Upload 7 files

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b8f3024a5d9c58ca88d63d828022c4a183ab4cd50126bd8886808b65cfcdcdfc
+oid sha256:153c42d5c27126006704d4feca531d6a691f60a91624225e1c9e395801fbee41
 size 40036040

checkpoint-496_adapter_config.json

@@ -0,0 +1,28 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "yentinglin/Taiwan-LLM-7B-v2.0-chat",
+  "bias": "none",
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 4,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "o_proj",
+    "k_proj",
+    "up_proj",
+    "v_proj",
+    "gate_proj",
+    "q_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-496_trainer_state.json

@@ -0,0 +1,3155 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9375,
+  "eval_steps": 25,
+  "global_step": 496,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0,
+      "loss": 4.0553,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 4.052073955535889,
+      "eval_runtime": 57.3905,
+      "eval_samples_per_second": 4.356,
+      "eval_steps_per_second": 4.356,
+      "step": 1
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 4.3696,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4e-05,
+      "loss": 4.2401,
+      "step": 3
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6e-05,
+      "loss": 4.3077,
+      "step": 4
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8e-05,
+      "loss": 4.0741,
+      "step": 5
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001,
+      "loss": 4.1673,
+      "step": 6
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00012,
+      "loss": 3.9378,
+      "step": 7
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00014,
+      "loss": 4.0167,
+      "step": 8
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00016,
+      "loss": 3.5818,
+      "step": 9
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00018,
+      "loss": 3.9022,
+      "step": 10
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002,
+      "loss": 3.7906,
+      "step": 11
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001999979107245606,
+      "loss": 3.6776,
+      "step": 12
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019999164298554375,
+      "loss": 3.4291,
+      "step": 13
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019998119704485014,
+      "loss": 3.6101,
+      "step": 14
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019996657333896877,
+      "loss": 3.5537,
+      "step": 15
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019994777247895855,
+      "loss": 3.386,
+      "step": 16
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019992479525042303,
+      "loss": 3.6359,
+      "step": 17
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019989764261347736,
+      "loss": 3.6132,
+      "step": 18
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019986631570270832,
+      "loss": 3.4794,
+      "step": 19
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019983081582712685,
+      "loss": 3.2395,
+      "step": 20
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019979114447011323,
+      "loss": 3.2974,
+      "step": 21
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019974730328935534,
+      "loss": 3.2698,
+      "step": 22
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001996992941167792,
+      "loss": 3.3091,
+      "step": 23
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019964711895847258,
+      "loss": 3.2017,
+      "step": 24
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019959077999460095,
+      "loss": 3.2469,
+      "step": 25
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 3.0720062255859375,
+      "eval_runtime": 61.8274,
+      "eval_samples_per_second": 4.044,
+      "eval_steps_per_second": 4.044,
+      "step": 25
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019953027957931658,
+      "loss": 3.1102,
+      "step": 26
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019946562024066014,
+      "loss": 3.1641,
+      "step": 27
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019939680468045499,
+      "loss": 2.9779,
+      "step": 28
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019932383577419432,
+      "loss": 3.104,
+      "step": 29
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019924671657092096,
+      "loss": 2.7981,
+      "step": 30
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019916545029310012,
+      "loss": 2.7673,
+      "step": 31
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019908004033648453,
+      "loss": 2.8478,
+      "step": 32
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019899049026997272,
+      "loss": 2.7539,
+      "step": 33
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019889680383545973,
+      "loss": 2.726,
+      "step": 34
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019879898494768093,
+      "loss": 2.81,
+      "step": 35
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019869703769404828,
+      "loss": 2.6177,
+      "step": 36
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019859096633447965,
+      "loss": 2.6046,
+      "step": 37
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 2.5004,
+      "step": 38
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001983664691986601,
+      "loss": 2.4993,
+      "step": 39
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019824805280313623,
+      "loss": 2.2775,
+      "step": 40
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019812553106273847,
+      "loss": 2.3306,
+      "step": 41
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019799890909710013,
+      "loss": 2.236,
+      "step": 42
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019786819219718443,
+      "loss": 2.2798,
+      "step": 43
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001977333858250636,
+      "loss": 2.2178,
+      "step": 44
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019759449561369038,
+      "loss": 1.9915,
+      "step": 45
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019745152736666302,
+      "loss": 2.3066,
+      "step": 46
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019730448705798239,
+      "loss": 2.1409,
+      "step": 47
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019715338083180269,
+      "loss": 1.9976,
+      "step": 48
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019699821500217434,
+      "loss": 2.1843,
+      "step": 49
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001968389960527806,
+      "loss": 2.1539,
+      "step": 50
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.9486372470855713,
+      "eval_runtime": 61.7944,
+      "eval_samples_per_second": 4.046,
+      "eval_steps_per_second": 4.046,
+      "step": 50
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001966757306366662,
+      "loss": 2.2096,
+      "step": 51
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019650842557595967,
+      "loss": 2.0241,
+      "step": 52
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019633708786158806,
+      "loss": 1.8814,
+      "step": 53
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019616172465298492,
+      "loss": 1.9953,
+      "step": 54
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019598234327779118,
+      "loss": 1.9807,
+      "step": 55
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001957989512315489,
+      "loss": 1.9743,
+      "step": 56
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019561155617738797,
+      "loss": 1.867,
+      "step": 57
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019542016594570615,
+      "loss": 1.8289,
+      "step": 58
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019522478853384155,
+      "loss": 1.9712,
