{
  "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": 2.0607,
      "step": 1
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.999352212103373e-05,
      "loss": 2.0069,
      "step": 5
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.9974091841168195e-05,
      "loss": 1.8242,
      "step": 10
    },
    {
      "epoch": 0.55,
      "eval_loss": 1.6627923250198364,
      "eval_runtime": 2.2915,
      "eval_samples_per_second": 36.22,
      "eval_steps_per_second": 1.309,
      "step": 13
    },
    {
      "epoch": 1.06,
      "learning_rate": 4.994171922976348e-05,
      "loss": 1.7045,
      "step": 15
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.9896421063288286e-05,
      "loss": 1.6023,
      "step": 20
    },
    {
      "epoch": 1.49,
      "learning_rate": 4.983822081662578e-05,
      "loss": 1.5291,
      "step": 25
    },
    {
      "epoch": 1.57,
      "eval_loss": 1.3837387561798096,
      "eval_runtime": 2.1494,
      "eval_samples_per_second": 38.615,
      "eval_steps_per_second": 1.396,
      "step": 27
    },
    {
      "epoch": 2.13,
      "learning_rate": 4.976714865090827e-05,
      "loss": 1.417,
      "step": 30
    },
    {
      "epoch": 2.34,
      "learning_rate": 4.9683241397886745e-05,
      "loss": 1.3126,
      "step": 35
    },
    {
      "epoch": 2.55,
      "learning_rate": 4.958654254084355e-05,
      "loss": 1.2362,
      "step": 40
    },
    {
      "epoch": 2.55,
      "eval_loss": 1.1785995960235596,
      "eval_runtime": 2.169,
      "eval_samples_per_second": 38.267,
      "eval_steps_per_second": 1.383,
      "step": 40
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.947710219205808e-05,
      "loss": 1.1757,
      "step": 45
    },
    {
      "epoch": 3.4,
      "learning_rate": 4.9354977066836986e-05,
      "loss": 1.1401,
      "step": 50
    },
    {
      "epoch": 3.57,
      "eval_loss": 1.097777247428894,
      "eval_runtime": 2.1856,
      "eval_samples_per_second": 37.976,
      "eval_steps_per_second": 1.373,
      "step": 54
    },
    {
      "epoch": 4.04,
      "learning_rate": 4.922023045412266e-05,
      "loss": 1.1249,
      "step": 55
    },
    {
      "epoch": 4.26,
      "learning_rate": 4.907293218369499e-05,
      "loss": 1.0927,
      "step": 60
    },
    {
      "epoch": 4.47,
      "learning_rate": 4.8913158589983374e-05,
      "loss": 1.0666,
      "step": 65
    },
    {
      "epoch": 4.55,
      "eval_loss": 1.042842984199524,
      "eval_runtime": 2.1605,
      "eval_samples_per_second": 38.418,
      "eval_steps_per_second": 1.389,
      "step": 67
    },
    {
      "epoch": 5.11,
      "learning_rate": 4.874099247250798e-05,
      "loss": 1.0616,
      "step": 70
    },
    {
      "epoch": 5.32,
      "learning_rate": 4.855652305297052e-05,
      "loss": 1.0397,
      "step": 75
    },
    {
      "epoch": 5.53,
      "learning_rate": 4.835984592901678e-05,
      "loss": 1.021,
      "step": 80
    },
    {
      "epoch": 5.57,
      "eval_loss": 1.0062576532363892,
      "eval_runtime": 2.1747,
      "eval_samples_per_second": 38.166,
      "eval_steps_per_second": 1.379,
      "step": 81
    },
    {
      "epoch": 6.17,
      "learning_rate": 4.815106302469513e-05,
      "loss": 0.9995,
      "step": 85
    },
    {
      "epoch": 6.38,
      "learning_rate": 4.793028253763633e-05,
      "loss": 0.9852,
      "step": 90
    },
    {
      "epoch": 6.55,
      "eval_loss": 0.9628537893295288,
      "eval_runtime": 2.1651,
      "eval_samples_per_second": 38.335,
      "eval_steps_per_second": 1.386,
      "step": 94
    },
    {
      "epoch": 7.02,
      "learning_rate": 4.769761888298222e-05,
      "loss": 0.9799,
      "step": 95
    },
    {
      "epoch": 7.23,
      "learning_rate": 4.74531926340924e-05,
      "loss": 0.9521,
      "step": 100
    },
    {
      "epoch": 7.45,
      "learning_rate": 4.719713046005938e-05,
      "loss": 0.9317,
      "step": 105
    },
    {
      "epoch": 7.57,
      "eval_loss": 0.9245708584785461,
      "eval_runtime": 2.1462,
