 1850
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.445255474452555e-06,
+      "loss": 0.4992,
+      "step": 1851
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.443795620437956e-06,
+      "loss": 0.476,
+      "step": 1852
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4423357664233576e-06,
+      "loss": 0.5212,
+      "step": 1853
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4408759124087593e-06,
+      "loss": 0.5409,
+      "step": 1854
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.439416058394161e-06,
+      "loss": 0.467,
+      "step": 1855
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.437956204379562e-06,
+      "loss": 0.4522,
+      "step": 1856
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4364963503649638e-06,
+      "loss": 0.5311,
+      "step": 1857
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4350364963503654e-06,
+      "loss": 0.4541,
+      "step": 1858
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.433576642335766e-06,
+      "loss": 0.4387,
+      "step": 1859
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.432116788321168e-06,
+      "loss": 0.5048,
+      "step": 1860
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.430656934306569e-06,
+      "loss": 0.4731,
+      "step": 1861
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4291970802919707e-06,
+      "loss": 0.4483,
+      "step": 1862
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4277372262773723e-06,
+      "loss": 0.4948,
+      "step": 1863
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.426277372262774e-06,
+      "loss": 0.4745,
+      "step": 1864
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.424817518248175e-06,
+      "loss": 0.48,
+      "step": 1865
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.423357664233577e-06,
+      "loss": 0.5105,
+      "step": 1866
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4218978102189785e-06,
+      "loss": 0.4882,
+      "step": 1867
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.42043795620438e-06,
+      "loss": 0.4211,
+      "step": 1868
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.418978102189781e-06,
+      "loss": 0.5123,
+      "step": 1869
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4175182481751825e-06,
+      "loss": 0.5041,
+      "step": 1870
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4160583941605837e-06,
+      "loss": 0.5204,
+      "step": 1871
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4145985401459854e-06,
+      "loss": 0.4954,
+      "step": 1872
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.413138686131387e-06,
+      "loss": 0.5261,
+      "step": 1873
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4116788321167882e-06,
+      "loss": 0.5065,
+      "step": 1874
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.41021897810219e-06,
+      "loss": 0.5083,
+      "step": 1875
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4087591240875915e-06,
+      "loss": 0.4714,
+      "step": 1876
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.407299270072993e-06,
+      "loss": 0.4903,
+      "step": 1877
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4058394160583944e-06,
+      "loss": 0.55,
+      "step": 1878
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4043795620437956e-06,
+      "loss": 0.4923,
+      "step": 1879
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4029197080291968e-06,
+      "loss": 0.4408,
+      "step": 1880
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4014598540145984e-06,
+      "loss": 0.443,
+      "step": 1881
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.4e-06,
+      "loss": 0.5354,
+      "step": 1882
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.3985401459854017e-06,
+      "loss": 0.4345,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.397080291970803e-06,
+      "loss": 0.5143,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3956204379562045e-06,
+      "loss": 0.4482,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.394160583941606e-06,
+      "loss": 0.5217,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.392700729927008e-06,
+      "loss": 0.5242,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.391240875912409e-06,
+      "loss": 0.5907,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.38978102189781e-06,
+      "loss": 0.4992,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3883211678832115e-06,
+      "loss": 0.4686,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.386861313868613e-06,
+      "loss": 0.5128,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3854014598540147e-06,
+      "loss": 0.4473,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.383941605839416e-06,
+      "loss": 0.4961,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3824817518248176e-06,
+      "loss": 0.4668,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3810218978102192e-06,
+      "loss": 0.4657,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.379562043795621e-06,
+      "loss": 0.4853,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.378102189781022e-06,
+      "loss": 0.5023,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3766423357664233e-06,
+      "loss": 0.4829,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3751824817518245e-06,
+      "loss": 0.5161,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.4823929965496063,
+      "eval_runtime": 1.1215,
+      "eval_samples_per_second": 17.833,
+      "eval_steps_per_second": 0.892,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9,
+      "eval_bench_accuracy_agieval": 0.3898305084745763,
+      "eval_bench_accuracy_arc_challenge": 0.8148148148148148,
+      "eval_bench_accuracy_arc_easy": 0.8888888888888888,
+      "eval_bench_accuracy_bigbench": 0.47540983606557374,
+      "eval_bench_accuracy_boolq": 0.7777777777777778,
+      "eval_bench_accuracy_mmlu": 0.46296296296296297,
+      "eval_bench_accuracy_openbookqa": 0.14814814814814814,
+      "eval_bench_accuracy_truthful_qa": 0.16981132075471697,
+      "eval_bench_accuracy_winogrande": 0.5185185185185185,
+      "eval_bench_average_accuracy": 0.5162403084895532,
+      "eval_bench_loss": 15.998585160956326,
+      "eval_bench_total_accuracy": 0.5150905432595574,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.373722627737226e-06,
+      "loss": 0.4724,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3722627737226278e-06,
+      "loss": 0.4625,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3708029197080294e-06,
+      "loss": 0.4646,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3693430656934306e-06,
+      "loss": 0.4576,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.3678832116788323e-06,
+      "loss": 0.5625,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.366423357664234e-06,
+      "loss": 0.4976,
+      "step": 1905
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3649635036496356e-06,
+      "loss": 0.5292,
+      "step": 1906
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3635036496350368e-06,
+      "loss": 0.4701,
+      "step": 1907
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3620437956204376e-06,
+      "loss": 0.5394,
+      "step": 1908
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.360583941605839e-06,
+      "loss": 0.4781,
+      "step": 1909
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.359124087591241e-06,
+      "loss": 0.5066,
+      "step": 1910
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3576642335766425e-06,
+      "loss": 0.514,
+      "step": 1911
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3562043795620437e-06,
+      "loss": 0.4827,
+      "step": 1912
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3547445255474453e-06,
+      "loss": 0.4622,
+      "step": 1913
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.353284671532847e-06,
+      "loss": 0.5499,
+      "step": 1914
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3518248175182486e-06,
+      "loss": 0.5524,
+      "step": 1915
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.35036496350365e-06,
+      "loss": 0.4582,
+      "step": 1916
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3489051094890514e-06,
+      "loss": 0.5076,
+      "step": 1917
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3474452554744522e-06,
+      "loss": 0.4992,
+      "step": 1918
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.345985401459854e-06,
+      "loss": 0.5205,
+      "step": 1919
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3445255474452555e-06,
+      "loss": 0.4403,
+      "step": 1920
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3430656934306567e-06,
+      "loss": 0.4658,
+      "step": 1921
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3416058394160584e-06,
+      "loss": 0.5101,
+      "step": 1922
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.34014598540146e-06,
+      "loss": 0.5262,
+      "step": 1923
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3386861313868616e-06,
+      "loss": 0.4674,
+      "step": 1924
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.337226277372263e-06,
+      "loss": 0.4847,
+      "step": 1925
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.3357664233576645e-06,
+      "loss": 0.44,
+      "step": 1926
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.334306569343066e-06,
+      "loss": 0.4487,
+      "step": 1927
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.332846715328467e-06,
+      "loss": 0.5427,
+      "step": 1928
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3313868613138686e-06,
+      "loss": 0.4788,
+      "step": 1929
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.32992700729927e-06,
+      "loss": 0.4951,
+      "step": 1930
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3284671532846714e-06,
+      "loss": 0.5254,
+      "step": 1931
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.327007299270073e-06,
+      "loss": 0.5025,
+      "step": 1932
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3255474452554747e-06,
+      "loss": 0.4838,
+      "step": 1933
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3240875912408763e-06,
+      "loss": 0.5122,
+      "step": 1934
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3226277372262775e-06,
+      "loss": 0.5787,
+      "step": 1935
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.321167883211679e-06,
+      "loss": 0.5083,
+      "step": 1936
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.31970802919708e-06,
+      "loss": 0.5092,
+      "step": 1937
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3182481751824816e-06,
+      "loss": 0.4752,
+      "step": 1938
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3167883211678832e-06,
+      "loss": 0.4874,
+      "step": 1939
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3153284671532845e-06,
+      "loss": 0.4338,
+      "step": 1940
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.313868613138686e-06,
+      "loss": 0.4973,
+      "step": 1941
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3124087591240877e-06,
+      "loss": 0.461,
+      "step": 1942
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3109489051094894e-06,
+      "loss": 0.5161,
+      "step": 1943
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3094890510948906e-06,
+      "loss": 0.4969,
+      "step": 1944
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3080291970802922e-06,
+      "loss": 0.4574,
+      "step": 1945
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.306569343065694e-06,
+      "loss": 0.4754,
+      "step": 1946
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3051094890510946e-06,
+      "loss": 0.4889,
+      "step": 1947
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3036496350364963e-06,
+      "loss": 0.5301,
+      "step": 1948
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.302189781021898e-06,
+      "loss": 0.5092,
+      "step": 1949
