{
  "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": 61.49272285065342,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.4296875,
      "logps/chosen": -135.0,
      "logps/rejected": -82.0,
      "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": 64.59200921165986,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.390625,
      "logps/chosen": -138.0,
      "logps/rejected": -93.0,
      "loss": 0.6867,
      "rewards/accuracies": 0.4027777910232544,
      "rewards/chosen": 0.003143310546875,
      "rewards/margins": 0.01043701171875,
      "rewards/rejected": -0.00732421875,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 50.07302232099878,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -1.4375,
      "logits/rejected": -1.375,
      "logps/chosen": -123.0,
      "logps/rejected": -95.5,
      "loss": 0.5975,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.1357421875,
      "rewards/margins": 0.2236328125,
      "rewards/rejected": -0.08837890625,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 26.095443217070695,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -1.4375,
      "logits/rejected": -1.3671875,
      "logps/chosen": -98.5,
      "logps/rejected": -95.5,
      "loss": 0.3751,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.484375,
      "rewards/margins": 0.875,
      "rewards/rejected": -0.390625,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 12.410203511154618,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -1.4140625,
      "logits/rejected": -1.3359375,
      "logps/chosen": -113.5,
      "logps/rejected": -109.5,
      "loss": 0.1565,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.734375,
      "rewards/margins": 2.28125,
      "rewards/rejected": -1.546875,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 3.1102180128507553,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -1.4296875,
      "logits/rejected": -1.3515625,
      "logps/chosen": -113.5,
      "logps/rejected": -136.0,
      "loss": 0.0414,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94140625,
      "rewards/margins": 4.71875,
      "rewards/rejected": -3.78125,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 1.6585282597010376,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -1.375,
      "logits/rejected": -1.28125,
      "logps/chosen": -115.5,
      "logps/rejected": -156.0,
      "loss": 0.025,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.9609375,
      "rewards/margins": 6.75,
      "rewards/rejected": -5.78125,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 1.6309695015065353,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -1.3515625,
      "logits/rejected": -1.265625,
      "logps/chosen": -121.5,
      "logps/rejected": -160.0,
      "loss": 0.019,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.28125,
      "rewards/margins": 8.125,
      "rewards/rejected": -6.84375,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 0.5423823083280593,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -1.28125,
      "logits/rejected": -1.2265625,
      "logps/chosen": -105.5,
      "logps/rejected": -184.0,
      "loss": 0.0037,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.21875,
      "rewards/margins": 10.0625,
      "rewards/rejected": -8.875,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 0.1564903117582722,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.2109375,
      "logps/chosen": -115.0,
      "logps/rejected": -226.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.25,
      "rewards/margins": 13.9375,
      "rewards/rejected": -12.6875,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 0.08472430011201164,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -1.2734375,
      "logits/rejected": -1.2265625,
      "logps/chosen": -102.5,
      "logps/rejected": -227.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.40625,
      "rewards/margins": 14.0625,
      "rewards/rejected": -12.6875,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 0.057237148188972964,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -1.3203125,
      "logits/rejected": -1.2265625,
      "logps/chosen": -107.5,
      "logps/rejected": -222.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5546875,
      "rewards/margins": 14.5,
      "rewards/rejected": -12.9375,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.00883849570169924,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -1.265625,
      "logits/rejected": -1.15625,
      "logps/chosen": -116.5,
      "logps/rejected": -223.0,
      "loss": 0.0009,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.140625,
      "rewards/margins": 14.25,
      "rewards/rejected": -13.0625,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 0.050923577333219915,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -1.265625,
      "logits/rejected": -1.1953125,
      "logps/chosen": -112.0,
      "logps/rejected": -229.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.5703125,
      "rewards/margins": 14.25,
      "rewards/rejected": -13.625,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 0.0027953076972493883,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -1.265625,
      "logits/rejected": -1.15625,
      "logps/chosen": -120.0,
      "logps/rejected": -229.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2109375,
      "rewards/margins": 14.9375,
      "rewards/rejected": -13.6875,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 0.09003195137744328,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.1796875,
      "logps/chosen": -127.5,
      "logps/rejected": -231.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2421875,
      "rewards/margins": 15.25,
      "rewards/rejected": -14.0,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 0.017992401732151855,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.171875,
      "logps/chosen": -114.5,
      "logps/rejected": -242.0,
      "loss": 0.0026,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.53515625,
      "rewards/margins": 15.0625,
      "rewards/rejected": -14.5,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 0.16506074037787472,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -1.2578125,
      "logits/rejected": -1.203125,
      "logps/chosen": -116.0,
      "logps/rejected": -240.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.494140625,
      "rewards/margins": 15.25,
      "rewards/rejected": -14.75,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.04434530409045823,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -1.2578125,
      "logits/rejected": -1.1875,
      "logps/chosen": -124.0,
