{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 939,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003194888178913738,
      "grad_norm": 66.45112919433747,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.4296875,
      "logps/chosen": -102.0,
      "logps/rejected": -229.0,
      "loss": 0.6914,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03194888178913738,
      "grad_norm": 68.56300885241505,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -1.3984375,
      "logits/rejected": -1.4375,
      "logps/chosen": -109.5,
      "logps/rejected": -240.0,
      "loss": 0.6873,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.006591796875,
      "rewards/margins": 0.0233154296875,
      "rewards/rejected": -0.0167236328125,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 63.45201175948467,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -1.4453125,
      "logits/rejected": -1.46875,
      "logps/chosen": -130.0,
      "logps/rejected": -247.0,
      "loss": 0.6476,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.0150146484375,
      "rewards/margins": 0.08642578125,
      "rewards/rejected": -0.0712890625,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 46.668149689919,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.484375,
      "logps/chosen": -122.5,
      "logps/rejected": -254.0,
      "loss": 0.484,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.0849609375,
      "rewards/margins": 0.51953125,
      "rewards/rejected": -0.435546875,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 12.835434682059827,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.484375,
      "logps/chosen": -130.0,
      "logps/rejected": -260.0,
      "loss": 0.2446,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.20703125,
      "rewards/margins": 1.515625,
      "rewards/rejected": -1.3125,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 1.4679619391009096,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.4296875,
      "logps/chosen": -109.0,
      "logps/rejected": -276.0,
      "loss": 0.0314,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.5234375,
      "rewards/margins": 4.03125,
      "rewards/rejected": -3.5,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 0.09784909559056462,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -1.3125,
      "logits/rejected": -1.4453125,
      "logps/chosen": -113.0,
      "logps/rejected": -310.0,
      "loss": 0.0027,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.73828125,
      "rewards/margins": 7.0625,
      "rewards/rejected": -6.3125,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 0.7877640759740914,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -1.34375,
      "logits/rejected": -1.3671875,
      "logps/chosen": -105.0,
      "logps/rejected": -326.0,
      "loss": 0.0012,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.91015625,
      "rewards/margins": 8.75,
      "rewards/rejected": -7.84375,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 0.12467848392417441,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.4453125,
      "logps/chosen": -119.0,
      "logps/rejected": -334.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94140625,
      "rewards/margins": 9.875,
      "rewards/rejected": -8.9375,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 0.012116127397948486,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -1.375,
      "logits/rejected": -1.40625,
      "logps/chosen": -116.5,
      "logps/rejected": -340.0,
      "loss": 0.0004,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0,
      "rewards/margins": 10.375,
      "rewards/rejected": -9.375,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 0.03810066891544163,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.3828125,
      "logps/chosen": -105.0,
      "logps/rejected": -344.0,
      "loss": 0.001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0078125,
      "rewards/margins": 10.5625,
      "rewards/rejected": -9.5625,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 0.015562454644345447,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -1.3984375,
      "logits/rejected": -1.328125,
      "logps/chosen": -114.5,
      "logps/rejected": -350.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94140625,
      "rewards/margins": 11.25,
      "rewards/rejected": -10.3125,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.007718916846088982,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -1.390625,
      "logits/rejected": -1.375,
      "logps/chosen": -108.0,
      "logps/rejected": -348.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.98828125,
      "rewards/margins": 11.3125,
      "rewards/rejected": -10.3125,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 0.006285136635001001,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -1.375,
      "logits/rejected": -1.3671875,
      "logps/chosen": -99.5,
      "logps/rejected": -362.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0078125,
      "rewards/margins": 12.375,
      "rewards/rejected": -11.375,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 0.015739482905232707,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -1.4375,
      "logits/rejected": -1.3671875,
      "logps/chosen": -111.5,
      "logps/rejected": -358.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.8828125,
      "rewards/margins": 11.875,
      "rewards/rejected": -11.0,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 0.009915409173787878,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -1.4375,
      "logits/rejected": -1.3828125,
      "logps/chosen": -108.5,
      "logps/rejected": -360.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1328125,