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.300729927007299e-06,
+      "loss": 0.4781,
+      "step": 1950
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2992700729927008e-06,
+      "loss": 0.5388,
+      "step": 1951
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2978102189781024e-06,
+      "loss": 0.4793,
+      "step": 1952
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2963503649635036e-06,
+      "loss": 0.5515,
+      "step": 1953
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2948905109489053e-06,
+      "loss": 0.5432,
+      "step": 1954
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.293430656934307e-06,
+      "loss": 0.4931,
+      "step": 1955
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2919708029197085e-06,
+      "loss": 0.5232,
+      "step": 1956
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2905109489051093e-06,
+      "loss": 0.4889,
+      "step": 1957
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.289051094890511e-06,
+      "loss": 0.4654,
+      "step": 1958
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.287591240875912e-06,
+      "loss": 0.502,
+      "step": 1959
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.286131386861314e-06,
+      "loss": 0.5226,
+      "step": 1960
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2846715328467155e-06,
+      "loss": 0.4683,
+      "step": 1961
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.283211678832117e-06,
+      "loss": 0.4992,
+      "step": 1962
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2817518248175183e-06,
+      "loss": 0.4463,
+      "step": 1963
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.28029197080292e-06,
+      "loss": 0.5192,
+      "step": 1964
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2788321167883216e-06,
+      "loss": 0.4254,
+      "step": 1965
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2773722627737232e-06,
+      "loss": 0.4817,
+      "step": 1966
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.275912408759124e-06,
+      "loss": 0.5023,
+      "step": 1967
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.2744525547445252e-06,
+      "loss": 0.4719,
+      "step": 1968
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.272992700729927e-06,
+      "loss": 0.4423,
+      "step": 1969
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2715328467153285e-06,
+      "loss": 0.496,
+      "step": 1970
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.27007299270073e-06,
+      "loss": 0.5288,
+      "step": 1971
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2686131386861314e-06,
+      "loss": 0.473,
+      "step": 1972
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.267153284671533e-06,
+      "loss": 0.514,
+      "step": 1973
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2656934306569346e-06,
+      "loss": 0.4777,
+      "step": 1974
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2642335766423363e-06,
+      "loss": 0.5383,
+      "step": 1975
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.262773722627737e-06,
+      "loss": 0.4412,
+      "step": 1976
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2613138686131387e-06,
+      "loss": 0.4579,
+      "step": 1977
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.25985401459854e-06,
+      "loss": 0.4891,
+      "step": 1978
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2583941605839415e-06,
+      "loss": 0.4393,
+      "step": 1979
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.256934306569343e-06,
+      "loss": 0.4594,
+      "step": 1980
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.255474452554745e-06,
+      "loss": 0.4633,
+      "step": 1981
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.254014598540146e-06,
+      "loss": 0.5002,
+      "step": 1982
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2525547445255477e-06,
+      "loss": 0.4562,
+      "step": 1983
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2510948905109493e-06,
+      "loss": 0.5092,
+      "step": 1984
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.249635036496351e-06,
+      "loss": 0.4862,
+      "step": 1985
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2481751824817517e-06,
+      "loss": 0.5156,
+      "step": 1986
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.246715328467153e-06,
+      "loss": 0.4774,
+      "step": 1987
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2452554744525546e-06,
+      "loss": 0.4738,
+      "step": 1988
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2437956204379562e-06,
+      "loss": 0.4746,
+      "step": 1989
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.242335766423358e-06,
+      "loss": 0.5155,
+      "step": 1990
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.240875912408759e-06,
+      "loss": 0.5276,
+      "step": 1991
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2394160583941607e-06,
+      "loss": 0.5282,
+      "step": 1992
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2379562043795624e-06,
+      "loss": 0.5272,
+      "step": 1993
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.236496350364964e-06,
+      "loss": 0.4982,
+      "step": 1994
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.235036496350365e-06,
+      "loss": 0.4932,
+      "step": 1995
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2335766423357664e-06,
+      "loss": 0.5117,
+      "step": 1996
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2321167883211676e-06,
+      "loss": 0.4666,
+      "step": 1997
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2306569343065693e-06,
+      "loss": 0.5377,
+      "step": 1998
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.229197080291971e-06,
+      "loss": 0.4778,
+      "step": 1999
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.227737226277372e-06,
+      "loss": 0.5051,
+      "step": 2000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2262773722627738e-06,
+      "loss": 0.5126,
+      "step": 2001
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2248175182481754e-06,
+      "loss": 0.4786,
+      "step": 2002
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.223357664233577e-06,
+      "loss": 0.5764,
+      "step": 2003
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2218978102189783e-06,
+      "loss": 0.4557,
+      "step": 2004
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.22043795620438e-06,
+      "loss": 0.4748,
+      "step": 2005
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2189781021897807e-06,
+      "loss": 0.5636,
+      "step": 2006
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2175182481751823e-06,
+      "loss": 0.4919,
+      "step": 2007
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.216058394160584e-06,
+      "loss": 0.4995,
+      "step": 2008
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.2145985401459856e-06,
+      "loss": 0.4528,
+      "step": 2009
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.213138686131387e-06,
+      "loss": 0.4732,
+      "step": 2010
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2116788321167884e-06,
+      "loss": 0.4594,
+      "step": 2011
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.21021897810219e-06,
+      "loss": 0.492,
+      "step": 2012
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2087591240875917e-06,
+      "loss": 0.4872,
+      "step": 2013
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.207299270072993e-06,
+      "loss": 0.4346,
+      "step": 2014
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2058394160583937e-06,
+      "loss": 0.4983,
+      "step": 2015
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2043795620437954e-06,
+      "loss": 0.4828,
+      "step": 2016
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.202919708029197e-06,
+      "loss": 0.5006,
+      "step": 2017
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2014598540145986e-06,
+      "loss": 0.4946,
+      "step": 2018
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2e-06,
+      "loss": 0.5394,
+      "step": 2019
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1985401459854015e-06,
+      "loss": 0.5003,
+      "step": 2020
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.197080291970803e-06,
+      "loss": 0.4772,
+      "step": 2021
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1956204379562048e-06,
+      "loss": 0.5383,
+      "step": 2022
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.194160583941606e-06,
+      "loss": 0.4752,
+      "step": 2023
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1927007299270076e-06,
+      "loss": 0.4659,
+      "step": 2024
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1912408759124084e-06,
+      "loss": 0.5155,
+      "step": 2025
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.18978102189781e-06,
+      "loss": 0.4519,
+      "step": 2026
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1883211678832117e-06,
+      "loss": 0.5505,
+      "step": 2027
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1868613138686133e-06,
+      "loss": 0.4971,
+      "step": 2028
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.1854014598540145e-06,
+      "loss": 0.485,
+      "step": 2029
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.183941605839416e-06,
+      "loss": 0.4781,
+      "step": 2030
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.182481751824818e-06,
+      "loss": 0.5126,
+      "step": 2031
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.181021897810219e-06,
+      "loss": 0.4775,
+      "step": 2032
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1795620437956207e-06,
+      "loss": 0.4967,
+      "step": 2033
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1781021897810223e-06,
+      "loss": 0.5238,
+      "step": 2034
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.176642335766423e-06,
+      "loss": 0.454,
+      "step": 2035
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1751824817518247e-06,
+      "loss": 0.5146,
+      "step": 2036
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1737226277372264e-06,
+      "loss": 0.4697,
+      "step": 2037
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1722627737226276e-06,
+      "loss": 0.4916,
+      "step": 2038
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1708029197080292e-06,
+      "loss": 0.4895,
+      "step": 2039
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.169343065693431e-06,
+      "loss": 0.5296,
+      "step": 2040
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1678832116788325e-06,
+      "loss": 0.4806,
+      "step": 2041
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1664233576642337e-06,
+      "loss": 0.4871,
+      "step": 2042
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1649635036496353e-06,
+      "loss": 0.4492,
+      "step": 2043
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.163503649635036e-06,
+      "loss": 0.4847,
+      "step": 2044
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1620437956204378e-06,
+      "loss": 0.4729,
+      "step": 2045
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1605839416058394e-06,
+      "loss": 0.5355,
+      "step": 2046
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1591240875912406e-06,
+      "loss": 0.4397,
+      "step": 2047
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1576642335766423e-06,
+      "loss": 0.4786,
+      "step": 2048
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.156204379562044e-06,
+      "loss": 0.485,
+      "step": 2049
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1547445255474455e-06,
+      "loss": 0.5042,
+      "step": 2050
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1532846715328468e-06,
+      "loss": 0.5587,
+      "step": 2051
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.1518248175182484e-06,
+      "loss": 0.4627,
+      "step": 2052
