diff --git "a/mistral-7b/trainer_state.json" "b/mistral-7b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/mistral-7b/trainer_state.json"
@@ -0,0 +1,8645 @@
+{
+  "best_metric": 1.8956754207611084,
+  "best_model_checkpoint": "./lora-out/checkpoint-704",
+  "epoch": 3.9774011299435026,
+  "eval_steps": 64,
+  "global_step": 1408,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.188679245283019e-07,
+      "loss": 1.8763,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0377358490566038e-06,
+      "loss": 2.0594,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5566037735849058e-06,
+      "loss": 1.8093,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0754716981132075e-06,
+      "loss": 1.8991,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5943396226415095e-06,
+      "loss": 1.9986,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1132075471698115e-06,
+      "loss": 1.9091,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632075471698113e-06,
+      "loss": 1.7758,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150943396226415e-06,
+      "loss": 1.9658,
+      "step": 8
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.669811320754717e-06,
+      "loss": 2.1666,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.188679245283019e-06,
+      "loss": 2.0388,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.7075471698113215e-06,
+      "loss": 1.9876,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 2.05,
+      "step": 12
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.745283018867925e-06,
+      "loss": 2.1043,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.264150943396226e-06,
+      "loss": 2.1008,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.783018867924528e-06,
+      "loss": 1.8744,
+      "step": 15
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.30188679245283e-06,
+      "loss": 1.9635,
+      "step": 16
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.820754716981133e-06,
+      "loss": 1.8415,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.339622641509433e-06,
+      "loss": 1.928,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.858490566037736e-06,
+      "loss": 2.084,
+      "step": 19
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 2.0203,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.089622641509434e-05,
+      "loss": 2.0735,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1415094339622643e-05,
+      "loss": 1.8626,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1933962264150944e-05,
+      "loss": 1.8078,
+      "step": 23
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 2.0432,
+      "step": 24
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.2971698113207549e-05,
+      "loss": 2.0748,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.349056603773585e-05,
+      "loss": 2.0721,
+      "step": 26
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.4009433962264152e-05,
+      "loss": 2.0951,
+      "step": 27
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.4528301886792452e-05,
+      "loss": 1.8304,
+      "step": 28
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5047169811320755e-05,
+      "loss": 2.016,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5566037735849056e-05,
+      "loss": 1.9287,
+      "step": 30
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.6084905660377358e-05,
+      "loss": 1.9704,
+      "step": 31
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.660377358490566e-05,
+      "loss": 1.9437,
+      "step": 32
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7122641509433963e-05,
+      "loss": 1.9436,
+      "step": 33
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7641509433962265e-05,
+      "loss": 1.9706,
+      "step": 34
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.8160377358490564e-05,
+      "loss": 1.856,
+      "step": 35
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.8679245283018867e-05,
+      "loss": 1.997,
+      "step": 36
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.919811320754717e-05,
+      "loss": 1.8972,
+      "step": 37
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.971698113207547e-05,
+      "loss": 1.9616,
+      "step": 38
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.0235849056603774e-05,
+      "loss": 2.0965,
+      "step": 39
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 1.9266,
+      "step": 40
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.127358490566038e-05,
+      "loss": 1.9404,
+      "step": 41
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.179245283018868e-05,
+      "loss": 2.0763,
+      "step": 42
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.2311320754716984e-05,
+      "loss": 1.8434,
+      "step": 43
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.2830188679245286e-05,
+      "loss": 1.9364,
+      "step": 44
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.3349056603773585e-05,
+      "loss": 1.9503,
+      "step": 45
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.3867924528301887e-05,
+      "loss": 1.9163,
+      "step": 46
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.438679245283019e-05,
+      "loss": 1.9609,
+      "step": 47
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.4905660377358492e-05,
+      "loss": 1.9648,
+      "step": 48
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.5424528301886795e-05,
+      "loss": 1.8921,
+      "step": 49
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.5943396226415097e-05,
+      "loss": 1.8229,
+      "step": 50
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.6462264150943396e-05,
+      "loss": 1.8533,
+      "step": 51
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.69811320754717e-05,
+      "loss": 1.8729,
+      "step": 52
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.75e-05,
+      "loss": 2.0393,
+      "step": 53
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.8018867924528303e-05,
+      "loss": 2.1571,
+      "step": 54
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.8537735849056606e-05,
+      "loss": 2.0352,
+      "step": 55
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9056603773584905e-05,
+      "loss": 2.1406,
+      "step": 56
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9575471698113207e-05,
+      "loss": 1.9978,
+      "step": 57
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 3.009433962264151e-05,
+      "loss": 2.1136,
+      "step": 58
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 3.061320754716981e-05,
+      "loss": 1.8786,
+      "step": 59
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 3.113207547169811e-05,
+      "loss": 1.8755,
+      "step": 60
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 3.165094339622642e-05,
+      "loss": 1.9923,
+      "step": 61
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 3.2169811320754716e-05,
+      "loss": 1.9311,
+      "step": 62
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 3.268867924528302e-05,
+      "loss": 1.954,
+      "step": 63
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 3.320754716981132e-05,
+      "loss": 1.8866,
+      "step": 64
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 1.9267239570617676,
+      "eval_runtime": 15.8456,
+      "eval_samples_per_second": 0.947,
+      "eval_steps_per_second": 0.252,
+      "step": 64
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 3.372641509433962e-05,
+      "loss": 1.9747,
+      "step": 65
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 3.4245283018867926e-05,
+      "loss": 1.9244,
+      "step": 66
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 3.4764150943396225e-05,
+      "loss": 2.0561,
+      "step": 67
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 3.528301886792453e-05,
+      "loss": 2.2111,
+      "step": 68
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 3.580188679245283e-05,
+      "loss": 2.1068,
+      "step": 69
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.632075471698113e-05,
+      "loss": 2.0595,
+      "step": 70
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.6839622641509434e-05,
+      "loss": 1.8751,
+      "step": 71
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.735849056603773e-05,
+      "loss": 2.0406,
+      "step": 72
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.787735849056604e-05,
+      "loss": 1.7887,
+      "step": 73
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.839622641509434e-05,
+      "loss": 1.928,
+      "step": 74
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.891509433962264e-05,
+      "loss": 2.0033,
+      "step": 75
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.943396226415094e-05,
+      "loss": 1.8148,
+      "step": 76
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.995283018867924e-05,
+      "loss": 1.8377,
+      "step": 77
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.047169811320755e-05,
+      "loss": 1.8968,
+      "step": 78
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.099056603773585e-05,
+      "loss": 1.8268,
+      "step": 79
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.150943396226415e-05,
+      "loss": 1.817,
+      "step": 80
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.202830188679246e-05,
+      "loss": 1.999,
+      "step": 81
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.254716981132076e-05,
+      "loss": 2.1163,
+      "step": 82
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.306603773584906e-05,
+      "loss": 1.818,
+      "step": 83
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.358490566037736e-05,
+      "loss": 2.043,
+      "step": 84
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.410377358490566e-05,
+      "loss": 2.0877,
+      "step": 85
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.462264150943397e-05,
+      "loss": 1.8506,
+      "step": 86
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5141509433962266e-05,
+      "loss": 1.9486,
+      "step": 87
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.566037735849057e-05,
+      "loss": 1.8371,
+      "step": 88
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.617924528301887e-05,
+      "loss": 2.0663,
+      "step": 89
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.669811320754717e-05,
+      "loss": 2.0218,
+      "step": 90
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7216981132075476e-05,
+      "loss": 1.7558,
+      "step": 91
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7735849056603775e-05,
+      "loss": 1.914,
+      "step": 92
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.825471698113208e-05,
+      "loss": 1.8807,
+      "step": 93
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.877358490566038e-05,
+      "loss": 1.771,
+      "step": 94
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9292452830188686e-05,
+      "loss": 1.8389,
+      "step": 95
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9811320754716985e-05,
+      "loss": 1.8625,
+      "step": 96
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.0330188679245284e-05,
+      "loss": 2.0654,
+      "step": 97
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.084905660377359e-05,
+      "loss": 1.8699,
+      "step": 98
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.136792452830189e-05,
+      "loss": 2.0099,
+      "step": 99
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.1886792452830194e-05,
+      "loss": 2.2083,
+      "step": 100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.240566037735849e-05,
+      "loss": 2.0373,
+      "step": 101
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.292452830188679e-05,
+      "loss": 2.0114,
+      "step": 102
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.34433962264151e-05,
+      "loss": 1.878,
+      "step": 103
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.39622641509434e-05,
+      "loss": 1.8008,
+      "step": 104
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.44811320754717e-05,
+      "loss": 1.9315,
+      "step": 105
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7969,
+      "step": 106
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9349,
+      "step": 107
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0334,
+      "step": 108
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0713,
+      "step": 109
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1895,
+      "step": 110
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8581,
+      "step": 111
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9417,
+      "step": 112
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9698,
+      "step": 113
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8789,
+      "step": 114
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1041,
+      "step": 115
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8171,
+      "step": 116
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9154,
+      "step": 117
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9921,
+      "step": 118
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.046,
+      "step": 119
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0098,
+      "step": 120
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9567,
+      "step": 121
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9345,
+      "step": 122
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9262,
+      "step": 123
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9932,
+      "step": 124
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0656,
+      "step": 125
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8495,
+      "step": 126
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0401,
+      "step": 127
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0231,
+      "step": 128
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 1.9198355674743652,
+      "eval_runtime": 15.8701,
+      "eval_samples_per_second": 0.945,
+      "eval_steps_per_second": 0.252,
+      "step": 128
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9994,
+      "step": 129
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8501,
+      "step": 130
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8624,
+      "step": 131
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9229,
+      "step": 132
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0773,
+      "step": 133
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0269,
+      "step": 134
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8407,
+      "step": 135
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9352,
