{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 788,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "learning_rate": 9.87341772151899e-06,
      "logits/chosen": -2.6043291091918945,
      "logits/rejected": -2.702824592590332,
      "logps/chosen": -68.50479125976562,
      "logps/rejected": -63.61880874633789,
      "loss": 0.6809,
      "rewards/accuracies": 0.7403846383094788,
      "rewards/chosen": -0.0004181800759397447,
      "rewards/margins": 0.025023292750120163,
      "rewards/rejected": -0.025441471487283707,
      "step": 39
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.974683544303798e-05,
      "logits/chosen": -2.6380443572998047,
      "logits/rejected": -2.621236562728882,
      "logps/chosen": -65.62651062011719,
      "logps/rejected": -66.03771209716797,
      "loss": 0.5601,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.02834504097700119,
      "rewards/margins": 0.32108908891677856,
      "rewards/rejected": -0.34943410754203796,
      "step": 78
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.898448519040903e-05,
      "logits/chosen": -2.6502482891082764,
      "logits/rejected": -2.6570136547088623,
      "logps/chosen": -67.91175079345703,
      "logps/rejected": -78.51203155517578,
      "loss": 0.3085,
      "rewards/accuracies": 0.9230769276618958,
      "rewards/chosen": -0.147563174366951,
      "rewards/margins": 1.4500247240066528,
      "rewards/rejected": -1.5975879430770874,
      "step": 117
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.7884344146685474e-05,
      "logits/chosen": -2.6012794971466064,
      "logits/rejected": -2.565227746963501,
      "logps/chosen": -71.57939910888672,
      "logps/rejected": -97.67864990234375,
      "loss": 0.1548,
      "rewards/accuracies": 0.9583333134651184,
      "rewards/chosen": -0.31067565083503723,
      "rewards/margins": 3.0861926078796387,
      "rewards/rejected": -3.3968684673309326,
      "step": 156
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.678420310296192e-05,
      "logits/chosen": -2.516798973083496,
      "logits/rejected": -2.4578933715820312,
      "logps/chosen": -72.1274185180664,
      "logps/rejected": -110.97193908691406,
      "loss": 0.1164,
      "rewards/accuracies": 0.9679487347602844,
      "rewards/chosen": -0.4628397524356842,
      "rewards/margins": 4.33509635925293,
      "rewards/rejected": -4.797935962677002,
      "step": 195
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.5684062059238364e-05,
      "logits/chosen": -2.4704339504241943,
      "logits/rejected": -2.3934192657470703,
      "logps/chosen": -75.9949951171875,
      "logps/rejected": -127.63720703125,
      "loss": 0.0745,
      "rewards/accuracies": 0.9722222685813904,
      "rewards/chosen": -0.9339821934700012,
      "rewards/margins": 5.540595054626465,
      "rewards/rejected": -6.474576950073242,
      "step": 234
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.4612129760225671e-05,
      "logits/chosen": -2.353947877883911,
      "logits/rejected": -2.2041804790496826,
      "logps/chosen": -81.29566955566406,
      "logps/rejected": -138.798583984375,
      "loss": 0.0749,
      "rewards/accuracies": 0.9615384340286255,
      "rewards/chosen": -1.3812835216522217,
      "rewards/margins": 6.240307331085205,
      "rewards/rejected": -7.621591567993164,
      "step": 273
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.3511988716502115e-05,
      "logits/chosen": -2.2809927463531494,
      "logits/rejected": -2.028799295425415,
      "logps/chosen": -90.3890380859375,
      "logps/rejected": -157.99842834472656,
      "loss": 0.077,
      "rewards/accuracies": 0.9807692170143127,
      "rewards/chosen": -2.2647416591644287,
      "rewards/margins": 7.266026973724365,
      "rewards/rejected": -9.530769348144531,
      "step": 312
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.2411847672778563e-05,
      "logits/chosen": -2.184739828109741,
      "logits/rejected": -1.9148027896881104,
      "logps/chosen": -90.59001159667969,
      "logps/rejected": -169.61471557617188,
      "loss": 0.0682,
      "rewards/accuracies": 0.9743589758872986,
      "rewards/chosen": -2.3255884647369385,
      "rewards/margins": 8.284488677978516,
      "rewards/rejected": -10.610077857971191,
      "step": 351
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.1311706629055009e-05,
      "logits/chosen": -2.1274867057800293,
      "logits/rejected": -1.8261984586715698,
      "logps/chosen": -92.77806854248047,
      "logps/rejected": -179.1676788330078,
      "loss": 0.0676,
      "rewards/accuracies": 0.9679487347602844,
      "rewards/chosen": -2.598703384399414,
      "rewards/margins": 9.01236343383789,
      "rewards/rejected": -11.611067771911621,
      "step": 390