      "logps/rejected": -258.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7265625,
      "rewards/margins": 17.0,
      "rewards/rejected": -16.25,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 0.008248687934695326,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -1.171875,
      "logits/rejected": -1.15625,
      "logps/chosen": -105.0,
      "logps/rejected": -255.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.8515625,
      "rewards/margins": 17.25,
      "rewards/rejected": -16.375,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 0.004339237156321609,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -1.1640625,
      "logits/rejected": -1.1171875,
      "logps/chosen": -81.5,
      "logps/rejected": -266.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.8828125,
      "rewards/margins": 18.875,
      "rewards/rejected": -17.0,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.2633647746468779,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.140625,
      "logps/chosen": -113.0,
      "logps/rejected": -268.0,
      "loss": 0.0004,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.828125,
      "rewards/margins": 18.625,
      "rewards/rejected": -16.75,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 0.0030891530649115573,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.1875,
      "logps/chosen": -99.0,
      "logps/rejected": -250.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.65625,
      "rewards/margins": 17.125,
      "rewards/rejected": -15.5,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 0.4831724706311816,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.140625,
      "logps/chosen": -95.0,
      "logps/rejected": -270.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.625,
      "rewards/margins": 19.0,
      "rewards/rejected": -17.5,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.00022049042774774556,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.1796875,
      "logps/chosen": -96.5,
      "logps/rejected": -246.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6953125,
      "rewards/margins": 17.0,
      "rewards/rejected": -15.3125,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 5.963574704969866,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -1.2578125,
      "logits/rejected": -1.15625,
      "logps/chosen": -128.0,
      "logps/rejected": -262.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.265625,
      "rewards/margins": 17.625,
      "rewards/rejected": -16.375,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 0.006798078235934642,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -1.2421875,
      "logits/rejected": -1.1640625,
      "logps/chosen": -111.5,
      "logps/rejected": -248.0,
      "loss": 0.0008,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.3125,
      "rewards/margins": 16.625,
      "rewards/rejected": -15.25,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.012926408245738876,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.1328125,
      "logps/chosen": -108.0,
      "logps/rejected": -258.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.9296875,
      "rewards/margins": 17.375,
      "rewards/rejected": -16.375,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 0.012047505266969924,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.1796875,
      "logps/chosen": -90.0,
      "logps/rejected": -260.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.203125,
      "rewards/margins": 17.75,
      "rewards/rejected": -16.5,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 0.0038855702170066918,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.1484375,
      "logps/chosen": -117.5,
      "logps/rejected": -258.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2109375,
      "rewards/margins": 18.0,
      "rewards/rejected": -16.875,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.11915291348972191,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.1640625,
      "logps/chosen": -98.5,
      "logps/rejected": -270.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.609375,
      "rewards/margins": 19.25,
      "rewards/rejected": -17.625,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 9.60700625296979e-05,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -1.1796875,
      "logits/rejected": -1.109375,
      "logps/chosen": -126.5,
      "logps/rejected": -266.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.34375,
      "rewards/margins": 18.625,
      "rewards/rejected": -17.25,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.171875,
      "eval_logits/rejected": -1.140625,
      "eval_logps/chosen": -141.0,
      "eval_logps/rejected": -264.0,
      "eval_loss": 1.0792873581522144e-05,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 1.4375,
      "eval_rewards/margins": 18.25,
      "eval_rewards/rejected": -16.875,
      "eval_runtime": 14.9791,
      "eval_samples_per_second": 13.352,
      "eval_steps_per_second": 0.467,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 0.0021804409507147635,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.1484375,
      "logps/chosen": -108.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.484375,
      "rewards/margins": 19.125,
      "rewards/rejected": -17.625,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.0002507624481599604,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.140625,
      "logps/chosen": -110.5,
      "logps/rejected": -250.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.359375,
      "rewards/margins": 17.75,
      "rewards/rejected": -16.375,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 0.00013068772996752224,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.1328125,
      "logps/chosen": -77.0,
      "logps/rejected": -274.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.65625,
      "rewards/margins": 19.625,
      "rewards/rejected": -18.0,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 0.0031721421859132516,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.1328125,
      "logps/chosen": -115.0,
      "logps/rejected": -266.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5,
      "rewards/margins": 19.25,
      "rewards/rejected": -17.75,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.0002774022241354511,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.15625,
      "logps/chosen": -115.5,
      "logps/rejected": -266.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.203125,
      "rewards/margins": 18.25,
      "rewards/rejected": -17.125,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 0.00025666850467225894,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.140625,