      "rewards/margins": 12.3125,
      "rewards/rejected": -11.1875,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 0.0029912599491742223,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -1.46875,
      "logits/rejected": -1.328125,
      "logps/chosen": -110.5,
      "logps/rejected": -356.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.046875,
      "rewards/margins": 12.5,
      "rewards/rejected": -11.4375,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 0.0029017468182494094,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -1.3359375,
      "logits/rejected": -1.359375,
      "logps/chosen": -92.0,
      "logps/rejected": -364.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2109375,
      "rewards/margins": 13.0,
      "rewards/rejected": -11.8125,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.18268301096109826,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.3671875,
      "logps/chosen": -98.5,
      "logps/rejected": -358.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0546875,
      "rewards/margins": 12.625,
      "rewards/rejected": -11.625,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 0.004003353039173915,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.4140625,
      "logps/chosen": -101.5,
      "logps/rejected": -354.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0703125,
      "rewards/margins": 12.6875,
      "rewards/rejected": -11.625,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 0.001476927281312559,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.359375,
      "logps/chosen": -98.0,
      "logps/rejected": -364.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 12.75,
      "rewards/rejected": -11.625,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.0076234997609973786,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -1.3828125,
      "logits/rejected": -1.3984375,
      "logps/chosen": -104.0,
      "logps/rejected": -358.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.15625,
      "rewards/margins": 12.6875,
      "rewards/rejected": -11.5,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 0.005396630147279373,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -1.296875,
      "logits/rejected": -1.359375,
      "logps/chosen": -108.0,
      "logps/rejected": -366.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1015625,
      "rewards/margins": 12.9375,
      "rewards/rejected": -11.875,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 0.0017048710378701922,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.34375,
      "logps/chosen": -131.0,
      "logps/rejected": -360.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.91796875,
      "rewards/margins": 12.5625,
      "rewards/rejected": -11.625,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.0019780892104056848,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.375,
      "logps/chosen": -98.5,
      "logps/rejected": -368.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2265625,
      "rewards/margins": 13.4375,
      "rewards/rejected": -12.25,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 0.0020127890370078044,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -1.484375,
      "logits/rejected": -1.40625,
      "logps/chosen": -99.5,
      "logps/rejected": -364.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 13.1875,
      "rewards/rejected": -12.0625,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 0.00864667456447465,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -1.484375,
      "logits/rejected": -1.453125,
      "logps/chosen": -113.5,
      "logps/rejected": -376.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0703125,
      "rewards/margins": 13.625,
      "rewards/rejected": -12.5,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.0007347459135287065,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.40625,
      "logps/chosen": -98.0,
      "logps/rejected": -366.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.21875,
      "rewards/margins": 13.0625,
      "rewards/rejected": -11.875,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 0.0029007856909616142,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -1.421875,
      "logits/rejected": -1.40625,
      "logps/chosen": -102.0,
      "logps/rejected": -372.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.203125,
      "rewards/margins": 13.625,
      "rewards/rejected": -12.4375,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 0.0014514297985436918,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -1.359375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -111.0,
      "logps/rejected": -368.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1015625,
      "rewards/margins": 13.3125,
      "rewards/rejected": -12.1875,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.0033845991144841616,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -1.34375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -115.0,
      "logps/rejected": -374.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.09375,
      "rewards/margins": 13.8125,
      "rewards/rejected": -12.6875,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 0.0007283116674165061,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -1.390625,
      "logits/rejected": -1.3359375,
      "logps/chosen": -116.0,
      "logps/rejected": -370.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.203125,
      "rewards/margins": 13.875,
      "rewards/rejected": -12.625,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.2578125,
      "eval_logits/rejected": -1.3125,
      "eval_logps/chosen": -90.0,
      "eval_logps/rejected": -382.0,
