{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 24.96185286103542,
  "eval_steps": 500,
  "global_step": 8825,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 7.952069716775599e-06,
      "loss": 0.9545,
      "step": 73
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.5904139433551197e-05,
      "loss": 0.9152,
      "step": 146
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.38562091503268e-05,
      "loss": 0.7448,
      "step": 219
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.1808278867102395e-05,
      "loss": 0.5783,
      "step": 292
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.9760348583877995e-05,
      "loss": 0.4973,
      "step": 365
    },
    {
      "epoch": 1.23,
      "learning_rate": 4.77124183006536e-05,
      "loss": 0.4414,
      "step": 438
    },
    {
      "epoch": 1.43,
      "learning_rate": 5.5664488017429196e-05,
      "loss": 0.4208,
      "step": 511
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.361655773420479e-05,
      "loss": 0.3861,
      "step": 584
    },
    {
      "epoch": 1.83,
      "learning_rate": 7.156862745098039e-05,
      "loss": 0.3802,
      "step": 657
    },
    {
      "epoch": 2.07,
      "learning_rate": 7.952069716775599e-05,
      "loss": 0.335,
      "step": 730
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.74727668845316e-05,
      "loss": 0.3216,
      "step": 803
    },
    {
      "epoch": 2.46,
      "learning_rate": 9.54248366013072e-05,
      "loss": 0.2991,
      "step": 876
    },
    {
      "epoch": 2.66,
      "learning_rate": 9.962456097856366e-05,
      "loss": 0.2919,
      "step": 949
    },
    {
      "epoch": 2.86,
      "learning_rate": 9.87404626377619e-05,
      "loss": 0.3062,
      "step": 1022
    },
    {
      "epoch": 3.1,
      "learning_rate": 9.785636429696016e-05,
      "loss": 0.2661,
      "step": 1095
    },
    {
      "epoch": 3.3,
      "learning_rate": 9.697226595615842e-05,
      "loss": 0.2547,
      "step": 1168
    },
    {
      "epoch": 3.5,
      "learning_rate": 9.608816761535667e-05,
      "loss": 0.2426,
      "step": 1241
    },
    {
      "epoch": 3.69,
      "learning_rate": 9.520406927455493e-05,
      "loss": 0.224,
      "step": 1314
    },
    {
      "epoch": 3.89,
      "learning_rate": 9.431997093375319e-05,
      "loss": 0.2164,
      "step": 1387
    },
    {
      "epoch": 4.13,
      "learning_rate": 9.343587259295143e-05,
      "loss": 0.1952,
      "step": 1460
    },
    {
      "epoch": 4.33,
      "learning_rate": 9.255177425214969e-05,
      "loss": 0.2025,
      "step": 1533
    },
    {
      "epoch": 4.53,
      "learning_rate": 9.166767591134794e-05,
      "loss": 0.1837,
      "step": 1606
    },
    {
      "epoch": 4.73,
      "learning_rate": 9.078357757054621e-05,
      "loss": 0.1911,
      "step": 1679
    },
    {
      "epoch": 4.93,
      "learning_rate": 8.989947922974447e-05,
      "loss": 0.1737,
      "step": 1752
    },
    {
      "epoch": 5.16,
      "learning_rate": 8.901538088894273e-05,
      "loss": 0.1633,
      "step": 1825
    },
    {
      "epoch": 5.36,
      "learning_rate": 8.813128254814099e-05,
      "loss": 0.1485,
      "step": 1898
    },
    {
      "epoch": 5.56,
      "learning_rate": 8.724718420733923e-05,
      "loss": 0.1479,
      "step": 1971
    },
    {
      "epoch": 5.76,
      "learning_rate": 8.636308586653749e-05,
      "loss": 0.1625,
      "step": 2044
    },
    {
      "epoch": 5.96,
      "learning_rate": 8.547898752573574e-05,
      "loss": 0.1602,
      "step": 2117
    },
    {
      "epoch": 6.2,
      "learning_rate": 8.4594889184934e-05,
      "loss": 0.1307,
      "step": 2190
    },
    {