      "logps/chosen": -110.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.53125,
      "rewards/margins": 19.125,
      "rewards/rejected": -17.625,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 8.707729010258159e-05,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.1796875,
      "logps/chosen": -109.0,
      "logps/rejected": -264.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.484375,
      "rewards/margins": 18.625,
      "rewards/rejected": -17.125,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.00020746386678032047,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.15625,
      "logps/chosen": -88.0,
      "logps/rejected": -270.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.625,
      "rewards/margins": 19.25,
      "rewards/rejected": -17.625,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 9.755428620835535e-05,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -1.265625,
      "logits/rejected": -1.1796875,
      "logps/chosen": -132.0,
      "logps/rejected": -264.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4140625,
      "rewards/margins": 18.0,
      "rewards/rejected": -16.625,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 0.0019240978142269086,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.1640625,
      "logps/chosen": -109.0,
      "logps/rejected": -280.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.828125,
      "rewards/margins": 19.875,
      "rewards/rejected": -18.125,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.008151304933624462,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.171875,
      "logps/chosen": -103.0,
      "logps/rejected": -256.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.71875,
      "rewards/margins": 18.0,
      "rewards/rejected": -16.25,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 0.0017297076073047298,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.15625,
      "logps/chosen": -104.0,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.609375,
      "rewards/margins": 19.125,
      "rewards/rejected": -17.5,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 0.0169796725593341,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.140625,
      "logps/chosen": -119.0,
      "logps/rejected": -270.0,
      "loss": 0.0009,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.515625,
      "rewards/margins": 18.125,
      "rewards/rejected": -17.625,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 0.002211490760678733,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -1.125,
      "logits/rejected": -1.1328125,
      "logps/chosen": -105.0,
      "logps/rejected": -286.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.27734375,
      "rewards/margins": 19.125,
      "rewards/rejected": -18.75,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 0.038971893726352264,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.09375,
      "logps/chosen": -123.5,
      "logps/rejected": -260.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.2890625,
      "rewards/margins": 17.0,
      "rewards/rejected": -17.375,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 0.0005523938555818159,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.1328125,
      "logps/chosen": -125.0,
      "logps/rejected": -276.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.43359375,
      "rewards/margins": 18.5,
      "rewards/rejected": -18.0,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.001508188141125635,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.125,
      "logps/chosen": -126.5,
      "logps/rejected": -276.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6796875,
      "rewards/margins": 18.75,
      "rewards/rejected": -18.125,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 0.000282269222742664,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -1.1640625,
      "logits/rejected": -1.1015625,
      "logps/chosen": -112.0,
      "logps/rejected": -270.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.94140625,
      "rewards/margins": 19.0,
      "rewards/rejected": -18.125,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 0.011861142581792664,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.140625,
      "logps/chosen": -111.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6640625,
      "rewards/margins": 18.375,
      "rewards/rejected": -17.75,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.014155663351858796,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.140625,
      "logps/chosen": -113.0,
      "logps/rejected": -258.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.734375,
      "rewards/margins": 17.75,
      "rewards/rejected": -17.0,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 0.0004801937477242006,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.1328125,
      "logps/chosen": -89.0,
      "logps/rejected": -288.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.3203125,
      "rewards/margins": 20.375,
      "rewards/rejected": -19.0,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 0.00018876422494270865,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.1328125,
      "logps/chosen": -117.0,
      "logps/rejected": -278.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.1875,
      "rewards/margins": 19.875,
      "rewards/rejected": -18.75,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.00021578457811446315,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -1.25,
      "logits/rejected": -1.1640625,
      "logps/chosen": -120.5,
      "logps/rejected": -278.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2421875,
      "rewards/margins": 19.25,
      "rewards/rejected": -18.0,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 6.197559111465712e-05,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -1.2109375,
      "logits/rejected": -1.125,
      "logps/chosen": -116.5,
      "logps/rejected": -286.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4375,
      "rewards/margins": 20.125,
      "rewards/rejected": -18.75,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 0.0004527745454286,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.125,
      "logps/chosen": -97.0,
      "logps/rejected": -284.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2578125,
      "rewards/margins": 20.0,
      "rewards/rejected": -18.75,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.0003072608527377377,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.1796875,
      "logps/chosen": -103.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4140625,