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.15036496350365e-06,
+      "loss": 0.4953,
+      "step": 2053
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.148905109489051e-06,
+      "loss": 0.4698,
+      "step": 2054
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1474452554744525e-06,
+      "loss": 0.5012,
+      "step": 2055
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.145985401459854e-06,
+      "loss": 0.4339,
+      "step": 2056
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1445255474452553e-06,
+      "loss": 0.4505,
+      "step": 2057
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.143065693430657e-06,
+      "loss": 0.5238,
+      "step": 2058
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1416058394160586e-06,
+      "loss": 0.5416,
+      "step": 2059
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1401459854014602e-06,
+      "loss": 0.4276,
+      "step": 2060
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1386861313868614e-06,
+      "loss": 0.4749,
+      "step": 2061
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.137226277372263e-06,
+      "loss": 0.475,
+      "step": 2062
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1357664233576647e-06,
+      "loss": 0.5316,
+      "step": 2063
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1343065693430655e-06,
+      "loss": 0.5377,
+      "step": 2064
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.132846715328467e-06,
+      "loss": 0.4944,
+      "step": 2065
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1313868613138684e-06,
+      "loss": 0.5152,
+      "step": 2066
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.12992700729927e-06,
+      "loss": 0.4756,
+      "step": 2067
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1284671532846716e-06,
+      "loss": 0.4924,
+      "step": 2068
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1270072992700733e-06,
+      "loss": 0.4925,
+      "step": 2069
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1255474452554745e-06,
+      "loss": 0.483,
+      "step": 2070
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.124087591240876e-06,
+      "loss": 0.5247,
+      "step": 2071
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1226277372262778e-06,
+      "loss": 0.4871,
+      "step": 2072
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.1211678832116794e-06,
+      "loss": 0.4402,
+      "step": 2073
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.11970802919708e-06,
+      "loss": 0.4838,
+      "step": 2074
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.118248175182482e-06,
+      "loss": 0.4442,
+      "step": 2075
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.116788321167883e-06,
+      "loss": 0.521,
+      "step": 2076
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1153284671532847e-06,
+      "loss": 0.4722,
+      "step": 2077
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1138686131386863e-06,
+      "loss": 0.5284,
+      "step": 2078
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1124087591240875e-06,
+      "loss": 0.4396,
+      "step": 2079
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.110948905109489e-06,
+      "loss": 0.5173,
+      "step": 2080
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.109489051094891e-06,
+      "loss": 0.4737,
+      "step": 2081
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1080291970802924e-06,
+      "loss": 0.4788,
+      "step": 2082
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1065693430656932e-06,
+      "loss": 0.4817,
+      "step": 2083
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.105109489051095e-06,
+      "loss": 0.5141,
+      "step": 2084
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.103649635036496e-06,
+      "loss": 0.5107,
+      "step": 2085
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1021897810218977e-06,
+      "loss": 0.5323,
+      "step": 2086
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1007299270072994e-06,
+      "loss": 0.4557,
+      "step": 2087
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.099270072992701e-06,
+      "loss": 0.4834,
+      "step": 2088
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.097810218978102e-06,
+      "loss": 0.5173,
+      "step": 2089
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.096350364963504e-06,
+      "loss": 0.4883,
+      "step": 2090
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0948905109489055e-06,
+      "loss": 0.4653,
+      "step": 2091
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.093430656934307e-06,
+      "loss": 0.4879,
+      "step": 2092
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.091970802919708e-06,
+      "loss": 0.4828,
+      "step": 2093
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.090510948905109e-06,
+      "loss": 0.5189,
+      "step": 2094
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0890510948905108e-06,
+      "loss": 0.5139,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0875912408759124e-06,
+      "loss": 0.5135,
+      "step": 2096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.086131386861314e-06,
+      "loss": 0.4851,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0846715328467153e-06,
+      "loss": 0.4499,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.083211678832117e-06,
+      "loss": 0.4946,
+      "step": 2099
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0817518248175185e-06,
+      "loss": 0.494,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.08029197080292e-06,
+      "loss": 0.523,
+      "step": 2101
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0788321167883214e-06,
+      "loss": 0.487,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0773722627737226e-06,
+      "loss": 0.4849,
+      "step": 2103
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.075912408759124e-06,
+      "loss": 0.4575,
+      "step": 2104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0744525547445254e-06,
+      "loss": 0.4709,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.072992700729927e-06,
+      "loss": 0.455,
+      "step": 2106
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0715328467153287e-06,
+      "loss": 0.4299,
+      "step": 2107
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.07007299270073e-06,
+      "loss": 0.4476,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0686131386861316e-06,
+      "loss": 0.4053,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0671532846715332e-06,
+      "loss": 0.4186,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.47802498936653137,
+      "eval_runtime": 1.1271,
+      "eval_samples_per_second": 17.744,
+      "eval_steps_per_second": 0.887,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0,
+      "eval_bench_accuracy_agieval": 0.3389830508474576,
+      "eval_bench_accuracy_arc_challenge": 0.8148148148148148,
+      "eval_bench_accuracy_arc_easy": 0.8703703703703703,
+      "eval_bench_accuracy_bigbench": 0.47540983606557374,
+      "eval_bench_accuracy_boolq": 0.7777777777777778,
+      "eval_bench_accuracy_mmlu": 0.5,
+      "eval_bench_accuracy_openbookqa": 0.24074074074074073,
+      "eval_bench_accuracy_truthful_qa": 0.3018867924528302,
+      "eval_bench_accuracy_winogrande": 0.5185185185185185,
+      "eval_bench_average_accuracy": 0.5376113223986759,
+      "eval_bench_loss": 16.656211761106928,
+      "eval_bench_total_accuracy": 0.5352112676056338,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0656934306569344e-06,
+      "loss": 0.4094,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.064233576642336e-06,
+      "loss": 0.4105,
+      "step": 2112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.062773722627737e-06,
+      "loss": 0.4497,
+      "step": 2113
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.0613138686131385e-06,
+      "loss": 0.453,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.05985401459854e-06,
+      "loss": 0.4626,
+      "step": 2115
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0583941605839418e-06,
+      "loss": 0.4278,
+      "step": 2116
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056934306569343e-06,
+      "loss": 0.4613,
+      "step": 2117
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0554744525547446e-06,
+      "loss": 0.4591,
+      "step": 2118
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0540145985401463e-06,
+      "loss": 0.4659,
+      "step": 2119
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052554744525548e-06,
+      "loss": 0.4349,
+      "step": 2120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051094890510949e-06,
+      "loss": 0.3768,
+      "step": 2121
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0496350364963503e-06,
+      "loss": 0.4932,
+      "step": 2122
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0481751824817515e-06,
+      "loss": 0.4414,
+      "step": 2123
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.046715328467153e-06,
+      "loss": 0.3943,
+      "step": 2124
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.045255474452555e-06,
+      "loss": 0.4432,
+      "step": 2125
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.043795620437956e-06,
+      "loss": 0.436,
+      "step": 2126
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0423357664233577e-06,
+      "loss": 0.4287,
+      "step": 2127
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0408759124087593e-06,
+      "loss": 0.4497,
+      "step": 2128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.039416058394161e-06,
+      "loss": 0.473,
+      "step": 2129
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.037956204379562e-06,
+      "loss": 0.3498,
+      "step": 2130
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.036496350364964e-06,
+      "loss": 0.4149,
+      "step": 2131
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0350364963503646e-06,
+      "loss": 0.4605,
+      "step": 2132
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0335766423357662e-06,
+      "loss": 0.4281,
+      "step": 2133
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.032116788321168e-06,
+      "loss": 0.3942,
+      "step": 2134
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0306569343065695e-06,
+      "loss": 0.4242,
+      "step": 2135
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0291970802919707e-06,
+      "loss": 0.4124,
+      "step": 2136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0277372262773723e-06,
+      "loss": 0.4146,
+      "step": 2137
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.026277372262774e-06,
+      "loss": 0.4065,
+      "step": 2138
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0248175182481756e-06,
+      "loss": 0.404,
+      "step": 2139
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.023357664233577e-06,
+      "loss": 0.4297,
+      "step": 2140
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0218978102189785e-06,
+      "loss": 0.4375,
+      "step": 2141
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0204379562043793e-06,
+      "loss": 0.4438,
+      "step": 2142
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.018978102189781e-06,
+      "loss": 0.4766,
+      "step": 2143
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0175182481751825e-06,
+      "loss": 0.4779,
+      "step": 2144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0160583941605838e-06,
+      "loss": 0.3961,
+      "step": 2145
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0145985401459854e-06,
+      "loss": 0.4444,
+      "step": 2146
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.013138686131387e-06,
+      "loss": 0.4674,
+      "step": 2147
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0116788321167887e-06,
+      "loss": 0.4476,
+      "step": 2148
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01021897810219e-06,
+      "loss": 0.4207,
+      "step": 2149