      "eval_samples_per_second": 38.673,
      "eval_steps_per_second": 1.398,
      "step": 108
    },
    {
      "epoch": 8.09,
      "learning_rate": 4.6929565060064864e-05,
      "loss": 0.9231,
      "step": 110
    },
    {
      "epoch": 8.3,
      "learning_rate": 4.665063509461097e-05,
      "loss": 0.8929,
      "step": 115
    },
    {
      "epoch": 8.51,
      "learning_rate": 4.6360485113662216e-05,
      "loss": 0.8735,
      "step": 120
    },
    {
      "epoch": 8.55,
      "eval_loss": 0.8884902596473694,
      "eval_runtime": 2.1658,
      "eval_samples_per_second": 38.323,
      "eval_steps_per_second": 1.385,
      "step": 121
    },
    {
      "epoch": 9.15,
      "learning_rate": 4.605926548173529e-05,
      "loss": 0.8606,
      "step": 125
    },
    {
      "epoch": 9.36,
      "learning_rate": 4.574713229997563e-05,
      "loss": 0.8246,
      "step": 130
    },
    {
      "epoch": 9.57,
      "learning_rate": 4.542424732526105e-05,
      "loss": 0.8147,
      "step": 135
    },
    {
      "epoch": 9.57,
      "eval_loss": 0.8334853053092957,
      "eval_runtime": 2.1468,
      "eval_samples_per_second": 38.663,
      "eval_steps_per_second": 1.397,
      "step": 135
    },
    {
      "epoch": 10.21,
      "learning_rate": 4.509077788637446e-05,
      "loss": 0.7726,
      "step": 140
    },
    {
      "epoch": 10.43,
      "learning_rate": 4.4746896797288965e-05,
      "loss": 0.763,
      "step": 145
    },
    {
      "epoch": 10.55,
      "eval_loss": 0.7860499024391174,
      "eval_runtime": 2.1649,
      "eval_samples_per_second": 38.339,
      "eval_steps_per_second": 1.386,
      "step": 148
    },
    {
      "epoch": 11.06,
      "learning_rate": 4.43927822676105e-05,
      "loss": 0.7436,
      "step": 150
    },
    {
      "epoch": 11.28,
      "learning_rate": 4.402861781022412e-05,
      "loss": 0.7065,
      "step": 155
    },
    {
      "epoch": 11.49,
      "learning_rate": 4.365459214619214e-05,
      "loss": 0.6926,
      "step": 160
    },
    {
      "epoch": 11.57,
      "eval_loss": 0.7296475172042847,
      "eval_runtime": 2.173,
      "eval_samples_per_second": 38.196,
      "eval_steps_per_second": 1.381,
      "step": 162
    },
    {
      "epoch": 12.13,
      "learning_rate": 4.3270899106953105e-05,
      "loss": 0.6764,
      "step": 165
    },
    {
      "epoch": 12.34,
      "learning_rate": 4.2877737533872485e-05,
      "loss": 0.6473,
      "step": 170
    },
    {
      "epoch": 12.55,
      "learning_rate": 4.2475311175197056e-05,
      "loss": 0.6332,
      "step": 175
    },
    {
      "epoch": 12.55,
      "eval_loss": 0.686292827129364,
      "eval_runtime": 2.1677,
      "eval_samples_per_second": 38.289,
      "eval_steps_per_second": 1.384,
      "step": 175
    },
    {
      "epoch": 13.19,
      "learning_rate": 4.206382858046636e-05,
      "loss": 0.6181,
      "step": 180
    },
    {
      "epoch": 13.4,
      "learning_rate": 4.164350299243601e-05,
      "loss": 0.5898,
      "step": 185
    },
    {
      "epoch": 13.57,
      "eval_loss": 0.6491469740867615,
      "eval_runtime": 2.1549,
      "eval_samples_per_second": 38.517,
      "eval_steps_per_second": 1.392,
      "step": 189
    },
    {
      "epoch": 14.04,
      "learning_rate": 4.12145522365689e-05,
      "loss": 0.5682,
      "step": 190
    },
    {
      "epoch": 14.26,
      "learning_rate": 4.077719860815132e-05,
      "loss": 0.5651,
      "step": 195
    },
    {
      "epoch": 14.47,
      "learning_rate": 4.033166875709291e-05,
      "loss": 0.536,
      "step": 200
    },
    {
      "epoch": 14.55,
      "eval_loss": 0.6179705262184143,
      "eval_runtime": 2.1686,
      "eval_samples_per_second": 38.273,
      "eval_steps_per_second": 1.383,
      "step": 202
    },
    {
      "epoch": 15.11,
      "learning_rate": 3.9878193570469746e-05,
      "loss": 0.5464,
      "step": 205
    },
    {
      "epoch": 15.32,
      "learning_rate": 3.941700805287168e-05,
      "loss": 0.5064,
      "step": 210
    },
    {