      "rewards/margins": 19.125,
      "rewards/rejected": -17.75,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 1.4102085909259622e-05,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.125,
      "logps/chosen": -118.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4921875,
      "rewards/margins": 19.125,
      "rewards/rejected": -17.625,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 0.008229293491203207,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.15625,
      "logps/chosen": -111.0,
      "logps/rejected": -270.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.265625,
      "rewards/margins": 18.875,
      "rewards/rejected": -17.625,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.0009905470831295794,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.1484375,
      "logps/chosen": -84.5,
      "logps/rejected": -274.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.640625,
      "rewards/margins": 19.75,
      "rewards/rejected": -18.125,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 0.0004928301143458307,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.1015625,
      "logps/chosen": -123.5,
      "logps/rejected": -268.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2421875,
      "rewards/margins": 19.0,
      "rewards/rejected": -17.75,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 0.0006560302576740441,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.1171875,
      "logps/chosen": -100.0,
      "logps/rejected": -288.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4609375,
      "rewards/margins": 20.625,
      "rewards/rejected": -19.125,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.15625,
      "eval_logits/rejected": -1.1171875,
      "eval_logps/chosen": -141.0,
      "eval_logps/rejected": -276.0,
      "eval_loss": 3.966204985772492e-06,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 1.4453125,
      "eval_rewards/margins": 19.5,
      "eval_rewards/rejected": -18.125,
      "eval_runtime": 15.1886,
      "eval_samples_per_second": 13.168,
      "eval_steps_per_second": 0.461,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.0077649630058205295,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.125,
      "logps/chosen": -107.5,
      "logps/rejected": -280.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.59375,
      "rewards/margins": 20.375,
      "rewards/rejected": -18.875,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 0.00032999132511516663,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -1.140625,
      "logits/rejected": -1.125,
      "logps/chosen": -111.0,
      "logps/rejected": -280.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5703125,
      "rewards/margins": 19.875,
      "rewards/rejected": -18.25,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 6.923858635050101e-05,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.1484375,
      "logps/chosen": -93.0,
      "logps/rejected": -276.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.484375,
      "rewards/margins": 19.875,
      "rewards/rejected": -18.25,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 7.462473136539333e-05,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.1875,
      "logps/chosen": -100.0,
      "logps/rejected": -264.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4921875,
      "rewards/margins": 19.0,
      "rewards/rejected": -17.5,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 0.0002834206448213677,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.1640625,
      "logps/chosen": -102.0,
      "logps/rejected": -268.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6015625,
      "rewards/margins": 19.25,
      "rewards/rejected": -17.625,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 0.0001231282211108334,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.1484375,
      "logps/chosen": -105.0,
      "logps/rejected": -266.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5703125,
      "rewards/margins": 19.25,
      "rewards/rejected": -17.625,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 0.00012015183446443852,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.140625,
      "logps/chosen": -133.0,
      "logps/rejected": -274.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.578125,
      "rewards/margins": 19.375,
      "rewards/rejected": -17.75,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 0.00022627234839590007,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -1.2421875,
      "logits/rejected": -1.140625,
      "logps/chosen": -110.5,
      "logps/rejected": -266.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.640625,
      "rewards/margins": 19.25,
      "rewards/rejected": -17.625,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 0.00012687874010370448,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.1328125,
      "logps/chosen": -97.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6875,
      "rewards/margins": 19.375,
      "rewards/rejected": -17.75,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.0008396801707681117,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -1.2109375,
      "logits/rejected": -1.1171875,
      "logps/chosen": -110.0,
      "logps/rejected": -288.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.78125,
      "rewards/margins": 21.0,
      "rewards/rejected": -19.25,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 1.2836392745225073e-05,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.15625,
      "logps/chosen": -99.5,
      "logps/rejected": -272.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.65625,
      "rewards/margins": 19.625,
      "rewards/rejected": -18.0,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 0.00012693793919841387,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.109375,
      "logps/chosen": -122.5,
      "logps/rejected": -292.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.3828125,
      "rewards/margins": 20.875,
      "rewards/rejected": -19.5,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 5.204424690298834e-05,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -1.2421875,
      "logits/rejected": -1.125,
      "logps/chosen": -119.0,
      "logps/rejected": -276.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.3984375,
      "rewards/margins": 19.5,
      "rewards/rejected": -18.125,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 0.00012545616212187522,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.140625,
      "logps/chosen": -124.5,
      "logps/rejected": -288.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.546875,
      "rewards/margins": 20.625,