+      "step": 136
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9501,
+      "step": 137
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9239,
+      "step": 138
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9907,
+      "step": 139
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9272,
+      "step": 140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8584,
+      "step": 141
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.5e-05,
+      "loss": 2.096,
+      "step": 142
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9345,
+      "step": 143
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8725,
+      "step": 144
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8895,
+      "step": 145
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7989,
+      "step": 146
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8534,
+      "step": 147
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9822,
+      "step": 148
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0224,
+      "step": 149
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0449,
+      "step": 150
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8775,
+      "step": 151
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8746,
+      "step": 152
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.932,
+      "step": 153
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0068,
+      "step": 154
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0045,
+      "step": 155
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9501,
+      "step": 156
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9689,
+      "step": 157
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1364,
+      "step": 158
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7746,
+      "step": 159
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.852,
+      "step": 160
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9932,
+      "step": 161
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8331,
+      "step": 162
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8655,
+      "step": 163
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0385,
+      "step": 164
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9515,
+      "step": 165
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0299,
+      "step": 166
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9165,
+      "step": 167
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9014,
+      "step": 168
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9239,
+      "step": 169
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8516,
+      "step": 170
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8151,
+      "step": 171
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1556,
+      "step": 172
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9864,
+      "step": 173
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9995,
+      "step": 174
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7933,
+      "step": 175
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9165,
+      "step": 176
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9907,
+      "step": 177
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9265,
+      "step": 178
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8285,
+      "step": 179
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0191,
+      "step": 180
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9367,
+      "step": 181
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8359,
+      "step": 182
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2265,
+      "step": 183
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8823,
+      "step": 184
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9437,
+      "step": 185
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9701,
+      "step": 186
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8259,
+      "step": 187
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9425,
+      "step": 188
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1224,
+      "step": 189
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.882,
+      "step": 190
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8594,
+      "step": 191
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0202,
+      "step": 192
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.9166834354400635,
+      "eval_runtime": 15.8953,
+      "eval_samples_per_second": 0.944,
+      "eval_steps_per_second": 0.252,
+      "step": 192
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8483,
+      "step": 193
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0376,
+      "step": 194
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8855,
+      "step": 195
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1538,
+      "step": 196
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9712,
+      "step": 197
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9211,
+      "step": 198
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0187,
+      "step": 199
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9248,
+      "step": 200
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9905,
+      "step": 201
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0332,
+      "step": 202
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9872,
+      "step": 203
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0161,
+      "step": 204
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0037,
+      "step": 205
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8186,
+      "step": 206
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8858,
+      "step": 207
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.834,
+      "step": 208
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0763,
+      "step": 209
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9497,
+      "step": 210
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 2.023,
+      "step": 211
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8244,
+      "step": 212
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9586,
+      "step": 213
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8326,
+      "step": 214
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0057,
+      "step": 215
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9797,
+      "step": 216
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6863,
+      "step": 217
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7368,
+      "step": 218
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8357,
+      "step": 219
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9472,
+      "step": 220
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9841,
+      "step": 221
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.889,
+      "step": 222
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0772,
+      "step": 223
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9318,
+      "step": 224
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7855,
+      "step": 225
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9101,
+      "step": 226
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0846,
+      "step": 227
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1986,
+      "step": 228
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9374,
+      "step": 229
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.81,
+      "step": 230
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9222,
+      "step": 231
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8392,
+      "step": 232
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0325,
+      "step": 233
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 2.09,
+      "step": 234
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8476,
+      "step": 235
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9211,
+      "step": 236
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8962,
+      "step": 237
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0457,
+      "step": 238
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9181,
+      "step": 239
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8752,
+      "step": 240
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9179,
+      "step": 241
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8738,
+      "step": 242
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8973,
+      "step": 243
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9131,
+      "step": 244
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0874,
+      "step": 245
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0245,
+      "step": 246
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9355,
+      "step": 247
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0169,
+      "step": 248
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8927,
+      "step": 249
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0183,
+      "step": 250
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9835,
+      "step": 251
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9892,
+      "step": 252
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0497,
+      "step": 253
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0267,
+      "step": 254
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.935,
+      "step": 255
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1717,
+      "step": 256
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.9131959676742554,
+      "eval_runtime": 15.8622,
+      "eval_samples_per_second": 0.946,
+      "eval_steps_per_second": 0.252,
+      "step": 256
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9212,
+      "step": 257
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8908,
+      "step": 258
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9697,
+      "step": 259
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7132,
+      "step": 260
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9855,
+      "step": 261
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9295,
+      "step": 262
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8864,
+      "step": 263
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8215,
+      "step": 264
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9387,
+      "step": 265
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0644,
+      "step": 266
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0394,
+      "step": 267
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 2.03,
+      "step": 268
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8564,
+      "step": 269
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8823,
+      "step": 270
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9494,
+      "step": 271
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7425,
+      "step": 272
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8835,
+      "step": 273
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.898,
+      "step": 274
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8898,
+      "step": 275
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8986,
+      "step": 276
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8679,
+      "step": 277
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8186,
+      "step": 278
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9745,
+      "step": 279
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9038,
+      "step": 280
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.833,
+      "step": 281
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9604,
+      "step": 282
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9541,
+      "step": 283
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8268,
+      "step": 284
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7634,
+      "step": 285
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9416,
+      "step": 286
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8609,
+      "step": 287
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0638,
+      "step": 288
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9296,
+      "step": 289
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9795,
+      "step": 290
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8656,
+      "step": 291
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0354,
+      "step": 292
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9194,
+      "step": 293
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9736,
+      "step": 294
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9813,
+      "step": 295
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6839,
+      "step": 296
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9092,
+      "step": 297
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8476,
+      "step": 298
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8747,
+      "step": 299
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1024,
+      "step": 300
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.166,
+      "step": 301
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9582,
+      "step": 302
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9381,
+      "step": 303
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0313,
+      "step": 304
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9334,
+      "step": 305
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9775,
+      "step": 306
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0269,
+      "step": 307
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0337,
+      "step": 308
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0188,
+      "step": 309
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6189,
+      "step": 310
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 2.011,
+      "step": 311
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.84,
+      "step": 312
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8876,
+      "step": 313
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0968,
+      "step": 314
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8737,
+      "step": 315
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9588,
+      "step": 316
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9933,
+      "step": 317
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.958,
+      "step": 318
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0382,
+      "step": 319
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8921,
+      "step": 320
