{
  "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": 54.91204506748968,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -3.609375,
      "logits/rejected": -3.8125,
      "logps/chosen": -124.5,
      "logps/rejected": -52.5,
      "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": 83.69103566326362,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.453125,
      "logps/chosen": -96.0,
      "logps/rejected": -71.0,
      "loss": 0.6912,
      "rewards/accuracies": 0.1944444477558136,
      "rewards/chosen": -0.0034637451171875,
      "rewards/margins": 0.0013885498046875,
      "rewards/rejected": -0.004852294921875,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 71.48080177755007,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.390625,
      "logps/chosen": -80.0,
      "logps/rejected": -92.0,
      "loss": 0.689,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -0.0177001953125,
      "rewards/margins": 0.017333984375,
      "rewards/rejected": -0.03515625,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 53.55900633266503,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.40625,
      "logps/chosen": -80.0,
      "logps/rejected": -89.5,
      "loss": 0.6722,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.042724609375,
      "rewards/margins": 0.044189453125,
      "rewards/rejected": -0.0869140625,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 66.9145932830632,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.515625,
      "logps/chosen": -82.5,
      "logps/rejected": -84.0,
      "loss": 0.659,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.08984375,
      "rewards/margins": 0.06201171875,
      "rewards/rejected": -0.15234375,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 43.670730033798854,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -3.625,
      "logits/rejected": -3.59375,
      "logps/chosen": -76.5,
      "logps/rejected": -95.5,
      "loss": 0.6322,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.1279296875,
      "rewards/margins": 0.15625,
      "rewards/rejected": -0.28515625,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 46.741290964112856,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.296875,
      "logps/chosen": -80.0,
      "logps/rejected": -91.0,
      "loss": 0.6724,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.06787109375,
      "rewards/margins": 0.3125,
      "rewards/rejected": -0.37890625,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 46.37071905241544,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -3.484375,
      "logits/rejected": -3.40625,
      "logps/chosen": -76.5,
      "logps/rejected": -90.0,
      "loss": 0.6188,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.0198974609375,
      "rewards/margins": 0.30859375,
      "rewards/rejected": -0.2890625,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 40.078593428645554,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -3.515625,
      "logits/rejected": -3.46875,
      "logps/chosen": -73.0,
      "logps/rejected": -103.5,
      "loss": 0.6507,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.0576171875,
      "rewards/margins": 0.78515625,
      "rewards/rejected": -0.84375,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 48.907912065438,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -3.625,
      "logits/rejected": -3.5625,
      "logps/chosen": -82.5,
      "logps/rejected": -99.0,
      "loss": 0.5341,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.058837890625,
      "rewards/margins": 0.65234375,
      "rewards/rejected": -0.59375,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 57.58557004423775,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -3.609375,
      "logits/rejected": -3.484375,
      "logps/chosen": -87.5,
      "logps/rejected": -95.0,
      "loss": 0.5099,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.1396484375,
      "rewards/margins": 0.78515625,
      "rewards/rejected": -0.64453125,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 42.378161307592016,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.5,
      "logps/chosen": -70.0,
      "logps/rejected": -108.0,
      "loss": 0.5245,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.08203125,
      "rewards/margins": 1.015625,
      "rewards/rejected": -0.9296875,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 50.38480454080293,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -3.484375,
      "logits/rejected": -3.5,
      "logps/chosen": -78.5,
      "logps/rejected": -103.5,
      "loss": 0.6502,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.1748046875,
      "rewards/margins": 0.9296875,
      "rewards/rejected": -1.1015625,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 44.620675269517484,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -3.5,
      "logits/rejected": -3.546875,
      "logps/chosen": -78.0,
      "logps/rejected": -103.5,
      "loss": 0.5064,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.056396484375,
      "rewards/margins": 1.015625,
      "rewards/rejected": -1.0703125,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 38.18139489697703,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.53125,
