{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 29.574468085106382,
  "eval_steps": 500,
  "global_step": 405,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 4.9999740874098114e-05,
      "loss": 1.6789,
      "step": 1
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.999352212103373e-05,
      "loss": 1.6341,
      "step": 5
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.9974091841168195e-05,
      "loss": 1.5276,
      "step": 10
    },
    {
      "epoch": 0.55,
      "eval_loss": 1.4329197406768799,
      "eval_runtime": 2.2791,
      "eval_samples_per_second": 36.419,
      "eval_steps_per_second": 1.316,
      "step": 13
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.994171922976348e-05,
      "loss": 1.4632,
      "step": 15
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.9896421063288286e-05,
      "loss": 1.3978,
      "step": 20
    },
    {
      "epoch": 1.49,
      "learning_rate": 4.983822081662578e-05,
      "loss": 1.352,
      "step": 25
    },
    {
      "epoch": 1.57,
      "eval_loss": 1.240639090538025,
      "eval_runtime": 2.1562,
      "eval_samples_per_second": 38.494,
      "eval_steps_per_second": 1.391,
      "step": 27
    },
    {
      "epoch": 2.13,
      "learning_rate": 4.976714865090827e-05,
      "loss": 1.2632,
      "step": 30
    },
    {
      "epoch": 2.34,
      "learning_rate": 4.9683241397886745e-05,
      "loss": 1.1895,
      "step": 35
    },
    {
      "epoch": 2.55,
      "learning_rate": 4.958654254084355e-05,
      "loss": 1.1329,
      "step": 40
    },
    {
      "epoch": 2.55,
      "eval_loss": 1.0908643007278442,
      "eval_runtime": 2.1459,
      "eval_samples_per_second": 38.679,
      "eval_steps_per_second": 1.398,
      "step": 40
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.947710219205808e-05,
      "loss": 1.0932,
      "step": 45
    },
    {
      "epoch": 3.4,
      "learning_rate": 4.9354977066836986e-05,
      "loss": 1.0628,
      "step": 50
    },
    {
      "epoch": 3.57,
      "eval_loss": 1.0299233198165894,
      "eval_runtime": 2.1556,
      "eval_samples_per_second": 38.505,
      "eval_steps_per_second": 1.392,
      "step": 54
    },
    {
      "epoch": 4.04,
      "learning_rate": 4.922023045412266e-05,
      "loss": 1.0544,
      "step": 55
    },
    {
      "epoch": 4.26,
      "learning_rate": 4.907293218369499e-05,
      "loss": 1.0274,
      "step": 60
    },
    {
      "epoch": 4.47,
      "learning_rate": 4.8913158589983374e-05,
      "loss": 1.0022,
      "step": 65
    },
    {
      "epoch": 4.55,
      "eval_loss": 0.9811795949935913,
      "eval_runtime": 2.141,
      "eval_samples_per_second": 38.767,
      "eval_steps_per_second": 1.401,
      "step": 67
    },
    {
      "epoch": 5.11,
      "learning_rate": 4.874099247250798e-05,
      "loss": 0.9986,
      "step": 70
    },
    {
      "epoch": 5.32,
      "learning_rate": 4.855652305297052e-05,
      "loss": 0.979,
      "step": 75
    },
    {
      "epoch": 5.53,
      "learning_rate": 4.835984592901678e-05,
      "loss": 0.957,
      "step": 80
    },
    {
      "epoch": 5.57,
      "eval_loss": 0.9445169568061829,
      "eval_runtime": 2.1636,
      "eval_samples_per_second": 38.362,
      "eval_steps_per_second": 1.387,
      "step": 81
    },
    {
      "epoch": 6.17,
      "learning_rate": 4.815106302469513e-05,
      "loss": 0.9335,
      "step": 85
    },
    {
      "epoch": 6.38,
      "learning_rate": 4.793028253763633e-05,
      "loss": 0.9148,
      "step": 90
    },
    {
      "epoch": 6.55,
      "eval_loss": 0.8947578072547913,
      "eval_runtime": 2.1571,
      "eval_samples_per_second": 38.478,
      "eval_steps_per_second": 1.391,
      "step": 94
    },
    {
      "epoch": 7.02,
      "learning_rate": 4.769761888298222e-05,
      "loss": 0.9063,
      "step": 95
    },
    {
      "epoch": 7.23,
      "learning_rate": 4.74531926340924e-05,
      "loss": 0.8713,
      "step": 100
    },
    {
      "epoch": 7.45,
      "learning_rate": 4.719713046005938e-05,
      "loss": 0.8443,
      "step": 105