+      "step": 59
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019502543210573881,
+      "loss": 1.6946,
+      "step": 60
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019482210499160765,
+      "loss": 1.8674,
+      "step": 61
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019461481568757506,
+      "loss": 1.6849,
+      "step": 62
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019440357285533,
+      "loss": 1.7003,
+      "step": 63
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019418838532176173,
+      "loss": 1.7021,
+      "step": 64
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 1.8302,
+      "step": 65
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001937462122819935,
+      "loss": 1.8549,
+      "step": 66
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000193519245252219,
+      "loss": 1.8946,
+      "step": 67
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001932883704732001,
+      "loss": 1.799,
+      "step": 68
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019305359759215685,
+      "loss": 1.7541,
+      "step": 69
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019281493641919368,
+      "loss": 1.7145,
+      "step": 70
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019257239692688907,
+      "loss": 1.8254,
+      "step": 71
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00019232598924987903,
+      "loss": 1.6161,
+      "step": 72
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019207572368443385,
+      "loss": 1.5698,
+      "step": 73
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00019182161068802741,
+      "loss": 1.8155,
+      "step": 74
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001915636608789006,
+      "loss": 1.6997,
+      "step": 75
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.6153544187545776,
+      "eval_runtime": 61.8069,
+      "eval_samples_per_second": 4.045,
+      "eval_steps_per_second": 4.045,
+      "step": 75
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00019130188503561741,
+      "loss": 1.6507,
+      "step": 76
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001910362940966147,
+      "loss": 1.6035,
+      "step": 77
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019076689915974488,
+      "loss": 1.4493,
+      "step": 78
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00019049371148181253,
+      "loss": 1.5494,
+      "step": 79
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001902167424781038,
+      "loss": 1.5386,
+      "step": 80
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00018993600372190932,
+      "loss": 1.467,
+      "step": 81
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00018965150694404094,
+      "loss": 1.6658,
+      "step": 82
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00018936326403234125,
+      "loss": 1.6175,
+      "step": 83
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00018907128703118695,
+      "loss": 1.6551,
+      "step": 84
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00018877558814098561,
+      "loss": 1.409,
+      "step": 85
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00018847617971766577,
+      "loss": 1.4879,
+      "step": 86
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001881730742721608,
+      "loss": 1.3067,
+      "step": 87
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00018786628446988593,
+      "loss": 1.6869,
+      "step": 88
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001875558231302091,
+      "loss": 1.5962,
+      "step": 89
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018724170322591537,
+      "loss": 1.3352,
+      "step": 90
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018692393788266479,
+      "loss": 1.5016,
+      "step": 91
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 1.6951,
+      "step": 92
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00018627752414301086,
+      "loss": 1.5388,
+      "step": 93
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00018594890275733453,
+      "loss": 1.3656,
+      "step": 94
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00018561668995302667,
+      "loss": 1.4599,
+      "step": 95
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001852808996117683,
+      "loss": 1.3428,
+      "step": 96
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00018494154576472976,
+      "loss": 1.2147,
+      "step": 97
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001845986425919841,
+      "loss": 1.4761,
+      "step": 98
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00018425220442191495,
+      "loss": 1.4571,
+      "step": 99
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00018390224573061747,
+      "loss": 1.5076,
+      "step": 100
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.49380362033844,
+      "eval_runtime": 61.7975,
+      "eval_samples_per_second": 4.045,
+      "eval_steps_per_second": 4.045,
+      "step": 100
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00018354878114129367,
+      "loss": 1.3776,
+      "step": 101
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00018319182542364117,
+      "loss": 1.437,
+      "step": 102
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00018283139349323634,
+      "loss": 1.3558,
+      "step": 103
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001824675004109107,
+      "loss": 1.4502,
+      "step": 104
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00018210016138212187,
+      "loss": 1.3451,
+      "step": 105
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00018172939175631808,
+      "loss": 1.403,
+      "step": 106
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00018135520702629675,
+      "loss": 1.1523,
+      "step": 107
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00018097762282755727,
+      "loss": 1.2302,
+      "step": 108
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00018059665493764743,
+      "loss": 1.4151,
+      "step": 109
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001802123192755044,
+      "loss": 1.4853,
+      "step": 110
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001798246319007893,
+      "loss": 1.2571,
+      "step": 111
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001794336090132164,
+      "loss": 1.3594,
+      "step": 112
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00017903926695187595,
+      "loss": 1.3997,
+      "step": 113
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00017864162219455163,
+      "loss": 1.5718,
+      "step": 114
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00017824069135703198,
+      "loss": 1.314,
+      "step": 115
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00017783649119241602,
+      "loss": 1.3595,
+      "step": 116
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00017742903859041325,
+      "loss": 1.4467,
+      "step": 117
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00017701835057663798,
+      "loss": 1.2837,
+      "step": 118
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 1.2669,
+      "step": 119
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00017618733709147662,
+      "loss": 1.3115,
+      "step": 120
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001757670463444118,
+      "loss": 1.1836,
+      "step": 121
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00017534358963276607,
+      "loss": 1.6092,
+      "step": 122
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00017491698465089362,
+      "loss": 1.2288,
+      "step": 123
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00017448724922470058,
+      "loss": 1.2305,
+      "step": 124