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.0239774330042314e-05,
      "logits/chosen": -2.037000894546509,
      "logits/rejected": -1.6896817684173584,
      "logps/chosen": -102.07405853271484,
      "logps/rejected": -199.1004180908203,
      "loss": 0.0448,
      "rewards/accuracies": 0.9903846383094788,
      "rewards/chosen": -3.419679641723633,
      "rewards/margins": 10.136916160583496,
      "rewards/rejected": -13.556596755981445,
      "step": 429
    },
    {
      "epoch": 2.38,
      "learning_rate": 9.13963328631876e-06,
      "logits/chosen": -2.0118300914764404,
      "logits/rejected": -1.6913644075393677,
      "logps/chosen": -97.3907241821289,
      "logps/rejected": -201.25863647460938,
      "loss": 0.0125,
      "rewards/accuracies": 0.9967948794364929,
      "rewards/chosen": -3.0104339122772217,
      "rewards/margins": 10.906888008117676,
      "rewards/rejected": -13.91732120513916,
      "step": 468
    },
    {
      "epoch": 2.57,
      "learning_rate": 8.039492242595204e-06,
      "logits/chosen": -1.9532654285430908,
      "logits/rejected": -1.5922006368637085,
      "logps/chosen": -102.87281799316406,
      "logps/rejected": -213.3865966796875,
      "loss": 0.0172,
      "rewards/accuracies": 0.9935897588729858,
      "rewards/chosen": -3.6911683082580566,
      "rewards/margins": 11.436351776123047,
      "rewards/rejected": -15.127519607543945,
      "step": 507
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.939351198871651e-06,
      "logits/chosen": -2.047797679901123,
      "logits/rejected": -1.7366411685943604,
      "logps/chosen": -97.63482666015625,
      "logps/rejected": -212.34762573242188,
      "loss": 0.0284,
      "rewards/accuracies": 0.9935897588729858,
      "rewards/chosen": -2.9373013973236084,
      "rewards/margins": 11.817158699035645,
      "rewards/rejected": -14.754460334777832,
      "step": 546
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.8392101551480965e-06,
      "logits/chosen": -1.9487961530685425,
      "logits/rejected": -1.665400743484497,
      "logps/chosen": -107.47830200195312,
      "logps/rejected": -230.8419647216797,
      "loss": 0.0048,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -4.116791725158691,
      "rewards/margins": 12.73591136932373,
      "rewards/rejected": -16.85270118713379,
      "step": 585
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.739069111424542e-06,
      "logits/chosen": -1.9433237314224243,
      "logits/rejected": -1.5677200555801392,
      "logps/chosen": -105.38491821289062,
      "logps/rejected": -234.572265625,
      "loss": 0.0039,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.7342302799224854,
      "rewards/margins": 13.451081275939941,
      "rewards/rejected": -17.185312271118164,
      "step": 624
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.6389280677009874e-06,
      "logits/chosen": -1.9452707767486572,
      "logits/rejected": -1.6384880542755127,
      "logps/chosen": -105.9559326171875,
      "logps/rejected": -233.3934326171875,
      "loss": 0.037,
      "rewards/accuracies": 0.9967948794364929,
      "rewards/chosen": -3.802997350692749,
      "rewards/margins": 13.106091499328613,
      "rewards/rejected": -16.909088134765625,
      "step": 663
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.538787023977433e-06,
      "logits/chosen": -1.9369711875915527,
      "logits/rejected": -1.6055175065994263,
      "logps/chosen": -101.39398956298828,
      "logps/rejected": -226.4627685546875,
      "loss": 0.0036,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -3.4717466831207275,
      "rewards/margins": 12.892216682434082,
      "rewards/rejected": -16.363964080810547,
      "step": 702
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.4386459802538789e-06,
      "logits/chosen": -1.9032098054885864,
      "logits/rejected": -1.5705137252807617,
      "logps/chosen": -106.39134979248047,
      "logps/rejected": -233.3502655029297,
      "loss": 0.005,
      "rewards/accuracies": 0.9967948794364929,
      "rewards/chosen": -3.9461498260498047,
      "rewards/margins": 13.194243431091309,
      "rewards/rejected": -17.140390396118164,
      "step": 741
    },
    {
      "epoch": 3.96,
      "learning_rate": 3.3850493653032445e-07,
      "logits/chosen": -1.9001864194869995,
      "logits/rejected": -1.5598957538604736,
      "logps/chosen": -109.82083892822266,
      "logps/rejected": -238.86415100097656,
      "loss": 0.0129,
      "rewards/accuracies": 0.9967948794364929,
      "rewards/chosen": -4.283285140991211,
      "rewards/margins": 13.32726764678955,
      "rewards/rejected": -17.610553741455078,
      "step": 780
    }
  ],
  "logging_steps": 39,
  "max_steps": 788,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}