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.910348892211914,
+      "eval_runtime": 15.8255,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 320
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6654,
+      "step": 321
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.91,
+      "step": 322
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9411,
+      "step": 323
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.807,
+      "step": 324
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9477,
+      "step": 325
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8288,
+      "step": 326
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0549,
+      "step": 327
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0891,
+      "step": 328
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8982,
+      "step": 329
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.825,
+      "step": 330
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8956,
+      "step": 331
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0764,
+      "step": 332
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9019,
+      "step": 333
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1776,
+      "step": 334
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0201,
+      "step": 335
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8298,
+      "step": 336
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1339,
+      "step": 337
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0821,
+      "step": 338
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9018,
+      "step": 339
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9474,
+      "step": 340
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8844,
+      "step": 341
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9427,
+      "step": 342
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9446,
+      "step": 343
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8503,
+      "step": 344
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8068,
+      "step": 345
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7826,
+      "step": 346
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8188,
+      "step": 347
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9254,
+      "step": 348
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0352,
+      "step": 349
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1565,
+      "step": 350
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.725,
+      "step": 351
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9508,
+      "step": 352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9927,
+      "step": 353
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8783,
+      "step": 354
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8725,
+      "step": 355
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8017,
+      "step": 356
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9587,
+      "step": 357
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8029,
+      "step": 358
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7235,
+      "step": 359
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8148,
+      "step": 360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6969,
+      "step": 361
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 2.056,
+      "step": 362
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1066,
+      "step": 363
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9098,
+      "step": 364
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7915,
+      "step": 365
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7709,
+      "step": 366
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7265,
+      "step": 367
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0535,
+      "step": 368
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9003,
+      "step": 369
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7629,
+      "step": 370
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8575,
+      "step": 371
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8103,
+      "step": 372
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7453,
+      "step": 373
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9325,
+      "step": 374
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9181,
+      "step": 375
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7825,
+      "step": 376
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8488,
+      "step": 377
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8314,
+      "step": 378
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9141,
+      "step": 379
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8282,
+      "step": 380
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7452,
+      "step": 381
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8025,
+      "step": 382
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0557,
+      "step": 383
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8759,
+      "step": 384
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.908756136894226,
+      "eval_runtime": 15.8254,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 384
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9679,
+      "step": 385
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8212,
+      "step": 386
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9418,
+      "step": 387
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8895,
+      "step": 388
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0485,
+      "step": 389
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9116,
+      "step": 390
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9387,
+      "step": 391
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9902,
+      "step": 392
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8657,
+      "step": 393
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7836,
+      "step": 394
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9202,
+      "step": 395
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9011,
+      "step": 396
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8977,
+      "step": 397
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7779,
+      "step": 398
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8189,
+      "step": 399
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8585,
+      "step": 400
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9763,
+      "step": 401
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7823,
+      "step": 402
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.689,
+      "step": 403
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.915,
+      "step": 404
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7241,
+      "step": 405
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.794,
+      "step": 406
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9473,
+      "step": 407
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.683,
+      "step": 408
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9877,
+      "step": 409
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.83,
+      "step": 410
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8058,
+      "step": 411
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8665,
+      "step": 412
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8601,
+      "step": 413
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.935,
+      "step": 414
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9988,
+      "step": 415
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8722,
+      "step": 416
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.836,
+      "step": 417
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9729,
+      "step": 418
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8294,
+      "step": 419
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7857,
+      "step": 420
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7951,
+      "step": 421
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9943,
+      "step": 422
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8178,
+      "step": 423
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0196,
+      "step": 424
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7755,
+      "step": 425
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7654,
+      "step": 426
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9323,
+      "step": 427
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7688,
+      "step": 428
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.955,
+      "step": 429
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.875,
+      "step": 430
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8321,
+      "step": 431
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8672,
+      "step": 432
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0068,
+      "step": 433
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8938,
+      "step": 434
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7994,
+      "step": 435
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9635,
+      "step": 436
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0295,
+      "step": 437
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9747,
+      "step": 438
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7277,
+      "step": 439
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9286,
+      "step": 440
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9488,
+      "step": 441
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7934,
+      "step": 442
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8474,
+      "step": 443
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0781,
+      "step": 444
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9829,
+      "step": 445
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9607,
+      "step": 446
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7645,
+      "step": 447
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5935,
+      "step": 448
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 1.9090455770492554,
+      "eval_runtime": 15.8424,
+      "eval_samples_per_second": 0.947,
+      "eval_steps_per_second": 0.252,
+      "step": 448
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9635,
+      "step": 449
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7286,
+      "step": 450
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8306,
+      "step": 451
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9006,
+      "step": 452
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 2.058,
+      "step": 453
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1269,
+      "step": 454
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8991,
+      "step": 455
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6794,
+      "step": 456
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9455,
+      "step": 457
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9056,
+      "step": 458
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8372,
+      "step": 459
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9977,
+      "step": 460
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.857,
+      "step": 461
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8302,
+      "step": 462
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9658,
+      "step": 463
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8063,
+      "step": 464
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9346,
+      "step": 465
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8321,
+      "step": 466
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9172,
+      "step": 467
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9369,
+      "step": 468
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8713,
+      "step": 469
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9548,
+      "step": 470
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9509,
+      "step": 471
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7716,
+      "step": 472
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7426,
+      "step": 473
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.029,
+      "step": 474
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8024,
+      "step": 475
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0286,
+      "step": 476
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8011,
+      "step": 477
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7678,
+      "step": 478
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9928,
+      "step": 479
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8499,
+      "step": 480
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6583,
+      "step": 481
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8256,
+      "step": 482
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.855,
+      "step": 483
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0679,
+      "step": 484
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8916,
+      "step": 485
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7041,
+      "step": 486
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2201,
+      "step": 487
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9308,
+      "step": 488
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1227,
+      "step": 489
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8805,
+      "step": 490
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7295,
+      "step": 491
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9037,
+      "step": 492
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8797,
+      "step": 493
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9664,
+      "step": 494
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7967,
+      "step": 495
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9072,
+      "step": 496
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7727,
+      "step": 497
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8644,
+      "step": 498
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8529,
+      "step": 499
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8128,
+      "step": 500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8639,
+      "step": 501
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9974,
+      "step": 502
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9035,
+      "step": 503
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.84,