      "logps/chosen": -78.5,
      "logps/rejected": -97.5,
      "loss": 0.5037,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.0908203125,
      "rewards/margins": 0.96875,
      "rewards/rejected": -0.875,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 80.12995502682458,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.5,
      "logps/chosen": -81.5,
      "logps/rejected": -91.5,
      "loss": 0.4751,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.26953125,
      "rewards/margins": 1.21875,
      "rewards/rejected": -0.94921875,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 43.55724368129485,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.453125,
      "logps/chosen": -73.0,
      "logps/rejected": -104.5,
      "loss": 0.4086,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.1025390625,
      "rewards/margins": 1.21875,
      "rewards/rejected": -1.1171875,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 57.948423255330844,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -3.65625,
      "logits/rejected": -3.5625,
      "logps/chosen": -77.0,
      "logps/rejected": -103.0,
      "loss": 0.421,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.171875,
      "rewards/margins": 1.609375,
      "rewards/rejected": -1.4375,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 57.14677545405883,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.40625,
      "logps/chosen": -72.0,
      "logps/rejected": -106.0,
      "loss": 0.4694,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.171875,
      "rewards/margins": 1.5546875,
      "rewards/rejected": -1.3828125,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 36.209893051679046,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.5,
      "logps/chosen": -81.0,
      "logps/rejected": -99.0,
      "loss": 0.465,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.045166015625,
      "rewards/margins": 1.25,
      "rewards/rejected": -1.2109375,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 38.81781204643516,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -3.609375,
      "logits/rejected": -3.453125,
      "logps/chosen": -82.0,
      "logps/rejected": -95.0,
      "loss": 0.368,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.193359375,
      "rewards/margins": 1.296875,
      "rewards/rejected": -1.1015625,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 42.14684997699955,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.4375,
      "logps/chosen": -74.0,
      "logps/rejected": -106.0,
      "loss": 0.3861,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.28515625,
      "rewards/margins": 2.0,
      "rewards/rejected": -1.71875,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 78.95472000856175,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.5,
      "logps/chosen": -78.0,
      "logps/rejected": -102.0,
      "loss": 0.4323,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.00067138671875,
      "rewards/margins": 1.453125,
      "rewards/rejected": -1.453125,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 39.48323175963722,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -3.4375,
      "logits/rejected": -3.390625,
      "logps/chosen": -82.0,
      "logps/rejected": -103.5,
      "loss": 0.4079,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.2099609375,
      "rewards/margins": 1.9140625,
      "rewards/rejected": -1.703125,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 39.15005497208971,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.4375,
      "logps/chosen": -78.0,
      "logps/rejected": -115.0,
      "loss": 0.4427,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.10107421875,
      "rewards/margins": 1.765625,
      "rewards/rejected": -1.6640625,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 34.86780232752544,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.421875,
      "logps/chosen": -74.5,
      "logps/rejected": -113.5,
      "loss": 0.3508,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.365234375,
      "rewards/margins": 2.09375,
      "rewards/rejected": -1.734375,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 33.65428910732692,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -3.609375,
      "logits/rejected": -3.484375,
      "logps/chosen": -71.0,
      "logps/rejected": -121.0,
      "loss": 0.3774,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.279296875,
      "rewards/margins": 2.96875,
      "rewards/rejected": -2.6875,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 24.522618840121858,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.484375,
      "logps/chosen": -82.5,
      "logps/rejected": -98.5,
      "loss": 0.3277,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.193359375,
      "rewards/margins": 1.7109375,
      "rewards/rejected": -1.515625,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 37.957743392782504,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.5,
      "logps/chosen": -77.0,
      "logps/rejected": -105.0,
      "loss": 0.4696,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.328125,