      "epoch": 6.4,
      "learning_rate": 8.371079084413226e-05,
      "loss": 0.1279,
      "step": 2263
    },
    {
      "epoch": 6.59,
      "learning_rate": 8.282669250333051e-05,
      "loss": 0.1231,
      "step": 2336
    },
    {
      "epoch": 6.79,
      "learning_rate": 8.194259416252876e-05,
      "loss": 0.1426,
      "step": 2409
    },
    {
      "epoch": 7.03,
      "learning_rate": 8.105849582172701e-05,
      "loss": 0.1245,
      "step": 2482
    },
    {
      "epoch": 7.23,
      "learning_rate": 8.017439748092528e-05,
      "loss": 0.1186,
      "step": 2555
    },
    {
      "epoch": 7.43,
      "learning_rate": 7.929029914012354e-05,
      "loss": 0.1081,
      "step": 2628
    },
    {
      "epoch": 7.63,
      "learning_rate": 7.84062007993218e-05,
      "loss": 0.1025,
      "step": 2701
    },
    {
      "epoch": 7.83,
      "learning_rate": 7.752210245852006e-05,
      "loss": 0.1173,
      "step": 2774
    },
    {
      "epoch": 8.06,
      "learning_rate": 7.66380041177183e-05,
      "loss": 0.0996,
      "step": 2847
    },
    {
      "epoch": 8.26,
      "learning_rate": 7.575390577691656e-05,
      "loss": 0.0861,
      "step": 2920
    },
    {
      "epoch": 8.46,
      "learning_rate": 7.486980743611481e-05,
      "loss": 0.1018,
      "step": 2993
    },
    {
      "epoch": 8.66,
      "learning_rate": 7.398570909531307e-05,
      "loss": 0.0977,
      "step": 3066
    },
    {
      "epoch": 8.86,
      "learning_rate": 7.310161075451133e-05,
      "loss": 0.1002,
      "step": 3139
    },
    {
      "epoch": 9.1,
      "learning_rate": 7.221751241370958e-05,
      "loss": 0.0839,
      "step": 3212
    },
    {
      "epoch": 9.29,
      "learning_rate": 7.133341407290783e-05,
      "loss": 0.0884,
      "step": 3285
    },
    {
      "epoch": 9.49,
      "learning_rate": 7.044931573210608e-05,
      "loss": 0.0825,
      "step": 3358
    },
    {
      "epoch": 9.69,
      "learning_rate": 6.956521739130436e-05,
      "loss": 0.0811,
      "step": 3431
    },
    {
      "epoch": 9.89,
      "learning_rate": 6.868111905050261e-05,
      "loss": 0.0851,
      "step": 3504
    },
    {
      "epoch": 10.13,
      "learning_rate": 6.779702070970087e-05,
      "loss": 0.0716,
      "step": 3577
    },
    {
      "epoch": 10.33,
      "learning_rate": 6.691292236889913e-05,
      "loss": 0.0689,
      "step": 3650
    },
    {
      "epoch": 10.53,
      "learning_rate": 6.602882402809738e-05,
      "loss": 0.0747,
      "step": 3723
    },
    {
      "epoch": 10.72,
      "learning_rate": 6.514472568729563e-05,
      "loss": 0.0728,
      "step": 3796
    },
    {
      "epoch": 10.92,
      "learning_rate": 6.426062734649388e-05,
      "loss": 0.0762,
      "step": 3869
    },
    {
      "epoch": 11.16,
      "learning_rate": 6.337652900569214e-05,
      "loss": 0.0693,
      "step": 3942
    },
    {
      "epoch": 11.36,
      "learning_rate": 6.24924306648904e-05,
      "loss": 0.0611,
      "step": 4015
    },
    {
      "epoch": 11.56,
      "learning_rate": 6.160833232408865e-05,
      "loss": 0.0616,
      "step": 4088
    },
    {
      "epoch": 11.76,
      "learning_rate": 6.072423398328692e-05,
      "loss": 0.064,
      "step": 4161
    },
    {
      "epoch": 11.96,
      "learning_rate": 5.984013564248516e-05,
      "loss": 0.0637,
      "step": 4234
    },
    {
      "epoch": 12.19,
      "learning_rate": 5.895603730168342e-05,
      "loss": 0.0548,
      "step": 4307
    },
    {
      "epoch": 12.39,
      "learning_rate": 5.8071938960881676e-05,
      "loss": 0.0583,
      "step": 4380
    },
    {
      "epoch": 12.59,
      "learning_rate": 5.718784062007993e-05,
      "loss": 0.0547,