+      "step": 504
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8928,
+      "step": 505
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7278,
+      "step": 506
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8777,
+      "step": 507
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9465,
+      "step": 508
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8905,
+      "step": 509
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8007,
+      "step": 510
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7495,
+      "step": 511
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8199,
+      "step": 512
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.9047073125839233,
+      "eval_runtime": 15.8349,
+      "eval_samples_per_second": 0.947,
+      "eval_steps_per_second": 0.253,
+      "step": 512
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8661,
+      "step": 513
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9312,
+      "step": 514
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7717,
+      "step": 515
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7454,
+      "step": 516
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9119,
+      "step": 517
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9785,
+      "step": 518
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9029,
+      "step": 519
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7402,
+      "step": 520
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8072,
+      "step": 521
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8136,
+      "step": 522
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8955,
+      "step": 523
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9862,
+      "step": 524
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8566,
+      "step": 525
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9079,
+      "step": 526
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9165,
+      "step": 527
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7425,
+      "step": 528
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.896,
+      "step": 529
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0463,
+      "step": 530
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8926,
+      "step": 531
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9633,
+      "step": 532
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9225,
+      "step": 533
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8044,
+      "step": 534
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8735,
+      "step": 535
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8568,
+      "step": 536
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9984,
+      "step": 537
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8493,
+      "step": 538
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6813,
+      "step": 539
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0485,
+      "step": 540
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9968,
+      "step": 541
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.973,
+      "step": 542
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7108,
+      "step": 543
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 2.141,
+      "step": 544
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9077,
+      "step": 545
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7523,
+      "step": 546
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7261,
+      "step": 547
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.93,
+      "step": 548
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9137,
+      "step": 549
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8861,
+      "step": 550
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7487,
+      "step": 551
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9515,
+      "step": 552
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8471,
+      "step": 553
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8204,
+      "step": 554
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7801,
+      "step": 555
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.973,
+      "step": 556
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0137,
+      "step": 557
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.811,
+      "step": 558
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7863,
+      "step": 559
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8528,
+      "step": 560
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7901,
+      "step": 561
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8324,
+      "step": 562
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8088,
+      "step": 563
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7647,
+      "step": 564
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.995,
+      "step": 565
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9418,
+      "step": 566
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9464,
+      "step": 567
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0446,
+      "step": 568
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9373,
+      "step": 569
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9599,
+      "step": 570
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7897,
+      "step": 571
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7883,
+      "step": 572
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.929,
+      "step": 573
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8037,
+      "step": 574
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0317,
+      "step": 575
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9032,
+      "step": 576
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.904633641242981,
+      "eval_runtime": 15.8817,
+      "eval_samples_per_second": 0.944,
+      "eval_steps_per_second": 0.252,
+      "step": 576
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8677,
+      "step": 577
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8722,
+      "step": 578
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8085,
+      "step": 579
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9854,
+      "step": 580
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9753,
+      "step": 581
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1195,
+      "step": 582
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8101,
+      "step": 583
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9143,
+      "step": 584
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8421,
+      "step": 585
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9934,
+      "step": 586
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.99,
+      "step": 587
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8455,
+      "step": 588
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9165,
+      "step": 589
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9551,
+      "step": 590
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8581,
+      "step": 591
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8379,
+      "step": 592
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0637,
+      "step": 593
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8534,
+      "step": 594
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0634,
+      "step": 595
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0384,
+      "step": 596
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2857,
+      "step": 597
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7737,
+      "step": 598
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7523,
+      "step": 599
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8293,
+      "step": 600
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8713,
+      "step": 601
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8498,
+      "step": 602
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.812,
+      "step": 603
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9033,
+      "step": 604
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8274,
+      "step": 605
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.803,
+      "step": 606
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9733,
+      "step": 607
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9907,
+      "step": 608
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8521,
+      "step": 609
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9774,
+      "step": 610
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7895,
+      "step": 611
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0517,
+      "step": 612
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8392,
+      "step": 613
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8374,
+      "step": 614
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.768,
+      "step": 615
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8617,
+      "step": 616
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0166,
+      "step": 617
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8767,
+      "step": 618
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7613,
+      "step": 619
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.929,
+      "step": 620
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8824,
+      "step": 621
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9365,
+      "step": 622
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7937,
+      "step": 623
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9532,
+      "step": 624
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9121,
+      "step": 625
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8889,
+      "step": 626
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9079,
+      "step": 627
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8388,
+      "step": 628
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8738,
+      "step": 629
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8281,
+      "step": 630
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9145,
+      "step": 631
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8998,
+      "step": 632
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8605,
+      "step": 633
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7396,
+      "step": 634
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9788,
+      "step": 635
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1603,
+      "step": 636
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.94,
+      "step": 637
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9619,
+      "step": 638
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9015,
+      "step": 639
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9345,
+      "step": 640
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 1.901041030883789,
+      "eval_runtime": 15.7651,
+      "eval_samples_per_second": 0.951,
+      "eval_steps_per_second": 0.254,
+      "step": 640
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9378,
+      "step": 641
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6842,
+      "step": 642
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7583,
+      "step": 643
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7447,
+      "step": 644
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 2.055,
+      "step": 645
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9641,
+      "step": 646
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9397,
+      "step": 647
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8886,
+      "step": 648
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8484,
+      "step": 649
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9315,
+      "step": 650
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6379,
+      "step": 651
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8233,
+      "step": 652
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8882,
+      "step": 653
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9685,
+      "step": 654
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.982,
+      "step": 655
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8245,
+      "step": 656
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8107,
+      "step": 657
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.992,
+      "step": 658
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0028,
+      "step": 659
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8672,
+      "step": 660
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8239,
+      "step": 661
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8306,
+      "step": 662
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7309,
+      "step": 663
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.914,
+      "step": 664
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.891,
+      "step": 665
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8386,
+      "step": 666
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7672,
+      "step": 667
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8875,
+      "step": 668
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9328,
+      "step": 669
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9852,
+      "step": 670
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7268,
+      "step": 671
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9357,
+      "step": 672
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9003,
+      "step": 673
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6654,
+      "step": 674
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7439,
+      "step": 675
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9208,
+      "step": 676
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7153,
+      "step": 677
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7507,
+      "step": 678
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7489,
+      "step": 679
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.822,
+      "step": 680
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8108,
+      "step": 681
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6284,
+      "step": 682
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8585,
+      "step": 683
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7957,
+      "step": 684
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0225,
+      "step": 685
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9307,