      "rewards/margins": 1.96875,
      "rewards/rejected": -1.640625,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 40.919686372535324,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.421875,
      "logps/chosen": -75.5,
      "logps/rejected": -99.0,
      "loss": 0.3743,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.1201171875,
      "rewards/margins": 1.9296875,
      "rewards/rejected": -1.8125,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 56.86349858521586,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -3.625,
      "logits/rejected": -3.640625,
      "logps/chosen": -79.5,
      "logps/rejected": -125.0,
      "loss": 0.3358,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.052734375,
      "rewards/margins": 1.9140625,
      "rewards/rejected": -1.859375,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 42.240708479135115,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -3.6875,
      "logits/rejected": -3.421875,
      "logps/chosen": -90.5,
      "logps/rejected": -95.0,
      "loss": 0.4604,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.353515625,
      "rewards/margins": 1.9609375,
      "rewards/rejected": -1.609375,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -3.546875,
      "eval_logits/rejected": -3.515625,
      "eval_logps/chosen": -74.5,
      "eval_logps/rejected": -104.5,
      "eval_loss": 0.5415624976158142,
      "eval_rewards/accuracies": 0.6785714030265808,
      "eval_rewards/chosen": -0.279296875,
      "eval_rewards/margins": 0.97265625,
      "eval_rewards/rejected": -1.25,
      "eval_runtime": 12.4672,
      "eval_samples_per_second": 16.042,
      "eval_steps_per_second": 0.561,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 9.346111921886239,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -3.4375,
      "logits/rejected": -3.328125,
      "logps/chosen": -76.0,
      "logps/rejected": -115.5,
      "loss": 0.2224,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.1806640625,
      "rewards/margins": 2.71875,
      "rewards/rejected": -2.53125,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 22.708486677603965,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -3.5,
      "logits/rejected": -3.4375,
      "logps/chosen": -71.0,
      "logps/rejected": -114.5,
      "loss": 0.173,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.6171875,
      "rewards/margins": 3.078125,
      "rewards/rejected": -2.453125,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 25.594601503918632,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -3.625,
      "logits/rejected": -3.328125,
      "logps/chosen": -73.5,
      "logps/rejected": -109.5,
      "loss": 0.1191,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.82421875,
      "rewards/margins": 3.421875,
      "rewards/rejected": -2.59375,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 25.095730138726623,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.390625,
      "logps/chosen": -78.5,
      "logps/rejected": -114.0,
      "loss": 0.156,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.7265625,
      "rewards/margins": 3.609375,
      "rewards/rejected": -2.890625,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 57.17457043068879,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -3.53125,
      "logits/rejected": -3.203125,
      "logps/chosen": -79.5,
      "logps/rejected": -109.5,
      "loss": 0.157,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.69140625,
      "rewards/margins": 3.5625,
      "rewards/rejected": -2.859375,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 12.603394535533164,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -3.375,
      "logits/rejected": -3.453125,
      "logps/chosen": -73.0,
      "logps/rejected": -114.5,
      "loss": 0.146,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.71875,
      "rewards/margins": 3.421875,
      "rewards/rejected": -2.703125,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 31.796547991836366,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -3.546875,
      "logits/rejected": -3.453125,
      "logps/chosen": -73.0,
      "logps/rejected": -135.0,
      "loss": 0.147,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.546875,
      "rewards/margins": 4.40625,
      "rewards/rejected": -3.84375,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 17.359416072010845,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.328125,
      "logps/chosen": -76.5,
      "logps/rejected": -114.0,
      "loss": 0.1611,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.515625,
      "rewards/margins": 3.484375,
      "rewards/rejected": -2.953125,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 21.91626573741424,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -3.53125,
      "logits/rejected": -3.359375,
      "logps/chosen": -78.0,
      "logps/rejected": -125.5,
      "loss": 0.1203,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.4453125,
      "rewards/margins": 4.0625,
      "rewards/rejected": -3.625,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 22.151247767510483,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -3.546875,