      "eval_loss": 1.5126615835470147e-05,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 1.140625,
      "eval_rewards/margins": 14.375,
      "eval_rewards/rejected": -13.1875,
      "eval_runtime": 11.1731,
      "eval_samples_per_second": 17.9,
      "eval_steps_per_second": 0.627,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 0.0007905572028997735,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -1.4609375,
      "logits/rejected": -1.359375,
      "logps/chosen": -120.0,
      "logps/rejected": -376.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1171875,
      "rewards/margins": 13.9375,
      "rewards/rejected": -12.8125,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.07616357831793605,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -1.3359375,
      "logits/rejected": -1.3203125,
      "logps/chosen": -103.0,
      "logps/rejected": -374.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2578125,
      "rewards/margins": 14.0,
      "rewards/rejected": -12.6875,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 0.0014919309897083502,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -1.359375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -108.0,
      "logps/rejected": -376.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2421875,
      "rewards/margins": 14.1875,
      "rewards/rejected": -13.0,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 0.001996979439874143,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.3671875,
      "logps/chosen": -102.5,
      "logps/rejected": -376.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1640625,
      "rewards/margins": 14.375,
      "rewards/rejected": -13.1875,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.0018549121455385391,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.4140625,
      "logps/chosen": -107.5,
      "logps/rejected": -372.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.265625,
      "rewards/margins": 13.9375,
      "rewards/rejected": -12.6875,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 0.00200072946427404,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -1.421875,
      "logits/rejected": -1.375,
      "logps/chosen": -116.5,
      "logps/rejected": -378.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2265625,
      "rewards/margins": 14.4375,
      "rewards/rejected": -13.25,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 0.015530216505332267,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -1.296875,
      "logits/rejected": -1.3671875,
      "logps/chosen": -98.0,
      "logps/rejected": -376.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.25,
      "rewards/margins": 14.5625,
      "rewards/rejected": -13.3125,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.0015197033576838867,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.296875,
      "logps/chosen": -110.0,
      "logps/rejected": -380.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2265625,
      "rewards/margins": 14.6875,
      "rewards/rejected": -13.5,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 0.0006111570124150037,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -1.484375,
      "logits/rejected": -1.28125,
      "logps/chosen": -119.5,
      "logps/rejected": -374.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.15625,
      "rewards/margins": 14.3125,
      "rewards/rejected": -13.1875,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 0.007744383515870189,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -1.4765625,
      "logits/rejected": -1.4296875,
      "logps/chosen": -112.0,
      "logps/rejected": -366.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 13.5625,
      "rewards/rejected": -12.4375,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.0012732364918080694,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -1.40625,
      "logits/rejected": -1.4375,
      "logps/chosen": -112.0,
      "logps/rejected": -378.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.265625,
      "rewards/margins": 14.4375,
      "rewards/rejected": -13.1875,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 0.0002836325231832676,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -1.375,
      "logits/rejected": -1.296875,
      "logps/chosen": -99.0,
      "logps/rejected": -388.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2265625,
      "rewards/margins": 15.6875,
      "rewards/rejected": -14.4375,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 0.0003737891758867377,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -1.3125,
      "logits/rejected": -1.3203125,
      "logps/chosen": -107.0,
      "logps/rejected": -384.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2734375,
      "rewards/margins": 15.5625,
      "rewards/rejected": -14.3125,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 8.578126063907685e-05,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -1.296875,
      "logits/rejected": -1.3515625,
      "logps/chosen": -109.0,
      "logps/rejected": -394.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0703125,
      "rewards/margins": 15.875,
      "rewards/rejected": -14.8125,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 0.00024371202591786317,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -1.40625,
      "logits/rejected": -1.359375,
      "logps/chosen": -111.0,
      "logps/rejected": -386.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.21875,
      "rewards/margins": 15.8125,
      "rewards/rejected": -14.5625,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 0.0011096809395769987,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.359375,