    },
    {
      "epoch": 7.57,
      "eval_loss": 0.8432280421257019,
      "eval_runtime": 2.1643,
      "eval_samples_per_second": 38.349,
      "eval_steps_per_second": 1.386,
      "step": 108
    },
    {
      "epoch": 8.09,
      "learning_rate": 4.6929565060064864e-05,
      "loss": 0.8298,
      "step": 110
    },
    {
      "epoch": 8.3,
      "learning_rate": 4.665063509461097e-05,
      "loss": 0.7866,
      "step": 115
    },
    {
      "epoch": 8.51,
      "learning_rate": 4.6360485113662216e-05,
      "loss": 0.7645,
      "step": 120
    },
    {
      "epoch": 8.55,
      "eval_loss": 0.7846970558166504,
      "eval_runtime": 2.1368,
      "eval_samples_per_second": 38.843,
      "eval_steps_per_second": 1.404,
      "step": 121
    },
    {
      "epoch": 9.15,
      "learning_rate": 4.605926548173529e-05,
      "loss": 0.738,
      "step": 125
    },
    {
      "epoch": 9.36,
      "learning_rate": 4.574713229997563e-05,
      "loss": 0.7076,
      "step": 130
    },
    {
      "epoch": 9.57,
      "learning_rate": 4.542424732526105e-05,
      "loss": 0.6952,
      "step": 135
    },
    {
      "epoch": 9.57,
      "eval_loss": 0.7192484140396118,
      "eval_runtime": 2.1619,
      "eval_samples_per_second": 38.391,
      "eval_steps_per_second": 1.388,
      "step": 135
    },
    {
      "epoch": 10.21,
      "learning_rate": 4.509077788637446e-05,
      "loss": 0.647,
      "step": 140
    },
    {
      "epoch": 10.43,
      "learning_rate": 4.4746896797288965e-05,
      "loss": 0.639,
      "step": 145
    },
    {
      "epoch": 10.55,
      "eval_loss": 0.6671289205551147,
      "eval_runtime": 2.1484,
      "eval_samples_per_second": 38.633,
      "eval_steps_per_second": 1.396,
      "step": 148
    },
    {
      "epoch": 11.06,
      "learning_rate": 4.43927822676105e-05,
      "loss": 0.6133,
      "step": 150
    },
    {
      "epoch": 11.28,
      "learning_rate": 4.402861781022412e-05,
      "loss": 0.5832,
      "step": 155
    },
    {
      "epoch": 11.49,
      "learning_rate": 4.365459214619214e-05,
      "loss": 0.5683,
      "step": 160
    },
    {
      "epoch": 11.57,
      "eval_loss": 0.6111546158790588,
      "eval_runtime": 2.1558,
      "eval_samples_per_second": 38.5,
      "eval_steps_per_second": 1.392,
      "step": 162
    },
    {
      "epoch": 12.13,
      "learning_rate": 4.3270899106953105e-05,
      "loss": 0.5655,
      "step": 165
    },
    {
      "epoch": 12.34,
      "learning_rate": 4.2877737533872485e-05,
      "loss": 0.5355,
      "step": 170
    },
    {
      "epoch": 12.55,
      "learning_rate": 4.2475311175197056e-05,
      "loss": 0.5223,
      "step": 175
    },
    {
      "epoch": 12.55,
      "eval_loss": 0.5776758193969727,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 38.208,
      "eval_steps_per_second": 1.381,
      "step": 175
    },
    {
      "epoch": 13.19,
      "learning_rate": 4.206382858046636e-05,
      "loss": 0.5186,
      "step": 180
    },
    {
      "epoch": 13.4,
      "learning_rate": 4.164350299243601e-05,
      "loss": 0.4958,
      "step": 185
    },
    {
      "epoch": 13.57,
      "eval_loss": 0.5592173337936401,
      "eval_runtime": 2.1614,
      "eval_samples_per_second": 38.402,
      "eval_steps_per_second": 1.388,
      "step": 189
    },
    {
      "epoch": 14.04,
      "learning_rate": 4.12145522365689e-05,
      "loss": 0.479,
      "step": 190
    },
    {
      "epoch": 14.26,
      "learning_rate": 4.077719860815132e-05,
      "loss": 0.4846,
      "step": 195
    },
    {
      "epoch": 14.47,
      "learning_rate": 4.033166875709291e-05,
      "loss": 0.4592,
      "step": 200
    },
    {
      "epoch": 14.55,
      "eval_loss": 0.5380522608757019,
      "eval_runtime": 2.1511,
      "eval_samples_per_second": 38.584,
      "eval_steps_per_second": 1.395,
      "step": 202
    },
    {
      "epoch": 15.11,
      "learning_rate": 3.9878193570469746e-05,
      "loss": 0.4718,
      "step": 205
    },
    {
      "epoch": 15.32,
      "learning_rate": 3.941700805287168e-05,