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0087591240875915e-06,
+      "loss": 0.4192,
+      "step": 2150
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.007299270072993e-06,
+      "loss": 0.3977,
+      "step": 2151
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.005839416058394e-06,
+      "loss": 0.4564,
+      "step": 2152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0043795620437956e-06,
+      "loss": 0.4449,
+      "step": 2153
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0029197080291972e-06,
+      "loss": 0.4489,
+      "step": 2154
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0014598540145984e-06,
+      "loss": 0.4325,
+      "step": 2155
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3e-06,
+      "loss": 0.4272,
+      "step": 2156
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9985401459854017e-06,
+      "loss": 0.403,
+      "step": 2157
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.997080291970803e-06,
+      "loss": 0.4411,
+      "step": 2158
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.995620437956204e-06,
+      "loss": 0.4423,
+      "step": 2159
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9941605839416058e-06,
+      "loss": 0.4179,
+      "step": 2160
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9927007299270074e-06,
+      "loss": 0.4488,
+      "step": 2161
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.991240875912409e-06,
+      "loss": 0.4246,
+      "step": 2162
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9897810218978103e-06,
+      "loss": 0.383,
+      "step": 2163
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9883211678832115e-06,
+      "loss": 0.4332,
+      "step": 2164
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.986861313868613e-06,
+      "loss": 0.4229,
+      "step": 2165
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9854014598540148e-06,
+      "loss": 0.4032,
+      "step": 2166
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9839416058394164e-06,
+      "loss": 0.4115,
+      "step": 2167
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9824817518248176e-06,
+      "loss": 0.3903,
+      "step": 2168
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.981021897810219e-06,
+      "loss": 0.4462,
+      "step": 2169
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9795620437956205e-06,
+      "loss": 0.4873,
+      "step": 2170
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.978102189781022e-06,
+      "loss": 0.4216,
+      "step": 2171
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9766423357664237e-06,
+      "loss": 0.3986,
+      "step": 2172
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.975182481751825e-06,
+      "loss": 0.4311,
+      "step": 2173
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.973722627737226e-06,
+      "loss": 0.4254,
+      "step": 2174
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.972262773722628e-06,
+      "loss": 0.3896,
+      "step": 2175
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9708029197080294e-06,
+      "loss": 0.4151,
+      "step": 2176
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9693430656934307e-06,
+      "loss": 0.4554,
+      "step": 2177
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.967883211678832e-06,
+      "loss": 0.3612,
+      "step": 2178
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9664233576642335e-06,
+      "loss": 0.4206,
+      "step": 2179
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.964963503649635e-06,
+      "loss": 0.4098,
+      "step": 2180
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9635036496350368e-06,
+      "loss": 0.4232,
+      "step": 2181
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.962043795620438e-06,
+      "loss": 0.4364,
+      "step": 2182
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.960583941605839e-06,
+      "loss": 0.418,
+      "step": 2183
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.959124087591241e-06,
+      "loss": 0.4041,
+      "step": 2184
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9576642335766425e-06,
+      "loss": 0.4189,
+      "step": 2185
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.956204379562044e-06,
+      "loss": 0.4942,
+      "step": 2186
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9547445255474453e-06,
+      "loss": 0.4467,
+      "step": 2187
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9532846715328466e-06,
+      "loss": 0.4254,
+      "step": 2188
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.951824817518248e-06,
+      "loss": 0.3987,
+      "step": 2189
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.95036496350365e-06,
+      "loss": 0.443,
+      "step": 2190
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.948905109489051e-06,
+      "loss": 0.4805,
+      "step": 2191
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9474452554744527e-06,
+      "loss": 0.4422,
+      "step": 2192
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.945985401459854e-06,
+      "loss": 0.4308,
+      "step": 2193
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9445255474452555e-06,
+      "loss": 0.4326,
+      "step": 2194
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.943065693430657e-06,
+      "loss": 0.4575,
+      "step": 2195
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9416058394160584e-06,
+      "loss": 0.4139,
+      "step": 2196
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.94014598540146e-06,
+      "loss": 0.3694,
+      "step": 2197
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9386861313868612e-06,
+      "loss": 0.4295,
+      "step": 2198
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.937226277372263e-06,
+      "loss": 0.4197,
+      "step": 2199
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9357664233576645e-06,
+      "loss": 0.4428,
+      "step": 2200
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9343065693430657e-06,
+      "loss": 0.4821,
+      "step": 2201
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9328467153284674e-06,
+      "loss": 0.4499,
+      "step": 2202
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9313868613138686e-06,
+      "loss": 0.4261,
+      "step": 2203
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9299270072992702e-06,
+      "loss": 0.4505,
+      "step": 2204
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9284671532846714e-06,
+      "loss": 0.4673,
+      "step": 2205
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.927007299270073e-06,
+      "loss": 0.4342,
+      "step": 2206
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9255474452554747e-06,
+      "loss": 0.3938,
+      "step": 2207
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.924087591240876e-06,
+      "loss": 0.4584,
+      "step": 2208
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9226277372262776e-06,
+      "loss": 0.4248,
+      "step": 2209
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9211678832116788e-06,
+      "loss": 0.4718,
+      "step": 2210
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9197080291970804e-06,
+      "loss": 0.4021,
+      "step": 2211
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.918248175182482e-06,
+      "loss": 0.3985,
+      "step": 2212
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9167883211678833e-06,
+      "loss": 0.4121,
+      "step": 2213
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.915328467153285e-06,
+      "loss": 0.423,
+      "step": 2214
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.913868613138686e-06,
+      "loss": 0.4193,
+      "step": 2215
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9124087591240877e-06,
+      "loss": 0.4195,
+      "step": 2216
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.910948905109489e-06,
+      "loss": 0.3955,
+      "step": 2217
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9094890510948906e-06,
+      "loss": 0.4124,
+      "step": 2218
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.908029197080292e-06,
+      "loss": 0.463,
+      "step": 2219
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9065693430656935e-06,
+      "loss": 0.415,
+      "step": 2220
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.905109489051095e-06,
+      "loss": 0.4122,
+      "step": 2221
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.9036496350364963e-06,
+      "loss": 0.4337,
+      "step": 2222
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.902189781021898e-06,
+      "loss": 0.4799,
+      "step": 2223
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.900729927007299e-06,
+      "loss": 0.4126,
+      "step": 2224
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.899270072992701e-06,
+      "loss": 0.466,
+      "step": 2225
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8978102189781024e-06,
+      "loss": 0.3785,
+      "step": 2226
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8963503649635036e-06,
+      "loss": 0.4105,
+      "step": 2227
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8948905109489053e-06,
+      "loss": 0.4097,
+      "step": 2228
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8934306569343065e-06,
+      "loss": 0.4501,
+      "step": 2229
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.891970802919708e-06,
+      "loss": 0.4185,
+      "step": 2230
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8905109489051098e-06,
+      "loss": 0.4364,
+      "step": 2231
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.889051094890511e-06,
+      "loss": 0.455,
+      "step": 2232
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8875912408759126e-06,
+      "loss": 0.4571,
+      "step": 2233
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.886131386861314e-06,
+      "loss": 0.4686,
+      "step": 2234
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8846715328467155e-06,
+      "loss": 0.4921,
+      "step": 2235
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.883211678832117e-06,
+      "loss": 0.4169,
+      "step": 2236
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8817518248175183e-06,
+      "loss": 0.4914,
+      "step": 2237
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8802919708029195e-06,
+      "loss": 0.4882,
+      "step": 2238
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.878832116788321e-06,
+      "loss": 0.491,
+      "step": 2239
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.877372262773723e-06,
+      "loss": 0.4061,
+      "step": 2240
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.8759124087591245e-06,
+      "loss": 0.4937,
+      "step": 2241
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8744525547445257e-06,
+      "loss": 0.4314,
+      "step": 2242
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.872992700729927e-06,
+      "loss": 0.4303,
+      "step": 2243
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8715328467153285e-06,
+      "loss": 0.4757,
+      "step": 2244
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.87007299270073e-06,
+      "loss": 0.4091,
+      "step": 2245
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.868613138686132e-06,
+      "loss": 0.4028,
+      "step": 2246
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.867153284671533e-06,
+      "loss": 0.4259,
+      "step": 2247
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8656934306569342e-06,
+      "loss": 0.3875,
+      "step": 2248
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.864233576642336e-06,
+      "loss": 0.4224,
+      "step": 2249
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8627737226277375e-06,
+      "loss": 0.4563,
+      "step": 2250
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8613138686131387e-06,
+      "loss": 0.4137,
+      "step": 2251
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.85985401459854e-06,
+      "loss": 0.4051,
+      "step": 2252