      "step": 4453
    },
    {
      "epoch": 12.79,
      "learning_rate": 5.630374227927819e-05,
      "loss": 0.0543,
      "step": 4526
    },
    {
      "epoch": 13.03,
      "learning_rate": 5.5419643938476454e-05,
      "loss": 0.0543,
      "step": 4599
    },
    {
      "epoch": 13.23,
      "learning_rate": 5.45355455976747e-05,
      "loss": 0.0479,
      "step": 4672
    },
    {
      "epoch": 13.43,
      "learning_rate": 5.3651447256872954e-05,
      "loss": 0.0517,
      "step": 4745
    },
    {
      "epoch": 13.62,
      "learning_rate": 5.276734891607121e-05,
      "loss": 0.0466,
      "step": 4818
    },
    {
      "epoch": 13.82,
      "learning_rate": 5.188325057526947e-05,
      "loss": 0.0499,
      "step": 4891
    },
    {
      "epoch": 14.06,
      "learning_rate": 5.0999152234467725e-05,
      "loss": 0.0479,
      "step": 4964
    },
    {
      "epoch": 14.26,
      "learning_rate": 5.011505389366599e-05,
      "loss": 0.0408,
      "step": 5037
    },
    {
      "epoch": 14.46,
      "learning_rate": 4.923095555286424e-05,
      "loss": 0.0425,
      "step": 5110
    },
    {
      "epoch": 14.66,
      "learning_rate": 4.8346857212062496e-05,
      "loss": 0.0452,
      "step": 5183
    },
    {
      "epoch": 14.86,
      "learning_rate": 4.746275887126075e-05,
      "loss": 0.0462,
      "step": 5256
    },
    {
      "epoch": 15.09,
      "learning_rate": 4.6578660530459004e-05,
      "loss": 0.0452,
      "step": 5329
    },
    {
      "epoch": 15.29,
      "learning_rate": 4.569456218965726e-05,
      "loss": 0.0371,
      "step": 5402
    },
    {
      "epoch": 15.49,
      "learning_rate": 4.481046384885552e-05,
      "loss": 0.0404,
      "step": 5475
    },
    {
      "epoch": 15.69,
      "learning_rate": 4.3926365508053775e-05,
      "loss": 0.0418,
      "step": 5548
    },
    {
      "epoch": 15.89,
      "learning_rate": 4.304226716725203e-05,
      "loss": 0.0405,
      "step": 5621
    },
    {
      "epoch": 16.13,
      "learning_rate": 4.215816882645028e-05,
      "loss": 0.0365,
      "step": 5694
    },
    {
      "epoch": 16.32,
      "learning_rate": 4.127407048564854e-05,
      "loss": 0.0368,
      "step": 5767
    },
    {
      "epoch": 16.52,
      "learning_rate": 4.03899721448468e-05,
      "loss": 0.0348,
      "step": 5840
    },
    {
      "epoch": 16.72,
      "learning_rate": 3.950587380404506e-05,
      "loss": 0.0347,
      "step": 5913
    },
    {
      "epoch": 16.92,
      "learning_rate": 3.862177546324331e-05,
      "loss": 0.0409,
      "step": 5986
    },
    {
      "epoch": 17.16,
      "learning_rate": 3.773767712244157e-05,
      "loss": 0.035,
      "step": 6059
    },
    {
      "epoch": 17.36,
      "learning_rate": 3.6853578781639824e-05,
      "loss": 0.0317,
      "step": 6132
    },
    {
      "epoch": 17.56,
      "learning_rate": 3.5969480440838074e-05,
      "loss": 0.0339,
      "step": 6205
    },
    {
      "epoch": 17.75,
      "learning_rate": 3.508538210003634e-05,
      "loss": 0.034,
      "step": 6278
    },
    {
      "epoch": 17.95,
      "learning_rate": 3.4201283759234595e-05,
      "loss": 0.0336,
      "step": 6351
    },
    {
      "epoch": 18.19,
      "learning_rate": 3.3317185418432845e-05,
      "loss": 0.0306,
      "step": 6424
    },
    {
      "epoch": 18.39,
      "learning_rate": 3.24330870776311e-05,
      "loss": 0.0312,
      "step": 6497
    },
    {
      "epoch": 18.59,
      "learning_rate": 3.154898873682936e-05,
      "loss": 0.0298,
      "step": 6570
    },
    {
      "epoch": 18.79,
      "learning_rate": 3.066489039602761e-05,
      "loss": 0.0322,
      "step": 6643
    },
    {