+      "step": 686
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8515,
+      "step": 687
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8868,
+      "step": 688
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0517,
+      "step": 689
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8739,
+      "step": 690
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.851,
+      "step": 691
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6956,
+      "step": 692
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9364,
+      "step": 693
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9086,
+      "step": 694
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9137,
+      "step": 695
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7439,
+      "step": 696
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9386,
+      "step": 697
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8278,
+      "step": 698
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8963,
+      "step": 699
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7445,
+      "step": 700
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9078,
+      "step": 701
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8682,
+      "step": 702
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9643,
+      "step": 703
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9067,
+      "step": 704
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 1.8956754207611084,
+      "eval_runtime": 15.8172,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 704
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9434,
+      "step": 705
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8199,
+      "step": 706
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7653,
+      "step": 707
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8696,
+      "step": 708
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8837,
+      "step": 709
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6851,
+      "step": 710
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7055,
+      "step": 711
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9095,
+      "step": 712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8649,
+      "step": 713
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7723,
+      "step": 714
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8985,
+      "step": 715
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6597,
+      "step": 716
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7673,
+      "step": 717
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.632,
+      "step": 718
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7904,
+      "step": 719
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7018,
+      "step": 720
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9637,
+      "step": 721
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8436,
+      "step": 722
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.649,
+      "step": 723
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5266,
+      "step": 724
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8776,
+      "step": 725
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8995,
+      "step": 726
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9047,
+      "step": 727
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7088,
+      "step": 728
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6831,
+      "step": 729
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7458,
+      "step": 730
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6726,
+      "step": 731
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.85,
+      "step": 732
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8863,
+      "step": 733
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6919,
+      "step": 734
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8132,
+      "step": 735
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6245,
+      "step": 736
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6661,
+      "step": 737
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5934,
+      "step": 738
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8433,
+      "step": 739
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7394,
+      "step": 740
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.78,
+      "step": 741
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7698,
+      "step": 742
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7158,
+      "step": 743
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6797,
+      "step": 744
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8162,
+      "step": 745
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9739,
+      "step": 746
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5963,
+      "step": 747
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7822,
+      "step": 748
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7986,
+      "step": 749
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6597,
+      "step": 750
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.803,
+      "step": 751
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7836,
+      "step": 752
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7958,
+      "step": 753
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8053,
+      "step": 754
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8482,
+      "step": 755
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7105,
+      "step": 756
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7662,
+      "step": 757
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.981,
+      "step": 758
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7128,
+      "step": 759
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7964,
+      "step": 760
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9118,
+      "step": 761
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6856,
+      "step": 762
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7717,
+      "step": 763
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.92,
+      "step": 764
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7106,
+      "step": 765
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9107,
+      "step": 766
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8034,
+      "step": 767
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8643,
+      "step": 768
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 1.9084140062332153,
+      "eval_runtime": 15.7691,
+      "eval_samples_per_second": 0.951,
+      "eval_steps_per_second": 0.254,
+      "step": 768
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7414,
+      "step": 769
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8657,
+      "step": 770
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7534,
+      "step": 771
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.695,
+      "step": 772
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9271,
+      "step": 773
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7213,
+      "step": 774
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6776,
+      "step": 775
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6942,
+      "step": 776
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8704,
+      "step": 777
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5163,
+      "step": 778
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5997,
+      "step": 779
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5882,
+      "step": 780
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6696,
+      "step": 781
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.665,
+      "step": 782
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8148,
+      "step": 783
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8108,
+      "step": 784
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7828,
+      "step": 785
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8787,
+      "step": 786
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7048,
+      "step": 787
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6589,
+      "step": 788
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8266,
+      "step": 789
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7714,
+      "step": 790
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8702,
+      "step": 791
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6824,
+      "step": 792
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.81,
+      "step": 793
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.751,
+      "step": 794
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8082,
+      "step": 795
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7196,
+      "step": 796
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8456,
+      "step": 797
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6487,
+      "step": 798
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4987,
+      "step": 799
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9665,
+      "step": 800
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8896,
+      "step": 801
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6338,
+      "step": 802
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.894,
+      "step": 803
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7418,
+      "step": 804
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8894,
+      "step": 805
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8593,
+      "step": 806
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7205,
+      "step": 807
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6907,
+      "step": 808
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8852,
+      "step": 809
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.76,
+      "step": 810
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7215,
+      "step": 811
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8215,
+      "step": 812
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8286,
+      "step": 813
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7595,
+      "step": 814
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8129,
+      "step": 815
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7214,
+      "step": 816
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7019,
+      "step": 817
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7345,
+      "step": 818
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7211,
+      "step": 819
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.802,
+      "step": 820
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6679,
+      "step": 821
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7677,
+      "step": 822
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.745,
+      "step": 823
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7046,
+      "step": 824
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7165,
+      "step": 825
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6705,
+      "step": 826
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7855,
+      "step": 827
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7884,
+      "step": 828
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8958,
+      "step": 829
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7122,
+      "step": 830
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 2.022,
+      "step": 831
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8046,
+      "step": 832
+    },
+    {
+      "epoch": 2.35,
+      "eval_loss": 1.9103764295578003,
+      "eval_runtime": 15.7656,
+      "eval_samples_per_second": 0.951,
+      "eval_steps_per_second": 0.254,
+      "step": 832
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6111,
+      "step": 833
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8109,
+      "step": 834
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.703,
+      "step": 835
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7922,
+      "step": 836
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0147,
+      "step": 837
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7194,
+      "step": 838
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9895,
+      "step": 839
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.787,
+      "step": 840
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9435,
+      "step": 841
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.74,
+      "step": 842
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7844,
+      "step": 843
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6824,
+      "step": 844
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9168,
+      "step": 845
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9959,
+      "step": 846
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8155,
+      "step": 847
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6206,
+      "step": 848
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7616,
+      "step": 849
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7026,
+      "step": 850
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7959,
+      "step": 851
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6107,
+      "step": 852
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7275,
+      "step": 853
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6842,
+      "step": 854
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7051,
+      "step": 855
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8819,
+      "step": 856
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7065,
+      "step": 857
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9535,
+      "step": 858
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.755,
+      "step": 859
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7594,
+      "step": 860
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6443,
+      "step": 861
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8298,
+      "step": 862
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8285,
+      "step": 863
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6745,
+      "step": 864
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.689,
+      "step": 865
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9458,
+      "step": 866
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8596,
+      "step": 867
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7854,
+      "step": 868
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9977,