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8583941605839416e-06,
+      "loss": 0.402,
+      "step": 2253
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.856934306569343e-06,
+      "loss": 0.4278,
+      "step": 2254
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.855474452554745e-06,
+      "loss": 0.4009,
+      "step": 2255
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.854014598540146e-06,
+      "loss": 0.4283,
+      "step": 2256
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8525547445255473e-06,
+      "loss": 0.4163,
+      "step": 2257
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.851094890510949e-06,
+      "loss": 0.4565,
+      "step": 2258
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8496350364963505e-06,
+      "loss": 0.4743,
+      "step": 2259
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.848175182481752e-06,
+      "loss": 0.4394,
+      "step": 2260
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8467153284671534e-06,
+      "loss": 0.4651,
+      "step": 2261
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.8452554744525546e-06,
+      "loss": 0.4427,
+      "step": 2262
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8437956204379562e-06,
+      "loss": 0.4511,
+      "step": 2263
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.842335766423358e-06,
+      "loss": 0.4362,
+      "step": 2264
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8408759124087595e-06,
+      "loss": 0.4818,
+      "step": 2265
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8394160583941603e-06,
+      "loss": 0.449,
+      "step": 2266
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.837956204379562e-06,
+      "loss": 0.4989,
+      "step": 2267
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8364963503649636e-06,
+      "loss": 0.4633,
+      "step": 2268
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8350364963503652e-06,
+      "loss": 0.4684,
+      "step": 2269
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8335766423357664e-06,
+      "loss": 0.4167,
+      "step": 2270
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8321167883211677e-06,
+      "loss": 0.4176,
+      "step": 2271
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8306569343065693e-06,
+      "loss": 0.3945,
+      "step": 2272
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.829197080291971e-06,
+      "loss": 0.4186,
+      "step": 2273
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8277372262773726e-06,
+      "loss": 0.4326,
+      "step": 2274
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8262773722627738e-06,
+      "loss": 0.4303,
+      "step": 2275
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.824817518248175e-06,
+      "loss": 0.4562,
+      "step": 2276
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8233576642335766e-06,
+      "loss": 0.433,
+      "step": 2277
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8218978102189783e-06,
+      "loss": 0.4213,
+      "step": 2278
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.82043795620438e-06,
+      "loss": 0.4841,
+      "step": 2279
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.818978102189781e-06,
+      "loss": 0.4475,
+      "step": 2280
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8175182481751823e-06,
+      "loss": 0.3824,
+      "step": 2281
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.816058394160584e-06,
+      "loss": 0.4968,
+      "step": 2282
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.8145985401459856e-06,
+      "loss": 0.3962,
+      "step": 2283
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.813138686131387e-06,
+      "loss": 0.4692,
+      "step": 2284
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.8116788321167885e-06,
+      "loss": 0.5105,
+      "step": 2285
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.8102189781021897e-06,
+      "loss": 0.4342,
+      "step": 2286
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.8087591240875913e-06,
+      "loss": 0.416,
+      "step": 2287
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.807299270072993e-06,
+      "loss": 0.4986,
+      "step": 2288
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.805839416058394e-06,
+      "loss": 0.4585,
+      "step": 2289
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.8043795620437954e-06,
+      "loss": 0.4187,
+      "step": 2290
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.802919708029197e-06,
+      "loss": 0.3804,
+      "step": 2291
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.8014598540145987e-06,
+      "loss": 0.4561,
+      "step": 2292
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.4536,
+      "step": 2293
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7985401459854015e-06,
+      "loss": 0.4496,
+      "step": 2294
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7970802919708027e-06,
+      "loss": 0.49,
+      "step": 2295
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7956204379562044e-06,
+      "loss": 0.3898,
+      "step": 2296
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.794160583941606e-06,
+      "loss": 0.444,
+      "step": 2297
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7927007299270072e-06,
+      "loss": 0.4148,
+      "step": 2298
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.791240875912409e-06,
+      "loss": 0.4947,
+      "step": 2299
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.78978102189781e-06,
+      "loss": 0.4786,
+      "step": 2300
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7883211678832117e-06,
+      "loss": 0.4473,
+      "step": 2301
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7868613138686133e-06,
+      "loss": 0.4295,
+      "step": 2302
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.7854014598540146e-06,
+      "loss": 0.3927,
+      "step": 2303
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.783941605839416e-06,
+      "loss": 0.404,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7824817518248174e-06,
+      "loss": 0.4612,
+      "step": 2305
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.781021897810219e-06,
+      "loss": 0.4268,
+      "step": 2306
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7795620437956207e-06,
+      "loss": 0.4621,
+      "step": 2307
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.778102189781022e-06,
+      "loss": 0.5076,
+      "step": 2308
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7766423357664235e-06,
+      "loss": 0.392,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7751824817518247e-06,
+      "loss": 0.4653,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7737226277372264e-06,
+      "loss": 0.4291,
+      "step": 2311
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.772262773722628e-06,
+      "loss": 0.3835,
+      "step": 2312
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7708029197080292e-06,
+      "loss": 0.4061,
+      "step": 2313
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.769343065693431e-06,
+      "loss": 0.4225,
+      "step": 2314
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.767883211678832e-06,
+      "loss": 0.4129,
+      "step": 2315
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7664233576642337e-06,
+      "loss": 0.5338,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.764963503649635e-06,
+      "loss": 0.4222,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7635036496350366e-06,
+      "loss": 0.4304,
+      "step": 2318
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7620437956204382e-06,
+      "loss": 0.4273,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7605839416058394e-06,
+      "loss": 0.4847,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.759124087591241e-06,
+      "loss": 0.4143,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 0.47176411747932434,
+      "eval_runtime": 1.1238,
+      "eval_samples_per_second": 17.796,
+      "eval_steps_per_second": 0.89,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1,
+      "eval_bench_accuracy_agieval": 0.3559322033898305,
+      "eval_bench_accuracy_arc_challenge": 0.8148148148148148,
+      "eval_bench_accuracy_arc_easy": 0.8888888888888888,
+      "eval_bench_accuracy_bigbench": 0.4262295081967213,
+      "eval_bench_accuracy_boolq": 0.7777777777777778,
+      "eval_bench_accuracy_mmlu": 0.5185185185185185,
+      "eval_bench_accuracy_openbookqa": 0.18518518518518517,
+      "eval_bench_accuracy_truthful_qa": 0.2641509433962264,
+      "eval_bench_accuracy_winogrande": 0.5185185185185185,
+      "eval_bench_average_accuracy": 0.5277795954096091,
+      "eval_bench_loss": 16.20269701854292,
+      "eval_bench_total_accuracy": 0.5251509054325956,
+      "step": 2321
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7576642335766423e-06,
+      "loss": 0.4341,
+      "step": 2322
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.756204379562044e-06,
+      "loss": 0.3782,
+      "step": 2323
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7547445255474456e-06,
+      "loss": 0.4457,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7532846715328468e-06,
+      "loss": 0.4109,
+      "step": 2325
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.7518248175182484e-06,
+      "loss": 0.4185,
+      "step": 2326
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7503649635036496e-06,
+      "loss": 0.3995,
+      "step": 2327
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7489051094890513e-06,
+      "loss": 0.4351,
+      "step": 2328
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7474452554744525e-06,
+      "loss": 0.4376,
+      "step": 2329
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.745985401459854e-06,
+      "loss": 0.4368,
+      "step": 2330
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7445255474452553e-06,
+      "loss": 0.4594,
+      "step": 2331
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.743065693430657e-06,
+      "loss": 0.4659,
+      "step": 2332
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7416058394160586e-06,
+      "loss": 0.4408,
+      "step": 2333
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.74014598540146e-06,
+      "loss": 0.4479,
+      "step": 2334
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7386861313868615e-06,
+      "loss": 0.5242,
+      "step": 2335
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7372262773722627e-06,
+      "loss": 0.4833,
+      "step": 2336
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7357664233576643e-06,
+      "loss": 0.4654,
+      "step": 2337
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.734306569343066e-06,
+      "loss": 0.5312,
+      "step": 2338
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.732846715328467e-06,
+      "loss": 0.4073,
+      "step": 2339
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.731386861313869e-06,
+      "loss": 0.4191,
+      "step": 2340
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.72992700729927e-06,
+      "loss": 0.4675,
+      "step": 2341
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7284671532846716e-06,
+      "loss": 0.4281,
+      "step": 2342
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7270072992700733e-06,
+      "loss": 0.4304,
+      "step": 2343
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7255474452554745e-06,
+      "loss": 0.4065,
+      "step": 2344
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7240875912408757e-06,
+      "loss": 0.4433,
+      "step": 2345
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.7226277372262774e-06,
+      "loss": 0.385,
+      "step": 2346
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.721167883211679e-06,
+      "loss": 0.417,
+      "step": 2347
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7197080291970806e-06,
+      "loss": 0.3925,
+      "step": 2348
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.718248175182482e-06,
+      "loss": 0.3837,