      "logits/rejected": -3.5,
      "logps/chosen": -78.5,
      "logps/rejected": -113.5,
      "loss": 0.1449,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.6796875,
      "rewards/margins": 3.765625,
      "rewards/rejected": -3.09375,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 12.839758011921711,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -3.515625,
      "logits/rejected": -3.390625,
      "logps/chosen": -76.0,
      "logps/rejected": -118.5,
      "loss": 0.1471,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.3203125,
      "rewards/margins": 3.8125,
      "rewards/rejected": -3.5,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 14.796849771267464,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -3.5,
      "logits/rejected": -3.453125,
      "logps/chosen": -75.0,
      "logps/rejected": -123.0,
      "loss": 0.1245,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6640625,
      "rewards/margins": 4.25,
      "rewards/rejected": -3.5625,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 50.3602171858621,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -3.640625,
      "logits/rejected": -3.453125,
      "logps/chosen": -86.0,
      "logps/rejected": -117.5,
      "loss": 0.1199,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.96484375,
      "rewards/margins": 4.6875,
      "rewards/rejected": -3.71875,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 24.93374901733837,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.390625,
      "logps/chosen": -75.0,
      "logps/rejected": -146.0,
      "loss": 0.1477,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.011962890625,
      "rewards/margins": 4.75,
      "rewards/rejected": -4.71875,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 25.852996383096112,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.46875,
      "logps/chosen": -73.0,
      "logps/rejected": -146.0,
      "loss": 0.133,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.25390625,
      "rewards/margins": 4.9375,
      "rewards/rejected": -4.65625,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 49.65712682603586,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -3.578125,
      "logits/rejected": -3.359375,
      "logps/chosen": -73.0,
      "logps/rejected": -147.0,
      "loss": 0.1198,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.0167236328125,
      "rewards/margins": 4.71875,
      "rewards/rejected": -4.71875,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 36.81147017790551,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -3.421875,
      "logits/rejected": -3.390625,
      "logps/chosen": -69.5,
      "logps/rejected": -142.0,
      "loss": 0.0979,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.34765625,
      "rewards/margins": 5.09375,
      "rewards/rejected": -4.75,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 21.233576498613207,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -3.609375,
      "logits/rejected": -3.46875,
      "logps/chosen": -86.5,
      "logps/rejected": -123.0,
      "loss": 0.1249,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.515625,
      "rewards/margins": 4.3125,
      "rewards/rejected": -3.796875,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 16.340242850202618,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.28125,
      "logps/chosen": -75.0,
      "logps/rejected": -135.0,
      "loss": 0.093,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.5078125,
      "rewards/margins": 5.5,
      "rewards/rejected": -5.0,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 14.733756298918374,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -3.390625,
      "logits/rejected": -3.40625,
      "logps/chosen": -70.0,
      "logps/rejected": -149.0,
      "loss": 0.1119,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.03564453125,
      "rewards/margins": 4.9375,
      "rewards/rejected": -5.0,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 23.33060659321782,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -3.484375,
      "logits/rejected": -3.453125,
      "logps/chosen": -75.5,
      "logps/rejected": -129.0,
      "loss": 0.1585,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.125,
      "rewards/margins": 4.28125,
      "rewards/rejected": -4.15625,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 22.427410271860705,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -3.34375,
      "logits/rejected": -3.25,
      "logps/chosen": -69.0,
      "logps/rejected": -148.0,
      "loss": 0.0962,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.040771484375,
      "rewards/margins": 5.125,
      "rewards/rejected": -5.15625,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 15.793382293172824,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -3.40625,
      "logits/rejected": -3.234375,
      "logps/chosen": -78.5,
      "logps/rejected": -133.0,
      "loss": 0.0915,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.173828125,
      "rewards/margins": 5.0,
      "rewards/rejected": -4.8125,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 45.93826349303006,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -3.484375,