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00017405440131090048,
+      "loss": 1.3794,
+      "step": 125
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 1.4039632081985474,
+      "eval_runtime": 61.7879,
+      "eval_samples_per_second": 4.046,
+      "eval_steps_per_second": 4.046,
+      "step": 125
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00017361845899626355,
+      "loss": 1.2283,
+      "step": 126
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00017317944049686124,
+      "loss": 1.2627,
+      "step": 127
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00017273736415730488,
+      "loss": 1.2911,
+      "step": 128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00017229224844997928,
+      "loss": 1.1225,
+      "step": 129
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00017184411197427077,
+      "loss": 1.3,
+      "step": 130
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00017139297345578994,
+      "loss": 1.1852,
+      "step": 131
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001709388517455893,
+      "loss": 1.1986,
+      "step": 132
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00017048176581937563,
+      "loss": 1.3396,
+      "step": 133
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00017002173477671686,
+      "loss": 1.3584,
+      "step": 134
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001695587778402442,
+      "loss": 1.2763,
+      "step": 135
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001690929143548488,
+      "loss": 1.3017,
+      "step": 136
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001686241637868734,
+      "loss": 1.2383,
+      "step": 137
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00016815254572329896,
+      "loss": 1.238,
+      "step": 138
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00016767807987092621,
+      "loss": 1.2702,
+      "step": 139
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00016720078605555224,
+      "loss": 1.3573,
+      "step": 140
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00016672068422114196,
+      "loss": 1.2863,
+      "step": 141
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001662377944289948,
+      "loss": 1.2192,
+      "step": 142
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001657521368569064,
+      "loss": 1.1934,
+      "step": 143
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00016526373179832546,
+      "loss": 1.1558,
+      "step": 144
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00016477259966150588,
+      "loss": 1.3684,
+      "step": 145
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 1.5386,
+      "step": 146
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0001637822363550706,
+      "loss": 1.2404,
+      "step": 147
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00016328304656828952,
+      "loss": 1.2416,
+      "step": 148
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00016278121246720987,
+      "loss": 1.3613,
+      "step": 149
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00016227675502122492,
+      "loss": 1.2755,
+      "step": 150
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.3754363059997559,
+      "eval_runtime": 61.7475,
+      "eval_samples_per_second": 4.049,
+      "eval_steps_per_second": 4.049,
+      "step": 150
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00016176969530934572,
+      "loss": 1.3912,
+      "step": 151
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001612600545193203,
+      "loss": 1.2778,
+      "step": 152
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00016074785394674837,
+      "loss": 1.2326,
+      "step": 153
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001602331149941915,
+      "loss": 1.2526,
+      "step": 154
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00015971585917027862,
+      "loss": 1.1176,
+      "step": 155
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001591961080888076,
+      "loss": 1.1617,
+      "step": 156
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001586738834678418,
+      "loss": 1.2921,
+      "step": 157
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00015814920712880267,
+      "loss": 1.3704,
+      "step": 158
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00015762210099555803,
+      "loss": 1.1412,
+      "step": 159
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00015709258709350582,
+      "loss": 1.1516,
+      "step": 160
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00015656068754865387,
+      "loss": 1.3323,
+      "step": 161
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00015602642458669528,
+      "loss": 1.2082,
+      "step": 162
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001554898205320797,
+      "loss": 1.1952,
+      "step": 163
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001549508978070806,
+      "loss": 1.172,
+      "step": 164
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00015440967893085828,
+      "loss": 1.3049,
+      "step": 165
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001538661865185188,
+      "loss": 1.1573,
+      "step": 166
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00015332044328016914,
+      "loss": 1.3122,
+      "step": 167
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001527724720199682,
+      "loss": 1.2487,
+      "step": 168
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00015222229563517385,
+      "loss": 1.2054,
+      "step": 169
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00015166993711518631,
+      "loss": 1.1871,
+      "step": 170
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00015111541954058734,
+      "loss": 1.1853,
+      "step": 171
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001505587660821759,
+      "loss": 1.2743,
+      "step": 172
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.1493,
+      "step": 173
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00014943914464238468,
+      "loss": 1.2354,
+      "step": 174
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00014887622344495643,
+      "loss": 1.1913,
+      "step": 175
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.3428610563278198,
+      "eval_runtime": 61.7873,
+      "eval_samples_per_second": 4.046,
+      "eval_steps_per_second": 4.046,
+      "step": 175
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00014831125992966385,
+      "loss": 1.2415,
+      "step": 176
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001477442777037949,
+      "loss": 1.1343,
+      "step": 177
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00014717530045899038,
+      "loss": 1.325,
+      "step": 178
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001466043519702539,
+      "loss": 1.3118,
+      "step": 179
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001460314560949586,
+      "loss": 1.2941,
+      "step": 180
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00014545663677185006,
+      "loss": 1.1848,
+      "step": 181
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00014487991802004623,
+      "loss": 1.1337,
+      "step": 182
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00014430132393803352,
+      "loss": 1.1494,
+      "step": 183
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001437208787026601,
+      "loss": 1.2076,
+      "step": 184
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00014313860656812536,
+      "loss": 1.1794,
+      "step": 185
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00014255453186496673,
+      "loss": 1.2402,
+      "step": 186
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001419686789990429,
+      "loss": 1.2741,
+      "step": 187