+      "step": 2349
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.716788321167883e-06,
+      "loss": 0.4313,
+      "step": 2350
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7153284671532847e-06,
+      "loss": 0.4519,
+      "step": 2351
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7138686131386863e-06,
+      "loss": 0.4362,
+      "step": 2352
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.712408759124088e-06,
+      "loss": 0.4211,
+      "step": 2353
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.710948905109489e-06,
+      "loss": 0.4243,
+      "step": 2354
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7094890510948904e-06,
+      "loss": 0.4361,
+      "step": 2355
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.708029197080292e-06,
+      "loss": 0.5023,
+      "step": 2356
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7065693430656937e-06,
+      "loss": 0.4281,
+      "step": 2357
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7051094890510953e-06,
+      "loss": 0.4271,
+      "step": 2358
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7036496350364965e-06,
+      "loss": 0.3991,
+      "step": 2359
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7021897810218977e-06,
+      "loss": 0.4372,
+      "step": 2360
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.7007299270072994e-06,
+      "loss": 0.3729,
+      "step": 2361
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.699270072992701e-06,
+      "loss": 0.4671,
+      "step": 2362
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.6978102189781022e-06,
+      "loss": 0.3825,
+      "step": 2363
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.6963503649635034e-06,
+      "loss": 0.4016,
+      "step": 2364
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.694890510948905e-06,
+      "loss": 0.4343,
+      "step": 2365
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.6934306569343067e-06,
+      "loss": 0.4427,
+      "step": 2366
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.6919708029197084e-06,
+      "loss": 0.4236,
+      "step": 2367
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.6905109489051096e-06,
+      "loss": 0.4524,
+      "step": 2368
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6890510948905108e-06,
+      "loss": 0.4842,
+      "step": 2369
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6875912408759124e-06,
+      "loss": 0.4083,
+      "step": 2370
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.686131386861314e-06,
+      "loss": 0.4225,
+      "step": 2371
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6846715328467157e-06,
+      "loss": 0.4207,
+      "step": 2372
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.683211678832117e-06,
+      "loss": 0.4337,
+      "step": 2373
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.681751824817518e-06,
+      "loss": 0.4555,
+      "step": 2374
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6802919708029198e-06,
+      "loss": 0.4547,
+      "step": 2375
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6788321167883214e-06,
+      "loss": 0.4447,
+      "step": 2376
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6773722627737226e-06,
+      "loss": 0.4275,
+      "step": 2377
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.675912408759124e-06,
+      "loss": 0.4463,
+      "step": 2378
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6744525547445255e-06,
+      "loss": 0.4677,
+      "step": 2379
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.672992700729927e-06,
+      "loss": 0.4193,
+      "step": 2380
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6715328467153287e-06,
+      "loss": 0.4163,
+      "step": 2381
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.67007299270073e-06,
+      "loss": 0.443,
+      "step": 2382
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.668613138686131e-06,
+      "loss": 0.4633,
+      "step": 2383
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.667153284671533e-06,
+      "loss": 0.4236,
+      "step": 2384
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6656934306569344e-06,
+      "loss": 0.4648,
+      "step": 2385
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.664233576642336e-06,
+      "loss": 0.3942,
+      "step": 2386
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6627737226277373e-06,
+      "loss": 0.4,
+      "step": 2387
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.6613138686131385e-06,
+      "loss": 0.456,
+      "step": 2388
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.65985401459854e-06,
+      "loss": 0.3959,
+      "step": 2389
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6583941605839418e-06,
+      "loss": 0.4647,
+      "step": 2390
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6569343065693434e-06,
+      "loss": 0.4606,
+      "step": 2391
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6554744525547446e-06,
+      "loss": 0.4391,
+      "step": 2392
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.654014598540146e-06,
+      "loss": 0.3848,
+      "step": 2393
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6525547445255475e-06,
+      "loss": 0.4252,
+      "step": 2394
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.651094890510949e-06,
+      "loss": 0.4302,
+      "step": 2395
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6496350364963503e-06,
+      "loss": 0.4236,
+      "step": 2396
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.648175182481752e-06,
+      "loss": 0.4252,
+      "step": 2397
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.646715328467153e-06,
+      "loss": 0.4409,
+      "step": 2398
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.645255474452555e-06,
+      "loss": 0.4109,
+      "step": 2399
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6437956204379565e-06,
+      "loss": 0.4157,
+      "step": 2400
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6423357664233577e-06,
+      "loss": 0.3921,
+      "step": 2401
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.640875912408759e-06,
+      "loss": 0.4814,
+      "step": 2402
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6394160583941605e-06,
+      "loss": 0.4601,
+      "step": 2403
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.637956204379562e-06,
+      "loss": 0.4028,
+      "step": 2404
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.636496350364964e-06,
+      "loss": 0.4442,
+      "step": 2405
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.635036496350365e-06,
+      "loss": 0.462,
+      "step": 2406
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6335766423357662e-06,
+      "loss": 0.4388,
+      "step": 2407
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.632116788321168e-06,
+      "loss": 0.3794,
+      "step": 2408
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6306569343065695e-06,
+      "loss": 0.4912,
+      "step": 2409
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.6291970802919707e-06,
+      "loss": 0.4487,
+      "step": 2410
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6277372262773724e-06,
+      "loss": 0.4103,
+      "step": 2411
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6262773722627736e-06,
+      "loss": 0.4079,
+      "step": 2412
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6248175182481752e-06,
+      "loss": 0.4056,
+      "step": 2413
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.623357664233577e-06,
+      "loss": 0.4529,
+      "step": 2414
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.621897810218978e-06,
+      "loss": 0.4445,
+      "step": 2415
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6204379562043797e-06,
+      "loss": 0.4023,
+      "step": 2416
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.618978102189781e-06,
+      "loss": 0.4428,
+      "step": 2417
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6175182481751826e-06,
+      "loss": 0.3888,
+      "step": 2418
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.616058394160584e-06,
+      "loss": 0.4265,
+      "step": 2419
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6145985401459854e-06,
+      "loss": 0.4404,
+      "step": 2420
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.613138686131387e-06,
+      "loss": 0.4324,
+      "step": 2421
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6116788321167883e-06,
+      "loss": 0.4059,
+      "step": 2422
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.61021897810219e-06,
+      "loss": 0.4217,
+      "step": 2423
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.608759124087591e-06,
+      "loss": 0.4285,
+      "step": 2424
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6072992700729928e-06,
+      "loss": 0.4161,
+      "step": 2425
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6058394160583944e-06,
+      "loss": 0.4232,
+      "step": 2426
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6043795620437956e-06,
+      "loss": 0.4131,
+      "step": 2427
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6029197080291972e-06,
+      "loss": 0.4382,
+      "step": 2428
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6014598540145985e-06,
+      "loss": 0.4667,
+      "step": 2429
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.6e-06,
+      "loss": 0.4157,
+      "step": 2430
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.5985401459854017e-06,
+      "loss": 0.3893,
+      "step": 2431
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.597080291970803e-06,
+      "loss": 0.4084,
+      "step": 2432
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5956204379562046e-06,
+      "loss": 0.4647,
+      "step": 2433
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.594160583941606e-06,
+      "loss": 0.4368,
+      "step": 2434
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5927007299270074e-06,
+      "loss": 0.454,
+      "step": 2435
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.591240875912409e-06,
+      "loss": 0.4446,
+      "step": 2436
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5897810218978103e-06,
+      "loss": 0.4472,
+      "step": 2437
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.588321167883212e-06,
+      "loss": 0.3899,
+      "step": 2438
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.586861313868613e-06,
+      "loss": 0.4139,
+      "step": 2439
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5854014598540148e-06,
+      "loss": 0.4391,
+      "step": 2440
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.583941605839416e-06,
+      "loss": 0.4233,
+      "step": 2441
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5824817518248176e-06,
+      "loss": 0.4029,
+      "step": 2442
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.581021897810219e-06,
+      "loss": 0.4427,
+      "step": 2443
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5795620437956205e-06,
+      "loss": 0.4879,
+      "step": 2444
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.578102189781022e-06,
+      "loss": 0.4052,
+      "step": 2445
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5766423357664233e-06,
+      "loss": 0.4086,
+      "step": 2446
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.575182481751825e-06,
+      "loss": 0.4513,
+      "step": 2447
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.573722627737226e-06,
+      "loss": 0.4119,
+      "step": 2448
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.572262773722628e-06,
+      "loss": 0.3751,
+      "step": 2449
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5708029197080295e-06,
+      "loss": 0.4383,
+      "step": 2450
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5693430656934307e-06,
+      "loss": 0.3965,
+      "step": 2451
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.5678832116788323e-06,