      "epoch": 15.53,
      "learning_rate": 3.894835120461584e-05,
      "loss": 0.5263,
      "step": 215
    },
    {
      "epoch": 15.57,
      "eval_loss": 0.5763722062110901,
      "eval_runtime": 2.1517,
      "eval_samples_per_second": 38.575,
      "eval_steps_per_second": 1.394,
      "step": 216
    },
    {
      "epoch": 16.17,
      "learning_rate": 3.8472465897889394e-05,
      "loss": 0.4961,
      "step": 220
    },
    {
      "epoch": 16.38,
      "learning_rate": 3.798959875088584e-05,
      "loss": 0.5071,
      "step": 225
    },
    {
      "epoch": 16.55,
      "eval_loss": 0.5683054327964783,
      "eval_runtime": 2.1772,
      "eval_samples_per_second": 38.123,
      "eval_steps_per_second": 1.378,
      "step": 229
    },
    {
      "epoch": 17.02,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4835,
      "step": 230
    },
    {
      "epoch": 17.23,
      "learning_rate": 3.700392337014789e-05,
      "loss": 0.4797,
      "step": 235
    },
    {
      "epoch": 17.45,
      "learning_rate": 3.6501625943278805e-05,
      "loss": 0.4756,
      "step": 240
    },
    {
      "epoch": 17.57,
      "eval_loss": 0.5597423315048218,
      "eval_runtime": 2.1615,
      "eval_samples_per_second": 38.399,
      "eval_steps_per_second": 1.388,
      "step": 243
    },
    {
      "epoch": 18.09,
      "learning_rate": 3.599336802514764e-05,
      "loss": 0.4652,
      "step": 245
    },
    {
      "epoch": 18.3,
      "learning_rate": 3.547941301041661e-05,
      "loss": 0.4466,
      "step": 250
    },
    {
      "epoch": 18.51,
      "learning_rate": 3.496002724615604e-05,
      "loss": 0.4693,
      "step": 255
    },
    {
      "epoch": 18.55,
      "eval_loss": 0.5342021584510803,
      "eval_runtime": 2.1635,
      "eval_samples_per_second": 38.365,
      "eval_steps_per_second": 1.387,
      "step": 256
    },
    {
      "epoch": 19.15,
      "learning_rate": 3.443547989381536e-05,
      "loss": 0.4602,
      "step": 260
    },
    {
      "epoch": 19.36,
      "learning_rate": 3.390604278973543e-05,
      "loss": 0.4522,
      "step": 265
    },
    {
      "epoch": 19.57,
      "learning_rate": 3.3371990304274656e-05,
      "loss": 0.4414,
      "step": 270
    },
    {
      "epoch": 19.57,
      "eval_loss": 0.5385740399360657,
      "eval_runtime": 2.1714,
      "eval_samples_per_second": 38.224,
      "eval_steps_per_second": 1.382,
      "step": 270
    },
    {
      "epoch": 20.21,
      "learning_rate": 3.283359919962206e-05,
      "loss": 0.4492,
      "step": 275
    },
    {
      "epoch": 20.43,
      "learning_rate": 3.2291148486370626e-05,
      "loss": 0.4266,
      "step": 280
    },
    {
      "epoch": 20.55,
      "eval_loss": 0.5345889925956726,
      "eval_runtime": 2.1657,
      "eval_samples_per_second": 38.324,
      "eval_steps_per_second": 1.385,
      "step": 283
    },
    {
      "epoch": 21.06,
      "learning_rate": 3.1744919278925605e-05,
      "loss": 0.437,
      "step": 285
    },
    {
      "epoch": 21.28,
      "learning_rate": 3.11951946498225e-05,
      "loss": 0.4219,
      "step": 290
    },
    {
      "epoch": 21.49,
      "learning_rate": 3.0642259483030266e-05,
      "loss": 0.4286,
      "step": 295
    },
    {
      "epoch": 21.57,
      "eval_loss": 0.5155086517333984,
      "eval_runtime": 2.1707,
      "eval_samples_per_second": 38.237,
      "eval_steps_per_second": 1.382,
      "step": 297
    },
    {
      "epoch": 22.13,
      "learning_rate": 3.008640032631585e-05,
      "loss": 0.4214,
      "step": 300
    },
    {
      "epoch": 22.34,
      "learning_rate": 2.9527905242746394e-05,
      "loss": 0.4136,
      "step": 305
    },
    {
      "epoch": 22.55,
      "learning_rate": 2.8967063661406285e-05,
      "loss": 0.4256,
      "step": 310
    },
    {
      "epoch": 22.55,
      "eval_loss": 0.5108062624931335,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 38.209,
      "eval_steps_per_second": 1.381,
      "step": 310
    },
    {
      "epoch": 23.19,