      "logps/chosen": -98.0,
      "logps/rejected": -386.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2421875,
      "rewards/margins": 15.625,
      "rewards/rejected": -14.4375,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.00012559595052717037,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -1.3125,
      "logits/rejected": -1.328125,
      "logps/chosen": -101.0,
      "logps/rejected": -398.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1640625,
      "rewards/margins": 16.5,
      "rewards/rejected": -15.3125,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 4.587907127048661e-05,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.265625,
      "logps/chosen": -106.5,
      "logps/rejected": -394.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1015625,
      "rewards/margins": 16.25,
      "rewards/rejected": -15.125,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 0.003436484517330256,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -1.40625,
      "logits/rejected": -1.359375,
      "logps/chosen": -125.0,
      "logps/rejected": -396.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0,
      "rewards/margins": 16.125,
      "rewards/rejected": -15.0625,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.00015189262808143838,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.3125,
      "logps/chosen": -119.0,
      "logps/rejected": -402.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.98046875,
      "rewards/margins": 16.5,
      "rewards/rejected": -15.5625,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 0.00023295877436104934,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -1.2578125,
      "logits/rejected": -1.3515625,
      "logps/chosen": -97.5,
      "logps/rejected": -398.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.171875,
      "rewards/margins": 16.625,
      "rewards/rejected": -15.4375,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 0.001046368209851116,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -1.359375,
      "logits/rejected": -1.3046875,
      "logps/chosen": -108.5,
      "logps/rejected": -406.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1328125,
      "rewards/margins": 16.625,
      "rewards/rejected": -15.5,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.00027227819240013357,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -1.421875,
      "logits/rejected": -1.3671875,
      "logps/chosen": -112.5,
      "logps/rejected": -404.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94921875,
      "rewards/margins": 16.75,
      "rewards/rejected": -15.8125,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 0.00016328713000194605,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.359375,
      "logps/chosen": -114.0,
      "logps/rejected": -410.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1015625,
      "rewards/margins": 17.25,
      "rewards/rejected": -16.125,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 0.00019838827176976435,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -1.4453125,
      "logits/rejected": -1.2890625,
      "logps/chosen": -107.5,
      "logps/rejected": -410.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.9765625,
      "rewards/margins": 17.125,
      "rewards/rejected": -16.125,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.0015154309315282544,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -1.28125,
      "logits/rejected": -1.359375,
      "logps/chosen": -98.5,
      "logps/rejected": -406.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 16.875,
      "rewards/rejected": -15.75,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 0.0002487440331129619,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -1.328125,
      "logits/rejected": -1.3203125,
      "logps/chosen": -101.0,
      "logps/rejected": -396.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1171875,
      "rewards/margins": 16.5,
      "rewards/rejected": -15.375,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 0.00015392331703359971,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -1.421875,
      "logits/rejected": -1.375,
      "logps/chosen": -112.0,
      "logps/rejected": -402.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.93359375,
      "rewards/margins": 17.0,
      "rewards/rejected": -16.125,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.0002166613584923672,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.3515625,
      "logps/chosen": -99.0,
      "logps/rejected": -404.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94921875,
      "rewards/margins": 17.0,
      "rewards/rejected": -16.125,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 0.00019977809817143256,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.2734375,
      "logps/chosen": -105.0,
      "logps/rejected": -406.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.03125,
      "rewards/margins": 17.0,
      "rewards/rejected": -16.0,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 7.015292642200733e-05,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -1.453125,
      "logits/rejected": -1.359375,
      "logps/chosen": -105.0,
      "logps/rejected": -412.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0390625,
      "rewards/margins": 17.375,
      "rewards/rejected": -16.375,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.2421875,
      "eval_logits/rejected": -1.28125,
      "eval_logps/chosen": -91.0,
      "eval_logps/rejected": -416.0,
      "eval_loss": 9.0727115775735e-07,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 1.0234375,
      "eval_rewards/margins": 17.75,
      "eval_rewards/rejected": -16.75,