+      "loss": 0.45,
+      "step": 2452
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5664233576642335e-06,
+      "loss": 0.4359,
+      "step": 2453
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.564963503649635e-06,
+      "loss": 0.4213,
+      "step": 2454
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.563503649635037e-06,
+      "loss": 0.4479,
+      "step": 2455
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.562043795620438e-06,
+      "loss": 0.4107,
+      "step": 2456
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5605839416058392e-06,
+      "loss": 0.4112,
+      "step": 2457
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.559124087591241e-06,
+      "loss": 0.3953,
+      "step": 2458
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5576642335766425e-06,
+      "loss": 0.4191,
+      "step": 2459
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.556204379562044e-06,
+      "loss": 0.4343,
+      "step": 2460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5547445255474454e-06,
+      "loss": 0.3863,
+      "step": 2461
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5532846715328466e-06,
+      "loss": 0.4334,
+      "step": 2462
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.551824817518248e-06,
+      "loss": 0.4359,
+      "step": 2463
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.55036496350365e-06,
+      "loss": 0.437,
+      "step": 2464
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5489051094890515e-06,
+      "loss": 0.4251,
+      "step": 2465
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5474452554744527e-06,
+      "loss": 0.4283,
+      "step": 2466
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.545985401459854e-06,
+      "loss": 0.4515,
+      "step": 2467
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5445255474452555e-06,
+      "loss": 0.4634,
+      "step": 2468
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.543065693430657e-06,
+      "loss": 0.437,
+      "step": 2469
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.541605839416059e-06,
+      "loss": 0.4102,
+      "step": 2470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5401459854014596e-06,
+      "loss": 0.4432,
+      "step": 2471
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.5386861313868613e-06,
+      "loss": 0.4255,
+      "step": 2472
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.537226277372263e-06,
+      "loss": 0.4233,
+      "step": 2473
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5357664233576645e-06,
+      "loss": 0.4466,
+      "step": 2474
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5343065693430657e-06,
+      "loss": 0.4171,
+      "step": 2475
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.532846715328467e-06,
+      "loss": 0.411,
+      "step": 2476
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5313868613138686e-06,
+      "loss": 0.424,
+      "step": 2477
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5299270072992702e-06,
+      "loss": 0.4165,
+      "step": 2478
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.528467153284672e-06,
+      "loss": 0.4754,
+      "step": 2479
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.527007299270073e-06,
+      "loss": 0.4447,
+      "step": 2480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5255474452554743e-06,
+      "loss": 0.4555,
+      "step": 2481
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.524087591240876e-06,
+      "loss": 0.5322,
+      "step": 2482
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5226277372262776e-06,
+      "loss": 0.3895,
+      "step": 2483
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.521167883211679e-06,
+      "loss": 0.4402,
+      "step": 2484
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.51970802919708e-06,
+      "loss": 0.4718,
+      "step": 2485
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5182481751824816e-06,
+      "loss": 0.4401,
+      "step": 2486
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5167883211678833e-06,
+      "loss": 0.4441,
+      "step": 2487
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.515328467153285e-06,
+      "loss": 0.4065,
+      "step": 2488
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.513868613138686e-06,
+      "loss": 0.4628,
+      "step": 2489
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5124087591240873e-06,
+      "loss": 0.4422,
+      "step": 2490
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.510948905109489e-06,
+      "loss": 0.3762,
+      "step": 2491
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5094890510948906e-06,
+      "loss": 0.4415,
+      "step": 2492
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5080291970802923e-06,
+      "loss": 0.4754,
+      "step": 2493
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.5065693430656935e-06,
+      "loss": 0.4684,
+      "step": 2494
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5051094890510947e-06,
+      "loss": 0.4522,
+      "step": 2495
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5036496350364963e-06,
+      "loss": 0.4323,
+      "step": 2496
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.502189781021898e-06,
+      "loss": 0.4176,
+      "step": 2497
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.5007299270072996e-06,
+      "loss": 0.438,
+      "step": 2498
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.499270072992701e-06,
+      "loss": 0.4069,
+      "step": 2499
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.497810218978102e-06,
+      "loss": 0.4146,
+      "step": 2500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4963503649635037e-06,
+      "loss": 0.4163,
+      "step": 2501
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4948905109489053e-06,
+      "loss": 0.4102,
+      "step": 2502
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4934306569343065e-06,
+      "loss": 0.4269,
+      "step": 2503
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.491970802919708e-06,
+      "loss": 0.4696,
+      "step": 2504
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4905109489051094e-06,
+      "loss": 0.437,
+      "step": 2505
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.489051094890511e-06,
+      "loss": 0.3888,
+      "step": 2506
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4875912408759126e-06,
+      "loss": 0.4182,
+      "step": 2507
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.486131386861314e-06,
+      "loss": 0.4778,
+      "step": 2508
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4846715328467155e-06,
+      "loss": 0.4051,
+      "step": 2509
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4832116788321167e-06,
+      "loss": 0.427,
+      "step": 2510
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4817518248175183e-06,
+      "loss": 0.43,
+      "step": 2511
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.48029197080292e-06,
+      "loss": 0.4712,
+      "step": 2512
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.478832116788321e-06,
+      "loss": 0.4259,
+      "step": 2513
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.4773722627737224e-06,
+      "loss": 0.4534,
+      "step": 2514
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.475912408759124e-06,
+      "loss": 0.4002,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4744525547445257e-06,
+      "loss": 0.4219,
+      "step": 2516
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4729927007299273e-06,
+      "loss": 0.4585,
+      "step": 2517
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4715328467153285e-06,
+      "loss": 0.4281,
+      "step": 2518
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4700729927007298e-06,
+      "loss": 0.4429,
+      "step": 2519
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4686131386861314e-06,
+      "loss": 0.441,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.467153284671533e-06,
+      "loss": 0.4238,
+      "step": 2521
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4656934306569342e-06,
+      "loss": 0.3915,
+      "step": 2522
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.464233576642336e-06,
+      "loss": 0.4058,
+      "step": 2523
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.462773722627737e-06,
+      "loss": 0.4309,
+      "step": 2524
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4613138686131387e-06,
+      "loss": 0.4387,
+      "step": 2525
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4598540145985404e-06,
+      "loss": 0.4649,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4583941605839416e-06,
+      "loss": 0.4273,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4569343065693432e-06,
+      "loss": 0.4284,
+      "step": 2528
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4554744525547444e-06,
+      "loss": 0.4061,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.454014598540146e-06,
+      "loss": 0.4347,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4525547445255477e-06,
+      "loss": 0.434,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.451094890510949e-06,
+      "loss": 0.445,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 0.4694254696369171,
+      "eval_runtime": 1.1219,
+      "eval_samples_per_second": 17.827,
+      "eval_steps_per_second": 0.891,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2,
+      "eval_bench_accuracy_agieval": 0.3389830508474576,
+      "eval_bench_accuracy_arc_challenge": 0.8518518518518519,
+      "eval_bench_accuracy_arc_easy": 0.8518518518518519,
+      "eval_bench_accuracy_bigbench": 0.47540983606557374,
+      "eval_bench_accuracy_boolq": 0.7222222222222222,
+      "eval_bench_accuracy_mmlu": 0.4444444444444444,
+      "eval_bench_accuracy_openbookqa": 0.2037037037037037,
+      "eval_bench_accuracy_truthful_qa": 0.24528301886792453,
+      "eval_bench_accuracy_winogrande": 0.5370370370370371,
+      "eval_bench_average_accuracy": 0.5189763352102297,
+      "eval_bench_loss": 16.107176263648345,
+      "eval_bench_total_accuracy": 0.5171026156941649,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4496350364963506e-06,
+      "loss": 0.4232,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4481751824817518e-06,
+      "loss": 0.4281,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4467153284671534e-06,
+      "loss": 0.4766,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.4452554744525546e-06,
+      "loss": 0.4605,
+      "step": 2536
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4437956204379563e-06,
+      "loss": 0.407,
+      "step": 2537
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.442335766423358e-06,
+      "loss": 0.4561,
+      "step": 2538
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.440875912408759e-06,
+      "loss": 0.422,
+      "step": 2539
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4394160583941608e-06,
+      "loss": 0.3979,
+      "step": 2540
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.437956204379562e-06,
+      "loss": 0.4535,
+      "step": 2541
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4364963503649636e-06,
+      "loss": 0.4056,
+      "step": 2542
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4350364963503652e-06,
+      "loss": 0.461,
+      "step": 2543
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4335766423357665e-06,
+      "loss": 0.4352,
+      "step": 2544
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.432116788321168e-06,
+      "loss": 0.4089,
+      "step": 2545
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4306569343065693e-06,
+      "loss": 0.4084,
+      "step": 2546
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.429197080291971e-06,
+      "loss": 0.3989,
+      "step": 2547
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4277372262773726e-06,
+      "loss": 0.4871,
+      "step": 2548
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.426277372262774e-06,