      "logits/rejected": -3.28125,
      "logps/chosen": -83.0,
      "logps/rejected": -132.0,
      "loss": 0.1383,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.0242919921875,
      "rewards/margins": 5.09375,
      "rewards/rejected": -5.09375,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 9.631296619786061,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.171875,
      "logps/chosen": -76.5,
      "logps/rejected": -145.0,
      "loss": 0.0963,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.220703125,
      "rewards/margins": 5.84375,
      "rewards/rejected": -5.625,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 36.84885036469591,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -3.609375,
      "logits/rejected": -3.328125,
      "logps/chosen": -80.0,
      "logps/rejected": -130.0,
      "loss": 0.1019,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.5234375,
      "rewards/margins": 4.875,
      "rewards/rejected": -4.34375,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 58.51630805316546,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -3.359375,
      "logits/rejected": -3.359375,
      "logps/chosen": -73.5,
      "logps/rejected": -137.0,
      "loss": 0.1676,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.1787109375,
      "rewards/margins": 4.40625,
      "rewards/rejected": -4.21875,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 27.20728584052076,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.421875,
      "logps/chosen": -79.0,
      "logps/rejected": -134.0,
      "loss": 0.1541,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.416015625,
      "rewards/margins": 5.0625,
      "rewards/rejected": -4.65625,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 7.349613633649374,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -3.671875,
      "logits/rejected": -3.34375,
      "logps/chosen": -84.0,
      "logps/rejected": -128.0,
      "loss": 0.1162,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.49609375,
      "rewards/margins": 4.75,
      "rewards/rejected": -4.25,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 25.280141244938278,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -3.296875,
      "logits/rejected": -3.390625,
      "logps/chosen": -73.5,
      "logps/rejected": -152.0,
      "loss": 0.1481,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.12451171875,
      "rewards/margins": 5.125,
      "rewards/rejected": -5.25,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 26.035177686239535,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -3.4375,
      "logits/rejected": -3.359375,
      "logps/chosen": -83.0,
      "logps/rejected": -132.0,
      "loss": 0.1593,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.05615234375,
      "rewards/margins": 4.6875,
      "rewards/rejected": -4.75,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -3.484375,
      "eval_logits/rejected": -3.4375,
      "eval_logps/chosen": -85.5,
      "eval_logps/rejected": -121.0,
      "eval_loss": 0.6119921803474426,
      "eval_rewards/accuracies": 0.6964285969734192,
      "eval_rewards/chosen": -1.3671875,
      "eval_rewards/margins": 1.53125,
      "eval_rewards/rejected": -2.890625,
      "eval_runtime": 12.5181,
      "eval_samples_per_second": 15.977,
      "eval_steps_per_second": 0.559,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 9.553959092849897,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.421875,
      "logps/chosen": -74.0,
      "logps/rejected": -147.0,
      "loss": 0.1076,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.0986328125,
      "rewards/margins": 4.9375,
      "rewards/rejected": -5.03125,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 8.457516826686375,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -3.40625,
      "logits/rejected": -3.46875,
      "logps/chosen": -77.0,
      "logps/rejected": -145.0,
      "loss": 0.0325,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.23046875,
      "rewards/margins": 5.5,
      "rewards/rejected": -5.28125,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 55.86329155487255,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -3.484375,
      "logits/rejected": -3.28125,
      "logps/chosen": -78.0,
      "logps/rejected": -154.0,
      "loss": 0.0294,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.31640625,
      "rewards/margins": 6.78125,
      "rewards/rejected": -6.46875,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 7.718735109632817,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -3.515625,
      "logits/rejected": -3.171875,
      "logps/chosen": -81.5,
      "logps/rejected": -155.0,
      "loss": 0.0293,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.28125,
      "rewards/margins": 7.21875,
      "rewards/rejected": -6.9375,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 8.509932612126983,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -3.53125,
      "logits/rejected": -3.265625,
      "logps/chosen": -77.5,
      "logps/rejected": -167.0,
      "loss": 0.0275,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.1279296875,