      "loss": 0.4436,
      "step": 210
    },
    {
      "epoch": 15.53,
      "learning_rate": 3.894835120461584e-05,
      "loss": 0.4602,
      "step": 215
    },
    {
      "epoch": 15.57,
      "eval_loss": 0.5099533796310425,
      "eval_runtime": 2.1624,
      "eval_samples_per_second": 38.383,
      "eval_steps_per_second": 1.387,
      "step": 216
    },
    {
      "epoch": 16.17,
      "learning_rate": 3.8472465897889394e-05,
      "loss": 0.4394,
      "step": 220
    },
    {
      "epoch": 16.38,
      "learning_rate": 3.798959875088584e-05,
      "loss": 0.4486,
      "step": 225
    },
    {
      "epoch": 16.55,
      "eval_loss": 0.5117282867431641,
      "eval_runtime": 2.1571,
      "eval_samples_per_second": 38.478,
      "eval_steps_per_second": 1.391,
      "step": 229
    },
    {
      "epoch": 17.02,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4298,
      "step": 230
    },
    {
      "epoch": 17.23,
      "learning_rate": 3.700392337014789e-05,
      "loss": 0.4295,
      "step": 235
    },
    {
      "epoch": 17.45,
      "learning_rate": 3.6501625943278805e-05,
      "loss": 0.4274,
      "step": 240
    },
    {
      "epoch": 17.57,
      "eval_loss": 0.5084130764007568,
      "eval_runtime": 2.1702,
      "eval_samples_per_second": 38.245,
      "eval_steps_per_second": 1.382,
      "step": 243
    },
    {
      "epoch": 18.09,
      "learning_rate": 3.599336802514764e-05,
      "loss": 0.4221,
      "step": 245
    },
    {
      "epoch": 18.3,
      "learning_rate": 3.547941301041661e-05,
      "loss": 0.4064,
      "step": 250
    },
    {
      "epoch": 18.51,
      "learning_rate": 3.496002724615604e-05,
      "loss": 0.4239,
      "step": 255
    },
    {
      "epoch": 18.55,
      "eval_loss": 0.4908837080001831,
      "eval_runtime": 2.1436,
      "eval_samples_per_second": 38.721,
      "eval_steps_per_second": 1.4,
      "step": 256
    },
    {
      "epoch": 19.15,
      "learning_rate": 3.443547989381536e-05,
      "loss": 0.421,
      "step": 260
    },
    {
      "epoch": 19.36,
      "learning_rate": 3.390604278973543e-05,
      "loss": 0.4126,
      "step": 265
    },
    {
      "epoch": 19.57,
      "learning_rate": 3.3371990304274656e-05,
      "loss": 0.4055,
      "step": 270
    },
    {
      "epoch": 19.57,
      "eval_loss": 0.5006158351898193,
      "eval_runtime": 2.157,
      "eval_samples_per_second": 38.479,
      "eval_steps_per_second": 1.391,
      "step": 270
    },
    {
      "epoch": 20.21,
      "learning_rate": 3.283359919962206e-05,
      "loss": 0.4136,
      "step": 275
    },
    {
      "epoch": 20.43,
      "learning_rate": 3.2291148486370626e-05,
      "loss": 0.3931,
      "step": 280
    },
    {
      "epoch": 20.55,
      "eval_loss": 0.4958841800689697,
      "eval_runtime": 2.1445,
      "eval_samples_per_second": 38.704,
      "eval_steps_per_second": 1.399,
      "step": 283
    },
    {
      "epoch": 21.06,
      "learning_rate": 3.1744919278925605e-05,
      "loss": 0.4056,
      "step": 285
    },
    {
      "epoch": 21.28,
      "learning_rate": 3.11951946498225e-05,
      "loss": 0.3901,
      "step": 290
    },
    {
      "epoch": 21.49,
      "learning_rate": 3.0642259483030266e-05,
      "loss": 0.3986,
      "step": 295
    },
    {
      "epoch": 21.57,
      "eval_loss": 0.4853080213069916,
      "eval_runtime": 2.1603,
      "eval_samples_per_second": 38.421,
      "eval_steps_per_second": 1.389,
      "step": 297
    },
    {
      "epoch": 22.13,
      "learning_rate": 3.008640032631585e-05,
      "loss": 0.3931,
      "step": 300
    },
    {
      "epoch": 22.34,
      "learning_rate": 2.9527905242746394e-05,
      "loss": 0.3849,
      "step": 305
    },
    {
      "epoch": 22.55,
      "learning_rate": 2.8967063661406285e-05,
      "loss": 0.3977,
      "step": 310
    },
    {
      "epoch": 22.55,
      "eval_loss": 0.4858959913253784,
      "eval_runtime": 2.1635,
      "eval_samples_per_second": 38.363,
      "eval_steps_per_second": 1.387,
      "step": 310
    },
    {
      "epoch": 23.19,