+      "step": 869
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7828,
+      "step": 870
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6712,
+      "step": 871
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6976,
+      "step": 872
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6637,
+      "step": 873
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8234,
+      "step": 874
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7762,
+      "step": 875
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7231,
+      "step": 876
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6682,
+      "step": 877
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8056,
+      "step": 878
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9182,
+      "step": 879
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8499,
+      "step": 880
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8832,
+      "step": 881
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6505,
+      "step": 882
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8629,
+      "step": 883
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7037,
+      "step": 884
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9912,
+      "step": 885
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6883,
+      "step": 886
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9611,
+      "step": 887
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6294,
+      "step": 888
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7118,
+      "step": 889
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9091,
+      "step": 890
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9107,
+      "step": 891
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8334,
+      "step": 892
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.872,
+      "step": 893
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7879,
+      "step": 894
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8291,
+      "step": 895
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5622,
+      "step": 896
+    },
+    {
+      "epoch": 2.53,
+      "eval_loss": 1.9111343622207642,
+      "eval_runtime": 15.8297,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 896
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8108,
+      "step": 897
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8616,
+      "step": 898
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7982,
+      "step": 899
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9036,
+      "step": 900
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8679,
+      "step": 901
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6097,
+      "step": 902
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8929,
+      "step": 903
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8855,
+      "step": 904
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8899,
+      "step": 905
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8021,
+      "step": 906
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8802,
+      "step": 907
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8358,
+      "step": 908
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7239,
+      "step": 909
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8453,
+      "step": 910
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.856,
+      "step": 911
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9009,
+      "step": 912
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7153,
+      "step": 913
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8069,
+      "step": 914
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6698,
+      "step": 915
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8077,
+      "step": 916
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6991,
+      "step": 917
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6358,
+      "step": 918
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8901,
+      "step": 919
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.914,
+      "step": 920
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8759,
+      "step": 921
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0676,
+      "step": 922
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9327,
+      "step": 923
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8873,
+      "step": 924
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8848,
+      "step": 925
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8782,
+      "step": 926
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7306,
+      "step": 927
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7191,
+      "step": 928
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7849,
+      "step": 929
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8355,
+      "step": 930
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7607,
+      "step": 931
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6837,
+      "step": 932
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8278,
+      "step": 933
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.736,
+      "step": 934
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5864,
+      "step": 935
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8364,
+      "step": 936
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7951,
+      "step": 937
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8251,
+      "step": 938
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9624,
+      "step": 939
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8844,
+      "step": 940
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9013,
+      "step": 941
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6256,
+      "step": 942
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9452,
+      "step": 943
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7198,
+      "step": 944
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6005,
+      "step": 945
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7862,
+      "step": 946
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6968,
+      "step": 947
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7225,
+      "step": 948
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7917,
+      "step": 949
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7547,
+      "step": 950
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8442,
+      "step": 951
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6128,
+      "step": 952
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6948,
+      "step": 953
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7104,
+      "step": 954
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8361,
+      "step": 955
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7884,
+      "step": 956
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0423,
+      "step": 957
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8477,
+      "step": 958
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6765,
+      "step": 959
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6571,
+      "step": 960
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 1.9046272039413452,
+      "eval_runtime": 15.821,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 960
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8389,
+      "step": 961
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8361,
+      "step": 962
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7566,
+      "step": 963
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6827,
+      "step": 964
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7623,
+      "step": 965
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.628,
+      "step": 966
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7955,
+      "step": 967
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7576,
+      "step": 968
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6056,
+      "step": 969
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7061,
+      "step": 970
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7672,
+      "step": 971
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7454,
+      "step": 972
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7006,
+      "step": 973
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7268,
+      "step": 974
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.904,
+      "step": 975
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7068,
+      "step": 976
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9171,
+      "step": 977
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6846,
+      "step": 978
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6668,
+      "step": 979
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8157,
+      "step": 980
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7917,
+      "step": 981
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7899,
+      "step": 982
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9337,
+      "step": 983
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6848,
+      "step": 984
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8253,
+      "step": 985
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6049,
+      "step": 986
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7543,
+      "step": 987
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8715,
+      "step": 988
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7701,
+      "step": 989
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8418,
+      "step": 990
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8947,
+      "step": 991
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8245,
+      "step": 992
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7866,
+      "step": 993
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6966,
+      "step": 994
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6865,
+      "step": 995
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8293,
+      "step": 996
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7333,
+      "step": 997
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9818,
+      "step": 998
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9954,
+      "step": 999
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6541,
+      "step": 1000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8485,
+      "step": 1001
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6253,
+      "step": 1002
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8744,
+      "step": 1003
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6481,
+      "step": 1004
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7262,
+      "step": 1005
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7663,
+      "step": 1006
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7832,
+      "step": 1007
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7518,
+      "step": 1008
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5817,
+      "step": 1009
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7278,
+      "step": 1010
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8573,
+      "step": 1011
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8839,
+      "step": 1012
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6467,
+      "step": 1013
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.93,
+      "step": 1014
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6739,
+      "step": 1015
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.685,
+      "step": 1016
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6035,
+      "step": 1017
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8281,
+      "step": 1018
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6309,
+      "step": 1019
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0225,
+      "step": 1020
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.783,
+      "step": 1021
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7345,
+      "step": 1022
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7778,
+      "step": 1023
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.687,
+      "step": 1024
+    },
+    {
+      "epoch": 2.89,
+      "eval_loss": 1.9048281908035278,
+      "eval_runtime": 15.8179,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 1024
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.87,
+      "step": 1025
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5939,
+      "step": 1026
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9544,
+      "step": 1027
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8658,
+      "step": 1028
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0409,
+      "step": 1029
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8278,
+      "step": 1030
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8789,
+      "step": 1031
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6719,
+      "step": 1032
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7059,
+      "step": 1033
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8003,
+      "step": 1034
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8126,
+      "step": 1035
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8602,
+      "step": 1036
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6596,
+      "step": 1037
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7602,
+      "step": 1038
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.664,
+      "step": 1039
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6623,
+      "step": 1040
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8087,
+      "step": 1041
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.638,
+      "step": 1042
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8812,
+      "step": 1043
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8309,
+      "step": 1044
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6749,
+      "step": 1045
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5502,
+      "step": 1046
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6402,
+      "step": 1047
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7163,
+      "step": 1048
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5623,
+      "step": 1049
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.723,
+      "step": 1050
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8435,
+      "step": 1051
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6106,