+      "loss": 0.4445,
+      "step": 2549
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.424817518248175e-06,
+      "loss": 0.401,
+      "step": 2550
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4233576642335767e-06,
+      "loss": 0.3973,
+      "step": 2551
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4218978102189783e-06,
+      "loss": 0.426,
+      "step": 2552
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4204379562043795e-06,
+      "loss": 0.436,
+      "step": 2553
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.418978102189781e-06,
+      "loss": 0.4157,
+      "step": 2554
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4175182481751824e-06,
+      "loss": 0.5141,
+      "step": 2555
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.416058394160584e-06,
+      "loss": 0.4014,
+      "step": 2556
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.4145985401459856e-06,
+      "loss": 0.4148,
+      "step": 2557
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.413138686131387e-06,
+      "loss": 0.379,
+      "step": 2558
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.4116788321167885e-06,
+      "loss": 0.4563,
+      "step": 2559
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.4102189781021897e-06,
+      "loss": 0.4231,
+      "step": 2560
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.4087591240875913e-06,
+      "loss": 0.4239,
+      "step": 2561
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.407299270072993e-06,
+      "loss": 0.4127,
+      "step": 2562
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.405839416058394e-06,
+      "loss": 0.4575,
+      "step": 2563
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.4043795620437954e-06,
+      "loss": 0.5009,
+      "step": 2564
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.402919708029197e-06,
+      "loss": 0.4615,
+      "step": 2565
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.4014598540145987e-06,
+      "loss": 0.3997,
+      "step": 2566
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.5106,
+      "step": 2567
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.3985401459854015e-06,
+      "loss": 0.4451,
+      "step": 2568
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.3970802919708027e-06,
+      "loss": 0.4434,
+      "step": 2569
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.3956204379562044e-06,
+      "loss": 0.3705,
+      "step": 2570
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.394160583941606e-06,
+      "loss": 0.4154,
+      "step": 2571
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.3927007299270077e-06,
+      "loss": 0.464,
+      "step": 2572
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.391240875912409e-06,
+      "loss": 0.3893,
+      "step": 2573
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.38978102189781e-06,
+      "loss": 0.4768,
+      "step": 2574
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.3883211678832117e-06,
+      "loss": 0.4092,
+      "step": 2575
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.3868613138686134e-06,
+      "loss": 0.4099,
+      "step": 2576
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.385401459854015e-06,
+      "loss": 0.422,
+      "step": 2577
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.383941605839416e-06,
+      "loss": 0.3885,
+      "step": 2578
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3824817518248174e-06,
+      "loss": 0.4014,
+      "step": 2579
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.381021897810219e-06,
+      "loss": 0.3844,
+      "step": 2580
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3795620437956207e-06,
+      "loss": 0.4448,
+      "step": 2581
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.378102189781022e-06,
+      "loss": 0.4682,
+      "step": 2582
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.376642335766423e-06,
+      "loss": 0.4207,
+      "step": 2583
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3751824817518248e-06,
+      "loss": 0.4456,
+      "step": 2584
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3737226277372264e-06,
+      "loss": 0.4052,
+      "step": 2585
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.372262773722628e-06,
+      "loss": 0.4595,
+      "step": 2586
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3708029197080293e-06,
+      "loss": 0.3994,
+      "step": 2587
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3693430656934305e-06,
+      "loss": 0.4265,
+      "step": 2588
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.367883211678832e-06,
+      "loss": 0.4312,
+      "step": 2589
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3664233576642337e-06,
+      "loss": 0.4248,
+      "step": 2590
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3649635036496354e-06,
+      "loss": 0.4244,
+      "step": 2591
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3635036496350366e-06,
+      "loss": 0.4427,
+      "step": 2592
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.362043795620438e-06,
+      "loss": 0.4868,
+      "step": 2593
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3605839416058394e-06,
+      "loss": 0.4389,
+      "step": 2594
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.359124087591241e-06,
+      "loss": 0.4175,
+      "step": 2595
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3576642335766427e-06,
+      "loss": 0.4014,
+      "step": 2596
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.3562043795620435e-06,
+      "loss": 0.4331,
+      "step": 2597
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.354744525547445e-06,
+      "loss": 0.448,
+      "step": 2598
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.353284671532847e-06,
+      "loss": 0.4731,
+      "step": 2599
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3518248175182484e-06,
+      "loss": 0.4047,
+      "step": 2600
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3503649635036496e-06,
+      "loss": 0.4704,
+      "step": 2601
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.348905109489051e-06,
+      "loss": 0.4419,
+      "step": 2602
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3474452554744525e-06,
+      "loss": 0.4138,
+      "step": 2603
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.345985401459854e-06,
+      "loss": 0.4514,
+      "step": 2604
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3445255474452558e-06,
+      "loss": 0.4409,
+      "step": 2605
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.343065693430657e-06,
+      "loss": 0.4236,
+      "step": 2606
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.341605839416058e-06,
+      "loss": 0.4005,
+      "step": 2607
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.34014598540146e-06,
+      "loss": 0.4233,
+      "step": 2608
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3386861313868615e-06,
+      "loss": 0.4366,
+      "step": 2609
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.337226277372263e-06,
+      "loss": 0.4455,
+      "step": 2610
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3357664233576643e-06,
+      "loss": 0.4509,
+      "step": 2611
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3343065693430655e-06,
+      "loss": 0.4259,
+      "step": 2612
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.332846715328467e-06,
+      "loss": 0.4023,
+      "step": 2613
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.331386861313869e-06,
+      "loss": 0.4208,
+      "step": 2614
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.32992700729927e-06,
+      "loss": 0.5032,
+      "step": 2615
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3284671532846717e-06,
+      "loss": 0.4269,
+      "step": 2616
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.327007299270073e-06,
+      "loss": 0.4884,
+      "step": 2617
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3255474452554745e-06,
+      "loss": 0.4254,
+      "step": 2618
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.324087591240876e-06,
+      "loss": 0.3964,
+      "step": 2619
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.3226277372262774e-06,
+      "loss": 0.3735,
+      "step": 2620
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.321167883211679e-06,
+      "loss": 0.4141,
+      "step": 2621
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3197080291970802e-06,
+      "loss": 0.4577,
+      "step": 2622
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.318248175182482e-06,
+      "loss": 0.4208,
+      "step": 2623
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3167883211678835e-06,
+      "loss": 0.3875,
+      "step": 2624
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3153284671532847e-06,
+      "loss": 0.4326,
+      "step": 2625
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.313868613138686e-06,
+      "loss": 0.4063,
+      "step": 2626
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3124087591240876e-06,
+      "loss": 0.4226,
+      "step": 2627
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.310948905109489e-06,
+      "loss": 0.4803,
+      "step": 2628
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3094890510948904e-06,
+      "loss": 0.4491,
+      "step": 2629
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.308029197080292e-06,
+      "loss": 0.4296,
+      "step": 2630
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3065693430656933e-06,
+      "loss": 0.4689,
+      "step": 2631
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.305109489051095e-06,
+      "loss": 0.3915,
+      "step": 2632
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3036496350364965e-06,
+      "loss": 0.4506,
+      "step": 2633
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3021897810218978e-06,
+      "loss": 0.443,
+      "step": 2634
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.3007299270072994e-06,
+      "loss": 0.4173,
+      "step": 2635
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.2992700729927006e-06,
+      "loss": 0.3914,
+      "step": 2636
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.2978102189781022e-06,
+      "loss": 0.4048,
+      "step": 2637
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.296350364963504e-06,
+      "loss": 0.4203,
+      "step": 2638
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.294890510948905e-06,
+      "loss": 0.4255,
+      "step": 2639
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.2934306569343067e-06,
+      "loss": 0.4094,
+      "step": 2640
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.291970802919708e-06,
+      "loss": 0.4183,
+      "step": 2641
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.2905109489051096e-06,
+      "loss": 0.4933,
+      "step": 2642
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.289051094890511e-06,
+      "loss": 0.396,
+      "step": 2643
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.2875912408759124e-06,
+      "loss": 0.4219,
+      "step": 2644
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.286131386861314e-06,
+      "loss": 0.4317,
+      "step": 2645
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.2846715328467153e-06,
+      "loss": 0.4422,
+      "step": 2646
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.283211678832117e-06,
+      "loss": 0.4363,
+      "step": 2647
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.281751824817518e-06,
+      "loss": 0.414,
+      "step": 2648
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.2802919708029198e-06,
+      "loss": 0.4408,
+      "step": 2649
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.2788321167883214e-06,
+      "loss": 0.414,
+      "step": 2650
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4210,
+  "num_train_epochs": 2,
+  "save_steps": 50,
+  "total_flos": 1.1114213452598477e+19,
+  "trial_name": null,
+  "trial_params": null
+}