      "epoch": 19.02,
      "learning_rate": 2.978079205522587e-05,
      "loss": 0.0301,
      "step": 6716
    },
    {
      "epoch": 19.22,
      "learning_rate": 2.8896693714424127e-05,
      "loss": 0.0297,
      "step": 6789
    },
    {
      "epoch": 19.42,
      "learning_rate": 2.801259537362238e-05,
      "loss": 0.0282,
      "step": 6862
    },
    {
      "epoch": 19.62,
      "learning_rate": 2.7128497032820637e-05,
      "loss": 0.0283,
      "step": 6935
    },
    {
      "epoch": 19.82,
      "learning_rate": 2.6244398692018894e-05,
      "loss": 0.0293,
      "step": 7008
    },
    {
      "epoch": 20.06,
      "learning_rate": 2.5360300351217148e-05,
      "loss": 0.0294,
      "step": 7081
    },
    {
      "epoch": 20.26,
      "learning_rate": 2.4476202010415405e-05,
      "loss": 0.0264,
      "step": 7154
    },
    {
      "epoch": 20.46,
      "learning_rate": 2.3592103669613662e-05,
      "loss": 0.026,
      "step": 7227
    },
    {
      "epoch": 20.65,
      "learning_rate": 2.270800532881192e-05,
      "loss": 0.0294,
      "step": 7300
    },
    {
      "epoch": 20.85,
      "learning_rate": 2.1823906988010173e-05,
      "loss": 0.0285,
      "step": 7373
    },
    {
      "epoch": 21.09,
      "learning_rate": 2.093980864720843e-05,
      "loss": 0.0269,
      "step": 7446
    },
    {
      "epoch": 21.29,
      "learning_rate": 2.0055710306406687e-05,
      "loss": 0.0258,
      "step": 7519
    },
    {
      "epoch": 21.49,
      "learning_rate": 1.917161196560494e-05,
      "loss": 0.0254,
      "step": 7592
    },
    {
      "epoch": 21.69,
      "learning_rate": 1.8287513624803197e-05,
      "loss": 0.0278,
      "step": 7665
    },
    {
      "epoch": 21.89,
      "learning_rate": 1.7403415284001454e-05,
      "loss": 0.0262,
      "step": 7738
    },
    {
      "epoch": 22.12,
      "learning_rate": 1.651931694319971e-05,
      "loss": 0.0245,
      "step": 7811
    },
    {
      "epoch": 22.32,
      "learning_rate": 1.5635218602397965e-05,
      "loss": 0.0249,
      "step": 7884
    },
    {
      "epoch": 22.52,
      "learning_rate": 1.4751120261596222e-05,
      "loss": 0.0244,
      "step": 7957
    },
    {
      "epoch": 22.72,
      "learning_rate": 1.3867021920794479e-05,
      "loss": 0.0255,
      "step": 8030
    },
    {
      "epoch": 22.92,
      "learning_rate": 1.2982923579992734e-05,
      "loss": 0.0256,
      "step": 8103
    },
    {
      "epoch": 23.16,
      "learning_rate": 1.209882523919099e-05,
      "loss": 0.0244,
      "step": 8176
    },
    {
      "epoch": 23.35,
      "learning_rate": 1.1214726898389245e-05,
      "loss": 0.0235,
      "step": 8249
    },
    {
      "epoch": 23.55,
      "learning_rate": 1.0330628557587502e-05,
      "loss": 0.0252,
      "step": 8322
    },
    {
      "epoch": 23.75,
      "learning_rate": 9.446530216785757e-06,
      "loss": 0.0248,
      "step": 8395
    },
    {
      "epoch": 23.95,
      "learning_rate": 8.562431875984014e-06,
      "loss": 0.024,
      "step": 8468
    },
    {
      "epoch": 24.19,
      "learning_rate": 7.67833353518227e-06,
      "loss": 0.0244,
      "step": 8541
    },
    {
      "epoch": 24.39,
      "learning_rate": 6.794235194380526e-06,
      "loss": 0.023,
      "step": 8614
    },
    {
      "epoch": 24.59,
      "learning_rate": 5.910136853578782e-06,
      "loss": 0.024,
      "step": 8687
    },
    {
      "epoch": 24.78,
      "learning_rate": 5.026038512777038e-06,
      "loss": 0.0238,
      "step": 8760
    }
  ],
  "logging_steps": 73,
  "max_steps": 9175,
  "num_train_epochs": 25,
  "save_steps": 500,
  "total_flos": 7.7154619490304e+17,
  "trial_name": null,
  "trial_params": null
}