+      "step": 1052
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8986,
+      "step": 1053
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7712,
+      "step": 1054
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6907,
+      "step": 1055
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6686,
+      "step": 1056
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7947,
+      "step": 1057
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7811,
+      "step": 1058
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7251,
+      "step": 1059
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6823,
+      "step": 1060
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9199,
+      "step": 1061
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9232,
+      "step": 1062
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.624,
+      "step": 1063
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5302,
+      "step": 1064
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6956,
+      "step": 1065
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6959,
+      "step": 1066
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6628,
+      "step": 1067
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5051,
+      "step": 1068
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.655,
+      "step": 1069
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4748,
+      "step": 1070
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8429,
+      "step": 1071
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6748,
+      "step": 1072
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6456,
+      "step": 1073
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7123,
+      "step": 1074
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9159,
+      "step": 1075
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4318,
+      "step": 1076
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6294,
+      "step": 1077
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5642,
+      "step": 1078
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6894,
+      "step": 1079
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.768,
+      "step": 1080
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5161,
+      "step": 1081
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7366,
+      "step": 1082
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7905,
+      "step": 1083
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8867,
+      "step": 1084
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6361,
+      "step": 1085
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4467,
+      "step": 1086
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4672,
+      "step": 1087
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5955,
+      "step": 1088
+    },
+    {
+      "epoch": 3.07,
+      "eval_loss": 1.9349616765975952,
+      "eval_runtime": 15.7354,
+      "eval_samples_per_second": 0.953,
+      "eval_steps_per_second": 0.254,
+      "step": 1088
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6836,
+      "step": 1089
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5363,
+      "step": 1090
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6366,
+      "step": 1091
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.584,
+      "step": 1092
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6099,
+      "step": 1093
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6588,
+      "step": 1094
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5309,
+      "step": 1095
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4058,
+      "step": 1096
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5406,
+      "step": 1097
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6967,
+      "step": 1098
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5394,
+      "step": 1099
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6335,
+      "step": 1100
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5532,
+      "step": 1101
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7605,
+      "step": 1102
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7374,
+      "step": 1103
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8527,
+      "step": 1104
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6454,
+      "step": 1105
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7072,
+      "step": 1106
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8284,
+      "step": 1107
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7174,
+      "step": 1108
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7427,
+      "step": 1109
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7253,
+      "step": 1110
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6433,
+      "step": 1111
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.471,
+      "step": 1112
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6554,
+      "step": 1113
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6282,
+      "step": 1114
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5739,
+      "step": 1115
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5.5e-05,
+      "loss": 2.015,
+      "step": 1116
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6217,
+      "step": 1117
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7584,
+      "step": 1118
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5836,
+      "step": 1119
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6397,
+      "step": 1120
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6218,
+      "step": 1121
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7019,
+      "step": 1122
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8836,
+      "step": 1123
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6691,
+      "step": 1124
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5246,
+      "step": 1125
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6446,
+      "step": 1126
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.521,
+      "step": 1127
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5843,
+      "step": 1128
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9895,
+      "step": 1129
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4804,
+      "step": 1130
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4273,
+      "step": 1131
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6483,
+      "step": 1132
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5402,
+      "step": 1133
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6839,
+      "step": 1134
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7466,
+      "step": 1135
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7803,
+      "step": 1136
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6396,
+      "step": 1137
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6403,
+      "step": 1138
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4621,
+      "step": 1139
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3392,
+      "step": 1140
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8388,
+      "step": 1141
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4276,
+      "step": 1142
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6452,
+      "step": 1143
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6529,
+      "step": 1144
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4917,
+      "step": 1145
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4027,
+      "step": 1146
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.678,
+      "step": 1147
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9175,
+      "step": 1148
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6452,
+      "step": 1149
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6226,
+      "step": 1150
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7327,
+      "step": 1151
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7397,
+      "step": 1152
+    },
+    {
+      "epoch": 3.25,
+      "eval_loss": 1.9380393028259277,
+      "eval_runtime": 15.7397,
+      "eval_samples_per_second": 0.953,
+      "eval_steps_per_second": 0.254,
+      "step": 1152
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6372,
+      "step": 1153
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.581,
+      "step": 1154
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7269,
+      "step": 1155
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8363,
+      "step": 1156
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7411,
+      "step": 1157
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4259,
+      "step": 1158
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6916,
+      "step": 1159
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3676,
+      "step": 1160
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4567,
+      "step": 1161
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7517,
+      "step": 1162
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.71,
+      "step": 1163
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4534,
+      "step": 1164
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.494,
+      "step": 1165
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4638,
+      "step": 1166
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5161,
+      "step": 1167
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5659,
+      "step": 1168
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6808,
+      "step": 1169
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.652,
+      "step": 1170
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7093,
+      "step": 1171
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5244,
+      "step": 1172
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5643,
+      "step": 1173
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5191,
+      "step": 1174
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4835,
+      "step": 1175
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6005,
+      "step": 1176
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5714,
+      "step": 1177
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9033,
+      "step": 1178
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7614,
+      "step": 1179
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.449,
+      "step": 1180
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5377,
+      "step": 1181
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5162,
+      "step": 1182
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6342,
+      "step": 1183
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5665,
+      "step": 1184
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4881,
+      "step": 1185
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7669,
+      "step": 1186
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6509,
+      "step": 1187
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.667,
+      "step": 1188
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6851,
+      "step": 1189
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4165,
+      "step": 1190
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.654,
+      "step": 1191
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8257,
+      "step": 1192
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6053,
+      "step": 1193
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3193,
+      "step": 1194
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.606,
+      "step": 1195
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8048,
+      "step": 1196
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6862,
+      "step": 1197
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7028,
+      "step": 1198
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6336,
+      "step": 1199
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6121,
+      "step": 1200
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6646,
+      "step": 1201
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5131,
+      "step": 1202
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6493,
+      "step": 1203
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6279,
+      "step": 1204
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5402,
+      "step": 1205
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5394,
+      "step": 1206
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5876,
+      "step": 1207
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5789,
+      "step": 1208
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4994,
+      "step": 1209
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6922,
+      "step": 1210
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7771,
+      "step": 1211
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6653,
+      "step": 1212
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7212,
+      "step": 1213
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5799,
+      "step": 1214
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8464,
+      "step": 1215
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4639,
+      "step": 1216
+    },
+    {
+      "epoch": 3.44,
+      "eval_loss": 1.9253300428390503,
+      "eval_runtime": 15.7681,
+      "eval_samples_per_second": 0.951,
+      "eval_steps_per_second": 0.254,
+      "step": 1216
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4409,
+      "step": 1217
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7769,
+      "step": 1218
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6987,
+      "step": 1219
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.595,
+      "step": 1220
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4972,
+      "step": 1221
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6825,
+      "step": 1222
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6901,
+      "step": 1223
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.2604,
+      "step": 1224
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6313,
+      "step": 1225
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5142,
+      "step": 1226
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6738,
+      "step": 1227
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5811,
+      "step": 1228
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7565,
+      "step": 1229
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4787,
+      "step": 1230
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6695,