      "learning_rate": 2.840416622740617e-05,
      "loss": 0.4066,
      "step": 315
    },
    {
      "epoch": 23.4,
      "learning_rate": 2.7839504651261872e-05,
      "loss": 0.418,
      "step": 320
    },
    {
      "epoch": 23.57,
      "eval_loss": 0.5229755640029907,
      "eval_runtime": 2.1749,
      "eval_samples_per_second": 38.163,
      "eval_steps_per_second": 1.379,
      "step": 324
    },
    {
      "epoch": 24.04,
      "learning_rate": 2.7273371557721176e-05,
      "loss": 0.4153,
      "step": 325
    },
    {
      "epoch": 24.26,
      "learning_rate": 2.6706060334116777e-05,
      "loss": 0.4139,
      "step": 330
    },
    {
      "epoch": 24.47,
      "learning_rate": 2.6137864978324095e-05,
      "loss": 0.407,
      "step": 335
    },
    {
      "epoch": 24.55,
      "eval_loss": 0.5164679884910583,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 38.209,
      "eval_steps_per_second": 1.381,
      "step": 337
    },
    {
      "epoch": 25.11,
      "learning_rate": 2.556907994640264e-05,
      "loss": 0.406,
      "step": 340
    },
    {
      "epoch": 25.32,
      "learning_rate": 2.5e-05,
      "loss": 0.3903,
      "step": 345
    },
    {
      "epoch": 25.53,
      "learning_rate": 2.4430920053597356e-05,
      "loss": 0.411,
      "step": 350
    },
    {
      "epoch": 25.57,
      "eval_loss": 0.5128384232521057,
      "eval_runtime": 2.1499,
      "eval_samples_per_second": 38.606,
      "eval_steps_per_second": 1.395,
      "step": 351
    },
    {
      "epoch": 26.17,
      "learning_rate": 2.3862135021675918e-05,
      "loss": 0.4,
      "step": 355
    },
    {
      "epoch": 26.38,
      "learning_rate": 2.329393966588323e-05,
      "loss": 0.3896,
      "step": 360
    },
    {
      "epoch": 26.55,
      "eval_loss": 0.5026661157608032,
      "eval_runtime": 2.1644,
      "eval_samples_per_second": 38.349,
      "eval_steps_per_second": 1.386,
      "step": 364
    },
    {
      "epoch": 27.02,
      "learning_rate": 2.2726628442278826e-05,
      "loss": 0.4005,
      "step": 365
    },
    {
      "epoch": 27.23,
      "learning_rate": 2.2160495348738123e-05,
      "loss": 0.4014,
      "step": 370
    },
    {
      "epoch": 27.45,
      "learning_rate": 2.159583377259384e-05,
      "loss": 0.39,
      "step": 375
    },
    {
      "epoch": 27.57,
      "eval_loss": 0.5062728524208069,
      "eval_runtime": 2.1531,
      "eval_samples_per_second": 38.549,
      "eval_steps_per_second": 1.393,
      "step": 378
    },
    {
      "epoch": 28.09,
      "learning_rate": 2.1032936338593718e-05,
      "loss": 0.3872,
      "step": 380
    },
    {
      "epoch": 28.3,
      "learning_rate": 2.0472094757253605e-05,
      "loss": 0.3929,
      "step": 385
    },
    {
      "epoch": 28.51,
      "learning_rate": 1.991359967368416e-05,
      "loss": 0.3928,
      "step": 390
    },
    {
      "epoch": 28.55,
      "eval_loss": 0.494553804397583,
      "eval_runtime": 2.1731,
      "eval_samples_per_second": 38.195,
      "eval_steps_per_second": 1.381,
      "step": 391
    },
    {
      "epoch": 29.15,
      "learning_rate": 1.9357740516969736e-05,
      "loss": 0.386,
      "step": 395
    },
    {
      "epoch": 29.36,
      "learning_rate": 1.8804805350177505e-05,
      "loss": 0.3948,
      "step": 400
    },
    {
      "epoch": 29.57,
      "learning_rate": 1.825508072107439e-05,
      "loss": 0.3818,
      "step": 405
    },
    {
      "epoch": 29.57,
      "eval_loss": 0.5063261985778809,
      "eval_runtime": 2.1689,
      "eval_samples_per_second": 38.269,
      "eval_steps_per_second": 1.383,
      "step": 405
    },
    {
      "epoch": 29.57,
      "step": 405,
      "total_flos": 2750244861247488.0,
      "train_loss": 0.7121148815861454,
      "train_runtime": 1989.4524,
      "train_samples_per_second": 11.249,
      "train_steps_per_second": 0.347
    }
  ],
  "logging_steps": 5,
  "max_steps": 690,
  "num_train_epochs": 30,
  "save_steps": 500,
  "total_flos": 2750244861247488.0,
  "trial_name": null,
  "trial_params": null
}