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00014138107245051392,
+      "loss": 1.2486,
+      "step": 188
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00014079173677281837,
+      "loss": 1.2796,
+      "step": 189
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0001402006965916474,
+      "loss": 1.0193,
+      "step": 190
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001396079766039157,
+      "loss": 1.1447,
+      "step": 191
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001390136015767295,
+      "loss": 1.0318,
+      "step": 192
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00013841759634635178,
+      "loss": 1.1047,
+      "step": 193
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00013781998581716427,
+      "loss": 1.1895,
+      "step": 194
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00013722079496062702,
+      "loss": 1.1119,
+      "step": 195
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0001366200488142348,
+      "loss": 1.0248,
+      "step": 196
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013601777248047105,
+      "loss": 1.0961,
+      "step": 197
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013541399112575878,
+      "loss": 1.2622,
+      "step": 198
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013480872997940905,
+      "loss": 1.1858,
+      "step": 199
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 1.1458,
+      "step": 200
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 1.3336366415023804,
+      "eval_runtime": 61.796,
+      "eval_samples_per_second": 4.046,
+      "eval_steps_per_second": 4.046,
+      "step": 200
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013359386953715421,
+      "loss": 1.3068,
+      "step": 201
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013298432100481079,
+      "loss": 1.0935,
+      "step": 202
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013237339420583212,
+      "loss": 1.1904,
+      "step": 203
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013176111466810532,
+      "loss": 1.1859,
+      "step": 204
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013114750797604247,
+      "loss": 1.1536,
+      "step": 205
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00013053259976951133,
+      "loss": 1.2067,
+      "step": 206
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012991641574276418,
+      "loss": 1.2649,
+      "step": 207
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012991641574276418,
+      "loss": 1.2189,
+      "step": 208
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012929898164336408,
+      "loss": 1.301,
+      "step": 209
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012868032327110904,
+      "loss": 1.1506,
+      "step": 210
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0001280604664769539,
+      "loss": 1.1964,
+      "step": 211
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012743943716193016,
+      "loss": 1.139,
+      "step": 212
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00012681726127606376,
+      "loss": 1.2197,
+      "step": 213
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001261939648172906,
+      "loss": 1.1713,
+      "step": 214
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0001255695738303704,
+      "loss": 1.2533,
+      "step": 215
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00012494411440579814,
+      "loss": 1.1755,
+      "step": 216
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00012431761267871417,
+      "loss": 0.9834,
+      "step": 217
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00012369009482781192,
+      "loss": 1.1805,
+      "step": 218
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00012306158707424403,
+      "loss": 1.0295,
+      "step": 219
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00012243211568052677,
+      "loss": 1.0861,
+      "step": 220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0001218017069494426,
+      "loss": 1.3924,
+      "step": 221
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0001211703872229411,
+      "loss": 1.1034,
+      "step": 222
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001205381828810382,
+      "loss": 1.0464,
+      "step": 223
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00011990512034071406,
+      "loss": 1.1107,
+      "step": 224
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00011927122605480898,
+      "loss": 1.1568,
+      "step": 225
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 1.31128990650177,
+      "eval_runtime": 61.808,
+      "eval_samples_per_second": 4.045,
+      "eval_steps_per_second": 4.045,
+      "step": 225
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00011863652651091823,
+      "loss": 1.0532,
+      "step": 226
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00011800104823028515,
+      "loss": 1.0956,
+      "step": 227
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 1.2576,
+      "step": 228
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00011672786170535552,
+      "loss": 1.081,
+      "step": 229
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00011609020666180575,
+      "loss": 1.097,
+      "step": 230
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00011545187928078406,
+      "loss": 1.1571,
+      "step": 231
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0001148129062351249,
+      "loss": 1.3143,
+      "step": 232
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00011417331422464205,
+      "loss": 1.2877,
+      "step": 233
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00011353312997501313,
+      "loss": 1.1032,
+      "step": 234
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00011289238023666266,
+      "loss": 1.1401,
+      "step": 235
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00011225109178364455,
+      "loss": 1.2596,
+      "step": 236
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00011160929141252303,
+      "loss": 1.1803,
+      "step": 237
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00011096700594125318,
+      "loss": 1.0706,
+      "step": 238
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00011032426220806018,
+      "loss": 1.1804,
+      "step": 239
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00010968108707031792,
+      "loss": 1.2334,
+      "step": 240
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00010903750740342682,
+      "loss": 1.1523,
+      "step": 241
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00010839355009969068,
+      "loss": 1.1336,
+      "step": 242
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0001077492420671931,
+      "loss": 1.1287,
+      "step": 243
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00010710461022867302,
+      "loss": 1.181,
+      "step": 244
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0001064596815203998,
+      "loss": 1.2882,
+      "step": 245
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010581448289104758,
+      "loss": 1.0872,
+      "step": 246
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010516904130056946,
+      "loss": 1.1117,
+      "step": 247
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010452338371907064,
+      "loss": 1.1874,
+      "step": 248
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0001038775371256817,
+      "loss": 1.0474,
+      "step": 249
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010323152850743107,
+      "loss": 1.1845,
+      "step": 250
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 1.2912580966949463,
+      "eval_runtime": 61.8585,
+      "eval_samples_per_second": 4.041,
+      "eval_steps_per_second": 4.041,