+      "step": 1231
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6173,
+      "step": 1232
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7813,
+      "step": 1233
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5165,
+      "step": 1234
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5299,
+      "step": 1235
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6262,
+      "step": 1236
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7077,
+      "step": 1237
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5139,
+      "step": 1238
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6732,
+      "step": 1239
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6222,
+      "step": 1240
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5087,
+      "step": 1241
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4548,
+      "step": 1242
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4974,
+      "step": 1243
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6454,
+      "step": 1244
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5739,
+      "step": 1245
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6256,
+      "step": 1246
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6639,
+      "step": 1247
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5513,
+      "step": 1248
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8193,
+      "step": 1249
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5447,
+      "step": 1250
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5471,
+      "step": 1251
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5737,
+      "step": 1252
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7482,
+      "step": 1253
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5658,
+      "step": 1254
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7131,
+      "step": 1255
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6725,
+      "step": 1256
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5497,
+      "step": 1257
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8631,
+      "step": 1258
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5978,
+      "step": 1259
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6475,
+      "step": 1260
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7079,
+      "step": 1261
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6711,
+      "step": 1262
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4475,
+      "step": 1263
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6583,
+      "step": 1264
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5936,
+      "step": 1265
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8351,
+      "step": 1266
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6518,
+      "step": 1267
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6995,
+      "step": 1268
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6543,
+      "step": 1269
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4819,
+      "step": 1270
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8292,
+      "step": 1271
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5661,
+      "step": 1272
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7389,
+      "step": 1273
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4943,
+      "step": 1274
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7641,
+      "step": 1275
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.665,
+      "step": 1276
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.974,
+      "step": 1277
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6404,
+      "step": 1278
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5531,
+      "step": 1279
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.561,
+      "step": 1280
+    },
+    {
+      "epoch": 3.62,
+      "eval_loss": 1.920297384262085,
+      "eval_runtime": 15.7579,
+      "eval_samples_per_second": 0.952,
+      "eval_steps_per_second": 0.254,
+      "step": 1280
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6378,
+      "step": 1281
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5811,
+      "step": 1282
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9128,
+      "step": 1283
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3653,
+      "step": 1284
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.706,
+      "step": 1285
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6308,
+      "step": 1286
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6404,
+      "step": 1287
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5833,
+      "step": 1288
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5877,
+      "step": 1289
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4376,
+      "step": 1290
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9683,
+      "step": 1291
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6446,
+      "step": 1292
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7979,
+      "step": 1293
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6559,
+      "step": 1294
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6149,
+      "step": 1295
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6367,
+      "step": 1296
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5529,
+      "step": 1297
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6524,
+      "step": 1298
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6392,
+      "step": 1299
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5092,
+      "step": 1300
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6769,
+      "step": 1301
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6808,
+      "step": 1302
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.745,
+      "step": 1303
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8714,
+      "step": 1304
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6815,
+      "step": 1305
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5297,
+      "step": 1306
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6942,
+      "step": 1307
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5466,
+      "step": 1308
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.796,
+      "step": 1309
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.842,
+      "step": 1310
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5013,
+      "step": 1311
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6676,
+      "step": 1312
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5002,
+      "step": 1313
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8792,
+      "step": 1314
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5142,
+      "step": 1315
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3362,
+      "step": 1316
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5927,
+      "step": 1317
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7502,
+      "step": 1318
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7012,
+      "step": 1319
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4472,
+      "step": 1320
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5482,
+      "step": 1321
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6584,
+      "step": 1322
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7355,
+      "step": 1323
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5272,
+      "step": 1324
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6538,
+      "step": 1325
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6033,
+      "step": 1326
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3744,
+      "step": 1327
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7475,
+      "step": 1328
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6737,
+      "step": 1329
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5556,
+      "step": 1330
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6193,
+      "step": 1331
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7167,
+      "step": 1332
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.622,
+      "step": 1333
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5722,
+      "step": 1334
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7205,
+      "step": 1335
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.81,
+      "step": 1336
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.672,
+      "step": 1337
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5235,
+      "step": 1338
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4878,
+      "step": 1339
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6181,
+      "step": 1340
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6526,
+      "step": 1341
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4841,
+      "step": 1342
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7102,
+      "step": 1343
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8727,
+      "step": 1344
+    },
+    {
+      "epoch": 3.8,
+      "eval_loss": 1.93291437625885,
+      "eval_runtime": 15.7409,
+      "eval_samples_per_second": 0.953,
+      "eval_steps_per_second": 0.254,
+      "step": 1344
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5924,
+      "step": 1345
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5196,
+      "step": 1346
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8256,
+      "step": 1347
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6012,
+      "step": 1348
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7743,
+      "step": 1349
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6774,
+      "step": 1350
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6101,
+      "step": 1351
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5228,
+      "step": 1352
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7871,
+      "step": 1353
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7288,
+      "step": 1354
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7653,
+      "step": 1355
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5456,
+      "step": 1356
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3417,
+      "step": 1357
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6922,
+      "step": 1358
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6126,
+      "step": 1359
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5273,
+      "step": 1360
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5465,
+      "step": 1361
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7021,
+      "step": 1362
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5739,
+      "step": 1363
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5776,
+      "step": 1364
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4621,
+      "step": 1365
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6688,
+      "step": 1366
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6526,
+      "step": 1367
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8029,
+      "step": 1368
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5807,
+      "step": 1369
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5816,
+      "step": 1370
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8104,
+      "step": 1371
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4314,
+      "step": 1372
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6695,
+      "step": 1373
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5866,
+      "step": 1374
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9254,
+      "step": 1375
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9509,
+      "step": 1376
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6422,
+      "step": 1377
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8006,
+      "step": 1378
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5843,
+      "step": 1379
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5775,
+      "step": 1380
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9611,
+      "step": 1381
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7163,
+      "step": 1382
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.627,
+      "step": 1383
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7201,
+      "step": 1384
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9958,
+      "step": 1385
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4505,
+      "step": 1386
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4589,
+      "step": 1387
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6366,
+      "step": 1388
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6653,
+      "step": 1389
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8071,
+      "step": 1390
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7386,
+      "step": 1391
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5111,
+      "step": 1392
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5644,
+      "step": 1393
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4966,
+      "step": 1394
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5706,
+      "step": 1395
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.487,
+      "step": 1396
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.723,
+      "step": 1397
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6159,
+      "step": 1398
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6758,
+      "step": 1399
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.661,
+      "step": 1400
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6187,
+      "step": 1401
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6006,
+      "step": 1402
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5472,
+      "step": 1403
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7232,
+      "step": 1404
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7497,
+      "step": 1405
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6407,
+      "step": 1406
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6974,
+      "step": 1407
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6312,
+      "step": 1408
+    },
+    {
+      "epoch": 3.98,
+      "eval_loss": 1.9206175804138184,
+      "eval_runtime": 15.8167,
+      "eval_samples_per_second": 0.948,
+      "eval_steps_per_second": 0.253,
+      "step": 1408
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1416,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 64,
+  "total_flos": 2.0148451966546084e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}