      "rewards/margins": 7.5625,
      "rewards/rejected": -7.6875,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 5.182963763871152,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -3.4375,
      "logits/rejected": -3.234375,
      "logps/chosen": -85.5,
      "logps/rejected": -162.0,
      "loss": 0.0422,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.0361328125,
      "rewards/margins": 7.46875,
      "rewards/rejected": -7.5,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 6.839940107975035,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -3.40625,
      "logits/rejected": -3.28125,
      "logps/chosen": -92.5,
      "logps/rejected": -186.0,
      "loss": 0.0273,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.85546875,
      "rewards/margins": 8.375,
      "rewards/rejected": -9.1875,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 40.31994684060318,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -3.4375,
      "logits/rejected": -3.3125,
      "logps/chosen": -81.5,
      "logps/rejected": -173.0,
      "loss": 0.0421,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.455078125,
      "rewards/margins": 8.5,
      "rewards/rejected": -8.9375,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 10.582151776223212,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -3.296875,
      "logits/rejected": -3.40625,
      "logps/chosen": -91.5,
      "logps/rejected": -198.0,
      "loss": 0.0331,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.6640625,
      "rewards/margins": 8.5625,
      "rewards/rejected": -10.1875,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 8.925593167475649,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -3.5,
      "logits/rejected": -3.5,
      "logps/chosen": -93.5,
      "logps/rejected": -176.0,
      "loss": 0.0499,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.91015625,
      "rewards/margins": 7.71875,
      "rewards/rejected": -8.625,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 7.581895717465126,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -3.5625,
      "logits/rejected": -3.390625,
      "logps/chosen": -94.5,
      "logps/rejected": -173.0,
      "loss": 0.0344,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.625,
      "rewards/margins": 8.0,
      "rewards/rejected": -8.625,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 1.850859647858044,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -3.28125,
      "logits/rejected": -3.3125,
      "logps/chosen": -84.0,
      "logps/rejected": -207.0,
      "loss": 0.0347,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.8125,
      "rewards/margins": 9.25,
      "rewards/rejected": -11.0625,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 5.116015057181206,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -3.375,
      "logits/rejected": -3.296875,
      "logps/chosen": -84.0,
      "logps/rejected": -206.0,
      "loss": 0.0276,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.0390625,
      "rewards/margins": 10.625,
      "rewards/rejected": -11.625,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 5.627455486007399,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -3.296875,
      "logits/rejected": -3.296875,
      "logps/chosen": -90.5,
      "logps/rejected": -174.0,
      "loss": 0.0372,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.58203125,
      "rewards/margins": 8.375,
      "rewards/rejected": -8.9375,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 1.9515967492229094,
      "learning_rate": 1e-07,
      "logits/chosen": -3.21875,
      "logits/rejected": -3.484375,
      "logps/chosen": -93.0,
      "logps/rejected": -182.0,
      "loss": 0.0304,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.0546875,
      "rewards/margins": 8.75,
      "rewards/rejected": -9.8125,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 37.33912179478409,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.421875,
      "logps/chosen": -88.0,
      "logps/rejected": -194.0,
      "loss": 0.0258,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.1015625,
      "rewards/margins": 8.625,
      "rewards/rejected": -9.6875,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 2.011094757599656,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -3.34375,
      "logits/rejected": -3.234375,
      "logps/chosen": -84.5,
      "logps/rejected": -192.0,
      "loss": 0.0232,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1796875,
      "rewards/margins": 8.75,
      "rewards/rejected": -9.9375,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 5.687933952633793,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -3.390625,
      "logits/rejected": -3.328125,
      "logps/chosen": -88.0,
      "logps/rejected": -178.0,
      "loss": 0.0292,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8359375,
      "rewards/margins": 7.9375,
      "rewards/rejected": -8.75,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 12.729895402222695,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.21875,
      "logps/chosen": -88.0,
      "logps/rejected": -186.0,
      "loss": 0.0502,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -1.0703125,