+      "step": 250
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00010258538485811765,
+      "loss": 1.3517,
+      "step": 251
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00010193913317718244,
+      "loss": 1.2637,
+      "step": 252
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00010129280046858086,
+      "loss": 1.1032,
+      "step": 253
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00010064641373965393,
+      "loss": 0.9574,
+      "step": 254
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0001,
+      "loss": 0.8302,
+      "step": 255
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.935358626034606e-05,
+      "loss": 1.1663,
+      "step": 256
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.870719953141917e-05,
+      "loss": 1.0611,
+      "step": 257
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 9.806086682281758e-05,
+      "loss": 1.1039,
+      "step": 258
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 9.741461514188242e-05,
+      "loss": 1.134,
+      "step": 259
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.676847149256895e-05,
+      "loss": 0.8512,
+      "step": 260
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.612246287431831e-05,
+      "loss": 1.01,
+      "step": 261
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.547661628092937e-05,
+      "loss": 1.1625,
+      "step": 262
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.483095869943055e-05,
+      "loss": 1.0704,
+      "step": 263
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.418551710895243e-05,
+      "loss": 0.8177,
+      "step": 264
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.354031847960022e-05,
+      "loss": 0.9913,
+      "step": 265
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.289538977132703e-05,
+      "loss": 1.166,
+      "step": 266
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.225075793280692e-05,
+      "loss": 1.125,
+      "step": 267
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.160644990030931e-05,
+      "loss": 0.9657,
+      "step": 268
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.09624925965732e-05,
+      "loss": 1.0344,
+      "step": 269
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.03189129296821e-05,
+      "loss": 1.0579,
+      "step": 270
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 8.967573779193987e-05,
+      "loss": 1.0357,
+      "step": 271
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 8.903299405874684e-05,
+      "loss": 1.1066,
+      "step": 272
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 8.839070858747697e-05,
+      "loss": 0.9976,
+      "step": 273
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 8.774890821635548e-05,
+      "loss": 0.8978,
+      "step": 274
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 8.710761976333734e-05,
+      "loss": 1.0132,
+      "step": 275
+    },
+    {
+      "epoch": 2.18,
+      "eval_loss": 1.297065258026123,
+      "eval_runtime": 61.8107,
+      "eval_samples_per_second": 4.045,
+      "eval_steps_per_second": 4.045,
+      "step": 275
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.646687002498692e-05,
+      "loss": 0.9386,
+      "step": 276
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.582668577535797e-05,
+      "loss": 0.8478,
+      "step": 277
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 8.518709376487515e-05,
+      "loss": 1.0554,
+      "step": 278
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.454812071921596e-05,
+      "loss": 1.0435,
+      "step": 279
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.390979333819426e-05,
+      "loss": 1.04,
+      "step": 280
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.327213829464449e-05,
+      "loss": 1.0513,
+      "step": 281
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 1.1234,
+      "step": 282
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.199895176971488e-05,
+      "loss": 1.0556,
+      "step": 283
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.13634734890818e-05,
+      "loss": 0.9852,
+      "step": 284
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.072877394519102e-05,
+      "loss": 0.9753,
+      "step": 285
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.009487965928596e-05,
+      "loss": 1.1802,
+      "step": 286
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 7.94618171189618e-05,
+      "loss": 0.8883,
+      "step": 287
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 7.882961277705895e-05,
+      "loss": 1.0442,
+      "step": 288
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 7.819829305055743e-05,
+      "loss": 0.9863,
+      "step": 289
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 7.756788431947326e-05,
+      "loss": 1.0352,
+      "step": 290
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 7.693841292575598e-05,
+      "loss": 1.08,
+      "step": 291
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 7.630990517218808e-05,
+      "loss": 1.0916,
+      "step": 292
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 7.568238732128585e-05,
+      "loss": 0.956,
+      "step": 293
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 7.505588559420189e-05,
+      "loss": 1.0018,
+      "step": 294
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 7.443042616962965e-05,
+      "loss": 1.0782,
+      "step": 295
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 7.380603518270941e-05,
+      "loss": 1.098,
+      "step": 296
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 7.318273872393625e-05,
+      "loss": 1.1226,
+      "step": 297
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 7.256056283806986e-05,
+      "loss": 0.9865,
+      "step": 298
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 7.193953352304613e-05,
+      "loss": 0.9185,
+      "step": 299
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 7.131967672889101e-05,
+      "loss": 0.9713,
+      "step": 300
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 1.2972489595413208,
+      "eval_runtime": 61.7846,
+      "eval_samples_per_second": 4.046,
+      "eval_steps_per_second": 4.046,
+      "step": 300
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 7.070101835663594e-05,
+      "loss": 1.0547,
+      "step": 301
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 7.008358425723585e-05,
+      "loss": 1.1558,
+      "step": 302
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 6.94674002304887e-05,
+      "loss": 1.0398,
+      "step": 303
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 6.885249202395754e-05,
+      "loss": 0.8272,
+      "step": 304
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 6.823888533189469e-05,
+      "loss": 0.9847,
+      "step": 305
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 6.762660579416791e-05,
+      "loss": 0.919,
+      "step": 306
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 6.701567899518924e-05,
+      "loss": 0.9437,
+      "step": 307
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 6.640613046284581e-05,
+      "loss": 1.0459,
+      "step": 308
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 0.98,
+      "step": 309
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 6.519127002059095e-05,
+      "loss": 1.1745,
+      "step": 310
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 6.458600887424125e-05,
+      "loss": 0.9702,
+      "step": 311
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 6.398222751952899e-05,
+      "loss": 0.9785,
+      "step": 312
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 6.337995118576521e-05,
+      "loss": 0.9573,
+      "step": 313
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 6.277920503937303e-05,