      "eval_runtime": 11.1997,
      "eval_samples_per_second": 17.858,
      "eval_steps_per_second": 0.625,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.00010862728759544983,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -1.3046875,
      "logits/rejected": -1.3125,
      "logps/chosen": -112.0,
      "logps/rejected": -410.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0,
      "rewards/margins": 17.25,
      "rewards/rejected": -16.25,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 4.385121869847701e-05,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -1.3359375,
      "logits/rejected": -1.3515625,
      "logps/chosen": -106.5,
      "logps/rejected": -408.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 17.5,
      "rewards/rejected": -16.375,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 0.0006704320313793596,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -1.3828125,
      "logits/rejected": -1.2890625,
      "logps/chosen": -109.5,
      "logps/rejected": -412.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.984375,
      "rewards/margins": 17.625,
      "rewards/rejected": -16.75,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 3.926751354406349e-05,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -1.296875,
      "logits/rejected": -1.3984375,
      "logps/chosen": -104.5,
      "logps/rejected": -412.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.90625,
      "rewards/margins": 17.875,
      "rewards/rejected": -17.0,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 7.442917181039221e-05,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -1.375,
      "logits/rejected": -1.2890625,
      "logps/chosen": -104.0,
      "logps/rejected": -414.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.96875,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.25,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 0.0001506499834402947,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -1.2734375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -104.5,
      "logps/rejected": -420.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.96484375,
      "rewards/margins": 18.625,
      "rewards/rejected": -17.75,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 4.3054154781333365e-05,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -1.3984375,
      "logits/rejected": -1.40625,
      "logps/chosen": -103.5,
      "logps/rejected": -416.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1328125,
      "rewards/margins": 18.125,
      "rewards/rejected": -17.0,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 3.607556222389406e-05,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -1.40625,
      "logits/rejected": -1.328125,
      "logps/chosen": -112.5,
      "logps/rejected": -412.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0703125,
      "rewards/margins": 18.0,
      "rewards/rejected": -16.875,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 2.166580956783726e-05,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -1.390625,
      "logits/rejected": -1.34375,
      "logps/chosen": -110.0,
      "logps/rejected": -414.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0546875,
      "rewards/margins": 18.0,
      "rewards/rejected": -17.0,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.0007768824282097016,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -1.34375,
      "logits/rejected": -1.328125,
      "logps/chosen": -114.0,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.9921875,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.25,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 1.3261401519699353e-05,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -1.28125,
      "logits/rejected": -1.3203125,
      "logps/chosen": -103.0,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 18.5,
      "rewards/rejected": -17.375,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 0.0005719833094700048,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -1.40625,
      "logits/rejected": -1.3671875,
      "logps/chosen": -113.0,
      "logps/rejected": -420.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0234375,
      "rewards/margins": 18.625,
      "rewards/rejected": -17.5,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 4.7443275644552423e-05,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -1.40625,
      "logits/rejected": -1.3046875,
      "logps/chosen": -118.0,
      "logps/rejected": -416.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.796875,
      "rewards/margins": 18.375,
      "rewards/rejected": -17.625,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 1.663293459020266e-05,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -1.3828125,
      "logits/rejected": -1.2578125,
      "logps/chosen": -104.0,
      "logps/rejected": -428.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.96875,
      "rewards/margins": 19.0,
      "rewards/rejected": -18.0,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 0.00010957271577923361,
      "learning_rate": 1e-07,
      "logits/chosen": -1.265625,
      "logits/rejected": -1.2890625,
      "logps/chosen": -90.0,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1328125,
      "rewards/margins": 18.5,
      "rewards/rejected": -17.375,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.0003475483172869348,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -1.2734375,
      "logits/rejected": -1.296875,
      "logps/chosen": -107.0,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.93359375,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.25,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 3.990205582948748e-05,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -1.3828125,