      "rewards/rejected": -19.125,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 0.0005769569863281744,
      "learning_rate": 1e-07,
      "logits/chosen": -1.21875,
      "logits/rejected": -1.125,
      "logps/chosen": -105.5,
      "logps/rejected": -278.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.484375,
      "rewards/margins": 20.0,
      "rewards/rejected": -18.5,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.0009218268433051736,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.109375,
      "logps/chosen": -107.0,
      "logps/rejected": -276.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6328125,
      "rewards/margins": 19.875,
      "rewards/rejected": -18.25,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 4.7673082430515324e-05,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.1015625,
      "logps/chosen": -114.0,
      "logps/rejected": -284.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.578125,
      "rewards/margins": 20.5,
      "rewards/rejected": -19.0,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 0.014277472004175404,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -1.2421875,
      "logits/rejected": -1.1328125,
      "logps/chosen": -88.0,
      "logps/rejected": -278.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6484375,
      "rewards/margins": 20.375,
      "rewards/rejected": -18.625,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 8.448849827923423e-05,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -1.2578125,
      "logits/rejected": -1.125,
      "logps/chosen": -118.5,
      "logps/rejected": -280.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5078125,
      "rewards/margins": 20.25,
      "rewards/rejected": -18.75,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 0.00556288956188279,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.140625,
      "logps/chosen": -101.0,
      "logps/rejected": -284.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.71875,
      "rewards/margins": 20.375,
      "rewards/rejected": -18.75,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 0.004449605087331906,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.1171875,
      "logps/chosen": -98.0,
      "logps/rejected": -288.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.75,
      "rewards/margins": 20.875,
      "rewards/rejected": -19.125,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 0.00017082233765247526,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -1.1953125,
      "logits/rejected": -1.1015625,
      "logps/chosen": -115.0,
      "logps/rejected": -278.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5703125,
      "rewards/margins": 20.375,
      "rewards/rejected": -18.75,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 6.476831280203482e-05,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -1.203125,
      "logits/rejected": -1.125,
      "logps/chosen": -125.0,
      "logps/rejected": -284.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.53125,
      "rewards/margins": 20.25,
      "rewards/rejected": -18.75,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 0.0005168510602309617,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -1.1796875,
      "logits/rejected": -1.1328125,
      "logps/chosen": -113.5,
      "logps/rejected": -278.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5390625,
      "rewards/margins": 19.875,
      "rewards/rejected": -18.375,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 0.000897340761523617,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -1.2109375,
      "logits/rejected": -1.125,
      "logps/chosen": -87.5,
      "logps/rejected": -288.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.703125,
      "rewards/margins": 21.0,
      "rewards/rejected": -19.25,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 2.971586501256541e-05,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -1.171875,
      "logits/rejected": -1.109375,
      "logps/chosen": -80.0,
      "logps/rejected": -286.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.8671875,
      "rewards/margins": 21.0,
      "rewards/rejected": -19.125,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 0.0033930981967334036,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -1.2421875,
      "logits/rejected": -1.1328125,
      "logps/chosen": -96.5,
      "logps/rejected": -280.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.7578125,
      "rewards/margins": 20.5,
      "rewards/rejected": -18.75,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.00026591254711212035,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.125,
      "logps/chosen": -91.5,
      "logps/rejected": -280.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.7265625,
      "rewards/margins": 20.625,
      "rewards/rejected": -18.875,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 6.534397185429311e-05,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -1.2265625,
      "logits/rejected": -1.1171875,
      "logps/chosen": -102.5,
      "logps/rejected": -282.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.640625,
      "rewards/margins": 20.0,
      "rewards/rejected": -18.375,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 5.403314331116278e-05,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -1.1875,
      "logits/rejected": -1.140625,
      "logps/chosen": -87.0,
      "logps/rejected": -284.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.78125,
      "rewards/margins": 20.75,
      "rewards/rejected": -18.875,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 9.594030191690998e-05,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -1.234375,
      "logits/rejected": -1.1328125,
      "logps/chosen": -119.5,
      "logps/rejected": -282.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6328125,
      "rewards/margins": 20.25,
      "rewards/rejected": -18.625,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.15625,
      "eval_logits/rejected": -1.1171875,
      "eval_logps/chosen": -138.0,
      "eval_logps/rejected": -276.0,
      "eval_loss": 1.7006927919283044e-06,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 1.75,
      "eval_rewards/margins": 19.875,
      "eval_rewards/rejected": -18.0,
      "eval_runtime": 16.9846,
      "eval_samples_per_second": 11.775,
      "eval_steps_per_second": 0.412,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.020377112339787087,
      "train_runtime": 5959.6747,
      "train_samples_per_second": 5.034,
      "train_steps_per_second": 0.158
    }
  ],
  "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
}