+      "loss": 0.9505,
+      "step": 314
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.218001418283576e-05,
+      "loss": 1.0688,
+      "step": 315
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.158240365364823e-05,
+      "loss": 0.9826,
+      "step": 316
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.098639842327052e-05,
+      "loss": 0.9873,
+      "step": 317
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.039202339608432e-05,
+      "loss": 0.969,
+      "step": 318
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5.979930340835265e-05,
+      "loss": 1.004,
+      "step": 319
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.920826322718165e-05,
+      "loss": 1.0032,
+      "step": 320
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.861892754948609e-05,
+      "loss": 1.0514,
+      "step": 321
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.80313210009571e-05,
+      "loss": 1.163,
+      "step": 322
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.744546813503328e-05,
+      "loss": 1.0799,
+      "step": 323
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.6861393431874675e-05,
+      "loss": 0.8988,
+      "step": 324
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5.6279121297339924e-05,
+      "loss": 1.0341,
+      "step": 325
+    },
+    {
+      "epoch": 2.58,
+      "eval_loss": 1.2901266813278198,
+      "eval_runtime": 61.8384,
+      "eval_samples_per_second": 4.043,
+      "eval_steps_per_second": 4.043,
+      "step": 325
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5.5698676061966515e-05,
+      "loss": 1.1851,
+      "step": 326
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.5120081979953785e-05,
+      "loss": 1.0744,
+      "step": 327
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.4543363228149946e-05,
+      "loss": 0.9741,
+      "step": 328
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5.3968543905041444e-05,
+      "loss": 0.9947,
+      "step": 329
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.339564802974615e-05,
+      "loss": 1.1458,
+      "step": 330
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5.282469954100968e-05,
+      "loss": 0.8714,
+      "step": 331
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.22557222962051e-05,
+      "loss": 0.8682,
+      "step": 332
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.168874007033615e-05,
+      "loss": 0.9135,
+      "step": 333
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.112377655504359e-05,
+      "loss": 1.0744,
+      "step": 334
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.056085535761532e-05,
+      "loss": 0.9575,
+      "step": 335
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.0823,
+      "step": 336
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 4.9441233917824106e-05,
+      "loss": 0.8986,
+      "step": 337
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 4.888458045941269e-05,
+      "loss": 0.9171,
+      "step": 338
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 4.833006288481371e-05,
+      "loss": 1.0258,
+      "step": 339
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.777770436482617e-05,
+      "loss": 1.0722,
+      "step": 340
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.722752798003184e-05,
+      "loss": 1.0347,
+      "step": 341
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.66795567198309e-05,
+      "loss": 1.034,
+      "step": 342
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.6133813481481246e-05,
+      "loss": 0.9559,
+      "step": 343
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.559032106914173e-05,
+      "loss": 0.9619,
+      "step": 344
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.50491021929194e-05,
+      "loss": 0.9795,
+      "step": 345
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.451017946792032e-05,
+      "loss": 0.9236,
+      "step": 346
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.397357541330476e-05,
+      "loss": 0.8438,
+      "step": 347
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.343931245134616e-05,
+      "loss": 0.9684,
+      "step": 348
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.2907412906494174e-05,
+      "loss": 0.9478,
+      "step": 349
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 4.2377899004441966e-05,
+      "loss": 1.0808,
+      "step": 350
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 1.2958766222000122,
+      "eval_runtime": 61.8006,
+      "eval_samples_per_second": 4.045,
+      "eval_steps_per_second": 4.045,
+      "step": 350
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 4.185079287119733e-05,
+      "loss": 1.112,
+      "step": 351
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 4.132611653215822e-05,
+      "loss": 0.983,
+      "step": 352
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 4.080389191119241e-05,
+      "loss": 1.1557,
+      "step": 353
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 4.028414082972141e-05,
+      "loss": 0.9879,
+      "step": 354
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.9766885005808565e-05,
+      "loss": 0.833,
+      "step": 355
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.9252146053251636e-05,
+      "loss": 1.0431,
+      "step": 356
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 3.873994548067972e-05,
+      "loss": 1.0371,
+      "step": 357
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 3.8230304690654304e-05,
+      "loss": 0.9455,
+      "step": 358
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 3.772324497877511e-05,
+      "loss": 0.9356,
+      "step": 359
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 3.721878753279017e-05,
+      "loss": 0.9807,
+      "step": 360
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.671695343171052e-05,
+      "loss": 0.9067,
+      "step": 361
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.621776364492939e-05,
+      "loss": 0.9393,
+      "step": 362
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 0.9856,
+      "step": 363
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.522740033849411e-05,
+      "loss": 1.0474,
+      "step": 364
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.4736268201674574e-05,
+      "loss": 0.9027,
+      "step": 365
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.424786314309365e-05,
+      "loss": 0.9521,
+      "step": 366
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.376220557100523e-05,
+      "loss": 0.9211,
+      "step": 367
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.3279315778858036e-05,
+      "loss": 0.9576,
+      "step": 368
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.279921394444776e-05,
+      "loss": 0.9643,
+      "step": 369
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.2321920129073816e-05,
+      "loss": 1.0322,
+      "step": 370
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.18474542767011e-05,
+      "loss": 0.9762,
+      "step": 371
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.137583621312665e-05,
+      "loss": 1.1331,
+      "step": 372
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.090708564515124e-05,
+      "loss": 1.104,
+      "step": 373
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.04412221597558e-05,
+      "loss": 0.9667,
+      "step": 374
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.997826522328315e-05,
+      "loss": 0.9492,
+      "step": 375
+    },
+    {
+      "epoch": 2.98,
+      "eval_loss": 1.2812517881393433,
+      "eval_runtime": 61.7806,
+      "eval_samples_per_second": 4.047,
+      "eval_steps_per_second": 4.047,
+      "step": 375
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.9518234180624393e-05,