      "logits/rejected": -1.34375,
      "logps/chosen": -102.5,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.109375,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.125,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 3.0246347907445157e-05,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.3046875,
      "logps/chosen": -120.0,
      "logps/rejected": -414.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94140625,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.375,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 1.7382836083050494e-05,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.328125,
      "logps/chosen": -102.5,
      "logps/rejected": -420.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0625,
      "rewards/margins": 18.5,
      "rewards/rejected": -17.5,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 1.1141783507026842e-05,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.359375,
      "logps/chosen": -111.0,
      "logps/rejected": -422.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0234375,
      "rewards/margins": 18.625,
      "rewards/rejected": -17.625,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 5.6649362804211735e-05,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -1.4375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -112.0,
      "logps/rejected": -428.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0078125,
      "rewards/margins": 19.25,
      "rewards/rejected": -18.25,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 4.1718570421510905e-05,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.375,
      "logps/chosen": -104.0,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.125,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 4.584929902297974e-05,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -1.3203125,
      "logits/rejected": -1.2890625,
      "logps/chosen": -111.5,
      "logps/rejected": -420.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.953125,
      "rewards/margins": 18.5,
      "rewards/rejected": -17.625,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 0.0008958556766046752,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -1.3671875,
      "logits/rejected": -1.3984375,
      "logps/chosen": -107.5,
      "logps/rejected": -414.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0625,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.25,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 5.93299351526953e-06,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -1.2734375,
      "logits/rejected": -1.3125,
      "logps/chosen": -117.0,
      "logps/rejected": -416.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0390625,
      "rewards/margins": 18.5,
      "rewards/rejected": -17.5,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 0.00014938750953581524,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -1.375,
      "logits/rejected": -1.359375,
      "logps/chosen": -96.0,
      "logps/rejected": -424.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.125,
      "rewards/margins": 18.875,
      "rewards/rejected": -17.75,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 0.00013613017137118136,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.2890625,
      "logps/chosen": -118.0,
      "logps/rejected": -418.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.9609375,
      "rewards/margins": 18.375,
      "rewards/rejected": -17.375,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.0003341949440027235,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -1.421875,
      "logits/rejected": -1.3515625,
      "logps/chosen": -110.0,
      "logps/rejected": -426.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.09375,
      "rewards/margins": 18.75,
      "rewards/rejected": -17.625,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 3.7993345789288195e-05,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -1.359375,
      "logits/rejected": -1.3515625,
      "logps/chosen": -122.0,
      "logps/rejected": -424.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0,
      "rewards/margins": 18.625,
      "rewards/rejected": -17.625,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 5.6917102751922324e-05,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -1.2890625,
      "logits/rejected": -1.34375,
      "logps/chosen": -107.0,
      "logps/rejected": -430.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.88671875,
      "rewards/margins": 19.0,
      "rewards/rejected": -18.0,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 0.00018882592412388543,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -1.4375,
      "logits/rejected": -1.3125,
      "logps/chosen": -108.0,
      "logps/rejected": -416.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.96484375,
      "rewards/margins": 18.0,
      "rewards/rejected": -17.125,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.2421875,
      "eval_logits/rejected": -1.2734375,
      "eval_logps/chosen": -91.5,
      "eval_logps/rejected": -432.0,
      "eval_loss": 2.9754883712485025e-07,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.99609375,
      "eval_rewards/margins": 19.125,
      "eval_rewards/rejected": -18.125,
      "eval_runtime": 13.8353,
      "eval_samples_per_second": 14.456,
      "eval_steps_per_second": 0.506,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.022397415345911637,
      "train_runtime": 4644.5622,
      "train_samples_per_second": 6.459,
      "train_steps_per_second": 0.202
    }
  ],
  "logging_steps": 10,
  "max_steps": 939,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}