      "learning_rate": 2.840416622740617e-05,
      "loss": 0.3785,
      "step": 315
    },
    {
      "epoch": 23.4,
      "learning_rate": 2.7839504651261872e-05,
      "loss": 0.3936,
      "step": 320
    },
    {
      "epoch": 23.57,
      "eval_loss": 0.4974050223827362,
      "eval_runtime": 2.1629,
      "eval_samples_per_second": 38.375,
      "eval_steps_per_second": 1.387,
      "step": 324
    },
    {
      "epoch": 24.04,
      "learning_rate": 2.7273371557721176e-05,
      "loss": 0.3901,
      "step": 325
    },
    {
      "epoch": 24.26,
      "learning_rate": 2.6706060334116777e-05,
      "loss": 0.3888,
      "step": 330
    },
    {
      "epoch": 24.47,
      "learning_rate": 2.6137864978324095e-05,
      "loss": 0.3821,
      "step": 335
    },
    {
      "epoch": 24.55,
      "eval_loss": 0.49520766735076904,
      "eval_runtime": 2.1468,
      "eval_samples_per_second": 38.662,
      "eval_steps_per_second": 1.397,
      "step": 337
    },
    {
      "epoch": 25.11,
      "learning_rate": 2.556907994640264e-05,
      "loss": 0.3813,
      "step": 340
    },
    {
      "epoch": 25.32,
      "learning_rate": 2.5e-05,
      "loss": 0.3672,
      "step": 345
    },
    {
      "epoch": 25.53,
      "learning_rate": 2.4430920053597356e-05,
      "loss": 0.3877,
      "step": 350
    },
    {
      "epoch": 25.57,
      "eval_loss": 0.4949447512626648,
      "eval_runtime": 2.1661,
      "eval_samples_per_second": 38.317,
      "eval_steps_per_second": 1.385,
      "step": 351
    },
    {
      "epoch": 26.17,
      "learning_rate": 2.3862135021675918e-05,
      "loss": 0.3777,
      "step": 355
    },
    {
      "epoch": 26.38,
      "learning_rate": 2.329393966588323e-05,
      "loss": 0.3681,
      "step": 360
    },
    {
      "epoch": 26.55,
      "eval_loss": 0.48658838868141174,
      "eval_runtime": 2.1579,
      "eval_samples_per_second": 38.464,
      "eval_steps_per_second": 1.39,
      "step": 364
    },
    {
      "epoch": 27.02,
      "learning_rate": 2.2726628442278826e-05,
      "loss": 0.3767,
      "step": 365
    },
    {
      "epoch": 27.23,
      "learning_rate": 2.2160495348738123e-05,
      "loss": 0.379,
      "step": 370
    },
    {
      "epoch": 27.45,
      "learning_rate": 2.159583377259384e-05,
      "loss": 0.3681,
      "step": 375
    },
    {
      "epoch": 27.57,
      "eval_loss": 0.492579847574234,
      "eval_runtime": 2.1665,
      "eval_samples_per_second": 38.311,
      "eval_steps_per_second": 1.385,
      "step": 378
    },
    {
      "epoch": 28.09,
      "learning_rate": 2.1032936338593718e-05,
      "loss": 0.3662,
      "step": 380
    },
    {
      "epoch": 28.3,
      "learning_rate": 2.0472094757253605e-05,
      "loss": 0.3705,
      "step": 385
    },
    {
      "epoch": 28.51,
      "learning_rate": 1.991359967368416e-05,
      "loss": 0.371,
      "step": 390
    },
    {
      "epoch": 28.55,
      "eval_loss": 0.4816737771034241,
      "eval_runtime": 2.1381,
      "eval_samples_per_second": 38.819,
      "eval_steps_per_second": 1.403,
      "step": 391
    },
    {
      "epoch": 29.15,
      "learning_rate": 1.9357740516969736e-05,
      "loss": 0.3648,
      "step": 395
    },
    {
      "epoch": 29.36,
      "learning_rate": 1.8804805350177505e-05,
      "loss": 0.3727,
      "step": 400
    },
    {
      "epoch": 29.57,
      "learning_rate": 1.825508072107439e-05,
      "loss": 0.3604,
      "step": 405
    },
    {
      "epoch": 29.57,
      "eval_loss": 0.4923429787158966,
      "eval_runtime": 2.1712,
      "eval_samples_per_second": 38.228,
      "eval_steps_per_second": 1.382,
      "step": 405
    },
    {
      "epoch": 29.57,
      "step": 405,
      "total_flos": 2750244861247488.0,
      "train_loss": 0.6371645744935965,
      "train_runtime": 1974.0079,
      "train_samples_per_second": 11.337,
      "train_steps_per_second": 0.35
    }
  ],
  "logging_steps": 5,
  "max_steps": 690,
  "num_train_epochs": 30,
  "save_steps": 500,
  "total_flos": 2750244861247488.0,
  "trial_name": null,
  "trial_params": null
}