+      "loss": 1.0619,
+      "step": 376
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.906114825441072e-05,
+      "loss": 0.9439,
+      "step": 377
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.8607026544210114e-05,
+      "loss": 0.905,
+      "step": 378
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.8155888025729273e-05,
+      "loss": 0.9329,
+      "step": 379
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.770775155002071e-05,
+      "loss": 1.012,
+      "step": 380
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.7262635842695127e-05,
+      "loss": 0.7935,
+      "step": 381
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.6820559503138797e-05,
+      "loss": 0.9079,
+      "step": 382
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.6381541003736486e-05,
+      "loss": 0.884,
+      "step": 383
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.594559868909956e-05,
+      "loss": 0.8965,
+      "step": 384
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.5512750775299432e-05,
+      "loss": 0.7722,
+      "step": 385
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.50830153491064e-05,
+      "loss": 0.8773,
+      "step": 386
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.465641036723393e-05,
+      "loss": 0.8756,
+      "step": 387
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.423295365558821e-05,
+      "loss": 0.8886,
+      "step": 388
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.381266290852341e-05,
+      "loss": 0.9629,
+      "step": 389
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 0.9036,
+      "step": 390
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.298164942336205e-05,
+      "loss": 0.8959,
+      "step": 391
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 2.2570961409586754e-05,
+      "loss": 0.8581,
+      "step": 392
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.2163508807583998e-05,
+      "loss": 0.7808,
+      "step": 393
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.1759308642968025e-05,
+      "loss": 0.8382,
+      "step": 394
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.1358377805448382e-05,
+      "loss": 0.8431,
+      "step": 395
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 2.0960733048124083e-05,
+      "loss": 0.9881,
+      "step": 396
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 2.0566390986783646e-05,
+      "loss": 0.7971,
+      "step": 397
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.01753680992107e-05,
+      "loss": 0.9015,
+      "step": 398
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.9787680724495617e-05,
+      "loss": 0.8186,
+      "step": 399
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.9403345062352573e-05,
+      "loss": 0.8436,
+      "step": 400
+    },
+    {
+      "epoch": 3.16,
+      "eval_loss": 1.319449782371521,
+      "eval_runtime": 61.8064,
+      "eval_samples_per_second": 4.045,
+      "eval_steps_per_second": 4.045,
+      "step": 400
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.9022377172442752e-05,
+      "loss": 0.7585,
+      "step": 401
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.864479297370325e-05,
+      "loss": 0.8746,
+      "step": 402
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.8270608243681953e-05,
+      "loss": 0.9837,
+      "step": 403
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.7899838617878163e-05,
+      "loss": 0.9746,
+      "step": 404
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.7532499589089323e-05,
+      "loss": 0.7808,
+      "step": 405
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.7168606506763695e-05,
+      "loss": 0.8529,
+      "step": 406
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.6808174576358848e-05,
+      "loss": 1.0696,
+      "step": 407
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.6451218858706374e-05,
+      "loss": 0.9803,
+      "step": 408
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.6097754269382536e-05,
+      "loss": 0.9196,
+      "step": 409
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.5747795578085046e-05,
+      "loss": 0.8965,
+      "step": 410
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.5401357408015893e-05,
+      "loss": 0.8625,
+      "step": 411
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.505845423527027e-05,
+      "loss": 0.8837,
+      "step": 412
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.47191003882317e-05,
+      "loss": 0.9438,
+      "step": 413
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.4383310046973365e-05,
+      "loss": 0.8853,
+      "step": 414
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.40510972426655e-05,
+      "loss": 0.8758,
+      "step": 415
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.3722475856989158e-05,
+      "loss": 0.8716,
+      "step": 416
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 0.9002,
+      "step": 417
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.307606211733522e-05,
+      "loss": 0.9933,
+      "step": 418
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.2758296774084632e-05,
+      "loss": 0.857,
+      "step": 419
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.2444176869790925e-05,
+      "loss": 0.8833,
+      "step": 420
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.213371553011411e-05,
+      "loss": 0.7823,
+      "step": 421
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.18269257278392e-05,
+      "loss": 0.806,
+      "step": 422
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.1523820282334219e-05,
+      "loss": 1.0458,
+      "step": 423
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.1224411859014417e-05,
+      "loss": 0.821,
+      "step": 424
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.092871296881307e-05,
+      "loss": 1.0561,
+      "step": 425
+    },
+    {
+      "epoch": 3.36,
+      "eval_loss": 1.3116503953933716,
+      "eval_runtime": 61.8632,
+      "eval_samples_per_second": 4.041,
+      "eval_steps_per_second": 4.041,
+      "step": 425
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0636735967658784e-05,
+      "loss": 0.9725,
+      "step": 426
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0348493055959062e-05,
+      "loss": 0.9468,
+      "step": 427
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0063996278090704e-05,
+      "loss": 0.8698,
+      "step": 428
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.783257521896227e-06,
+      "loss": 0.794,
+      "step": 429
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.506288518187467e-06,
+      "loss": 0.7835,
+      "step": 430
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.233100840255127e-06,
+      "loss": 0.9157,
+      "step": 431
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 8.963705903385345e-06,
+      "loss": 0.8811,
+      "step": 432
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.698114964382598e-06,
+      "loss": 0.7957,
+      "step": 433
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.436339121099412e-06,
+      "loss": 0.8907,
+      "step": 434
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.178389311972612e-06,
+      "loss": 0.8602,
+      "step": 435
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 7.92427631556617e-06,
+      "loss": 0.8855,
+      "step": 436
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 7.674010750120964e-06,
+      "loss": 0.9052,
+      "step": 437
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 7.427603073110967e-06,
+      "loss": 1.01,
+      "step": 438
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 7.185063580806317e-06,
+      "loss": 0.9242,
+      "step": 439