      "rewards/margins": 8.375,
      "rewards/rejected": -9.4375,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 8.744192592460582,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.265625,
      "logps/chosen": -89.5,
      "logps/rejected": -183.0,
      "loss": 0.0109,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.482421875,
      "rewards/margins": 8.9375,
      "rewards/rejected": -9.375,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 5.875109775991397,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.296875,
      "logps/chosen": -92.5,
      "logps/rejected": -210.0,
      "loss": 0.0231,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -1.453125,
      "rewards/margins": 10.0,
      "rewards/rejected": -11.4375,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 5.917586908189079,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.15625,
      "logps/chosen": -92.5,
      "logps/rejected": -198.0,
      "loss": 0.0262,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.71875,
      "rewards/margins": 8.8125,
      "rewards/rejected": -10.5,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 65.62542835716923,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -3.265625,
      "logits/rejected": -3.1875,
      "logps/chosen": -87.5,
      "logps/rejected": -205.0,
      "loss": 0.0222,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.8046875,
      "rewards/margins": 8.8125,
      "rewards/rejected": -10.625,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 7.503774618994616,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -3.1875,
      "logits/rejected": -3.4375,
      "logps/chosen": -96.5,
      "logps/rejected": -178.0,
      "loss": 0.0519,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -1.4609375,
      "rewards/margins": 7.9375,
      "rewards/rejected": -9.375,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 116.40230177947107,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -3.453125,
      "logits/rejected": -3.21875,
      "logps/chosen": -100.5,
      "logps/rejected": -172.0,
      "loss": 0.0278,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.65234375,
      "rewards/margins": 9.0,
      "rewards/rejected": -9.625,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 15.038161889193097,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.4375,
      "logps/chosen": -91.0,
      "logps/rejected": -199.0,
      "loss": 0.0442,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.6015625,
      "rewards/margins": 9.25,
      "rewards/rejected": -10.8125,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 15.841560904662904,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -3.265625,
      "logits/rejected": -3.328125,
      "logps/chosen": -94.0,
      "logps/rejected": -205.0,
      "loss": 0.0237,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.65625,
      "rewards/margins": 9.1875,
      "rewards/rejected": -10.8125,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 14.97273512190642,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -3.296875,
      "logits/rejected": -3.34375,
      "logps/chosen": -92.0,
      "logps/rejected": -212.0,
      "loss": 0.0231,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.6484375,
      "rewards/margins": 9.9375,
      "rewards/rejected": -11.625,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 9.521298988731209,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -3.34375,
      "logits/rejected": -3.390625,
      "logps/chosen": -96.0,
      "logps/rejected": -193.0,
      "loss": 0.0208,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.2265625,
      "rewards/margins": 9.125,
      "rewards/rejected": -10.3125,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 1.099193484912796,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -3.328125,
      "logits/rejected": -3.328125,
      "logps/chosen": -93.0,
      "logps/rejected": -200.0,
      "loss": 0.0315,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -1.671875,
      "rewards/margins": 9.25,
      "rewards/rejected": -10.9375,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 47.638493930140854,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.25,
      "logps/chosen": -89.0,
      "logps/rejected": -206.0,
      "loss": 0.052,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.6875,
      "rewards/margins": 10.1875,
      "rewards/rejected": -11.875,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -3.34375,
      "eval_logits/rejected": -3.328125,
      "eval_logps/chosen": -109.5,
      "eval_logps/rejected": -155.0,
      "eval_loss": 0.8959326148033142,
      "eval_rewards/accuracies": 0.6964285969734192,
      "eval_rewards/chosen": -3.78125,
      "eval_rewards/margins": 2.546875,
      "eval_rewards/rejected": -6.34375,
      "eval_runtime": 15.6058,
      "eval_samples_per_second": 12.816,
      "eval_steps_per_second": 0.449,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.2217866757426399,
      "train_runtime": 4852.7852,
      "train_samples_per_second": 6.182,
      "train_steps_per_second": 0.193
    }
  ],
  "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
}