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 6.946402407843155e-06,
+      "loss": 0.7778,
+      "step": 440
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 6.7116295267999455e-06,
+      "loss": 0.9008,
+      "step": 441
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 6.480754747781037e-06,
+      "loss": 0.8944,
+      "step": 442
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 6.253787718006498e-06,
+      "loss": 0.9784,
+      "step": 443
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 0.827,
+      "step": 444
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.811614678238275e-06,
+      "loss": 0.8666,
+      "step": 445
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5.596427144670002e-06,
+      "loss": 0.9152,
+      "step": 446
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5.385184312424974e-06,
+      "loss": 0.9269,
+      "step": 447
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5.177895008392353e-06,
+      "loss": 0.8852,
+      "step": 448
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 4.974567894261217e-06,
+      "loss": 0.8283,
+      "step": 449
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.775211466158469e-06,
+      "loss": 0.9367,
+      "step": 450
+    },
+    {
+      "epoch": 3.57,
+      "eval_loss": 1.3111974000930786,
+      "eval_runtime": 61.8414,
+      "eval_samples_per_second": 4.043,
+      "eval_steps_per_second": 4.043,
+      "step": 450
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.5798340542938855e-06,
+      "loss": 0.7843,
+      "step": 451
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 4.3884438226120424e-06,
+      "loss": 0.9498,
+      "step": 452
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 4.20104876845111e-06,
+      "loss": 1.0189,
+      "step": 453
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.017656722208807e-06,
+      "loss": 0.9243,
+      "step": 454
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.8382753470150854e-06,
+      "loss": 0.8153,
+      "step": 455
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.662912138411967e-06,
+      "loss": 0.8717,
+      "step": 456
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.4915744240403558e-06,
+      "loss": 0.8371,
+      "step": 457
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.3242693633337983e-06,
+      "loss": 0.9489,
+      "step": 458
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.161003947219421e-06,
+      "loss": 0.9656,
+      "step": 459
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.0017849978256516e-06,
+      "loss": 0.8834,
+      "step": 460
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 2.8466191681973507e-06,
+      "loss": 0.8328,
+      "step": 461
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 2.6955129420176196e-06,
+      "loss": 0.948,
+      "step": 462
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 2.548472633337007e-06,
+      "loss": 0.6954,
+      "step": 463
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 2.4055043863096428e-06,
+      "loss": 0.9203,
+      "step": 464
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 2.266614174936443e-06,
+      "loss": 0.9194,
+      "step": 465
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.1318078028155888e-06,
+      "loss": 0.8099,
+      "step": 466
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.0010909028998827e-06,
+      "loss": 0.9428,
+      "step": 467
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.874468937261531e-06,
+      "loss": 0.8694,
+      "step": 468
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.751947196863779e-06,
+      "loss": 0.8197,
+      "step": 469
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.6335308013398886e-06,
+      "loss": 0.9786,
+      "step": 470
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 0.8906,
+      "step": 471
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.409033665520354e-06,
+      "loss": 0.8925,
+      "step": 472
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.3029623059517493e-06,
+      "loss": 0.8353,
+      "step": 473
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.201015052319099e-06,
+      "loss": 0.8909,
+      "step": 474
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.1031961645402877e-06,
+      "loss": 0.9294,
+      "step": 475
+    },
+    {
+      "epoch": 3.77,
+      "eval_loss": 1.3116556406021118,
+      "eval_runtime": 61.8408,
+      "eval_samples_per_second": 4.043,
+      "eval_steps_per_second": 4.043,
+      "step": 475
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.0095097300273026e-06,
+      "loss": 0.9107,
+      "step": 476
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.199596635154683e-07,
+      "loss": 1.0287,
+      "step": 477
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 8.345497068998897e-07,
+      "loss": 0.8321,
+      "step": 478
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 7.532834290790436e-07,
+      "loss": 0.8194,
+      "step": 479
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 6.761642258056978e-07,
+      "loss": 0.9389,
+      "step": 480
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 6.031953195450135e-07,
+      "loss": 0.9331,
+      "step": 481
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.343797593398536e-07,
+      "loss": 0.8131,
+      "step": 482
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.6972042068341714e-07,
+      "loss": 0.8739,
+      "step": 483
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 4.092200053990691e-07,
+      "loss": 1.1704,
+      "step": 484
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.5288104152743083e-07,
+      "loss": 0.7947,
+      "step": 485
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.007058832207976e-07,
+      "loss": 1.0854,
+      "step": 486
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.5269671064467313e-07,
+      "loss": 0.7943,
+      "step": 487
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.088555298867978e-07,
+      "loss": 0.8828,
+      "step": 488
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.6918417287318245e-07,
+      "loss": 0.769,
+      "step": 489
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.3368429729168076e-07,
+      "loss": 1.0282,
+      "step": 490
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.023573865226446e-07,
+      "loss": 0.958,
+      "step": 491
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 7.520474957699586e-08,
+      "loss": 0.9203,
+      "step": 492
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5.2227521041470216e-08,
+      "loss": 0.9252,
+      "step": 493
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.3426661031255026e-08,
+      "loss": 1.0799,
+      "step": 494
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.8802955149865852e-08,
+      "loss": 0.9298,
+      "step": 495
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 8.357014456272794e-09,
+      "loss": 0.8072,
+      "step": 496
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 496,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "total_flos": 1.6132667358786355e+17,
+  "trial_name": null,
+  "trial_params": null
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e0f9164e794b55ad50d42a14241f43c652268b1fbbef7eeb693efd5292290cd
+size 80218501

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16894d75beb0b2e64d157ec733c62af03eedf21d64d5fe6ff9a6126671b074ab
+size 14575

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f893550683b6eab7c97aef6d41a568d827a96bd9f9d8dd6860446961dd8df8f
+size 627

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4dd619dab9182cb5c3d41449df0f56b3c988747f0ca62d01d956c1090a75990f
+oid sha256:29182789cc886c4cb1e203599999900c580a72f92286132414128006e2d749f4
 size 4603