{
  "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": 93.47416883117657,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -3.984375,
      "logits/rejected": -3.859375,
      "logps/chosen": -141.0,
      "logps/rejected": -97.0,
      "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": 87.75391323145124,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -4.09375,
      "logits/rejected": -4.125,
      "logps/chosen": -131.0,
      "logps/rejected": -106.5,
      "loss": 0.6825,
      "rewards/accuracies": 0.3611111044883728,
      "rewards/chosen": -0.00103759765625,
      "rewards/margins": 0.01214599609375,
      "rewards/rejected": -0.01318359375,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 75.41700547336782,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -3.90625,
      "logits/rejected": -3.921875,
      "logps/chosen": -129.0,
      "logps/rejected": -111.5,
      "loss": 0.545,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.04638671875,
      "rewards/margins": 0.375,
      "rewards/rejected": -0.328125,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 40.75333838571047,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -3.875,
      "logits/rejected": -3.84375,
      "logps/chosen": -101.5,
      "logps/rejected": -124.0,
      "loss": 0.2353,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.169921875,
      "rewards/margins": 1.828125,
      "rewards/rejected": -1.65625,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 13.43470393143394,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -3.59375,
      "logits/rejected": -3.59375,
      "logps/chosen": -119.5,
      "logps/rejected": -152.0,
      "loss": 0.0636,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1416015625,
      "rewards/margins": 4.5,
      "rewards/rejected": -4.34375,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 4.492643052908927,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -3.46875,
      "logits/rejected": -3.453125,
      "logps/chosen": -121.5,
      "logps/rejected": -177.0,
      "loss": 0.0497,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.025634765625,
      "rewards/margins": 6.53125,
      "rewards/rejected": -6.5,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 1.4050188499898384,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -3.65625,
      "logits/rejected": -3.609375,
      "logps/chosen": -125.0,
      "logps/rejected": -186.0,
      "loss": 0.0263,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.09033203125,
      "rewards/margins": 8.3125,
      "rewards/rejected": -8.1875,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 1.686556030425558,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -3.421875,
      "logits/rejected": -3.40625,
      "logps/chosen": -136.0,
      "logps/rejected": -211.0,
      "loss": 0.012,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.04931640625,
      "rewards/margins": 10.3125,
      "rewards/rejected": -10.3125,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 15.791799446785957,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -3.421875,
      "logits/rejected": -3.375,
      "logps/chosen": -115.5,
      "logps/rejected": -229.0,
      "loss": 0.0178,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.41015625,
      "rewards/margins": 12.6875,
      "rewards/rejected": -12.3125,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 3.6084615404152705,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -3.375,
      "logits/rejected": -3.4375,
      "logps/chosen": -125.0,
      "logps/rejected": -244.0,
      "loss": 0.0105,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.609375,
      "rewards/margins": 13.75,
      "rewards/rejected": -13.125,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 11.896415949522481,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -3.375,
      "logits/rejected": -3.390625,
      "logps/chosen": -108.5,
      "logps/rejected": -242.0,
      "loss": 0.0052,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.173828125,
      "rewards/margins": 13.5,
      "rewards/rejected": -13.375,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 18.67185475239393,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.265625,
      "logps/chosen": -120.5,
      "logps/rejected": -244.0,
      "loss": 0.0056,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.1796875,
      "rewards/margins": 13.6875,
      "rewards/rejected": -13.8125,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.3483119672235051,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.171875,
      "logps/chosen": -131.0,
      "logps/rejected": -294.0,
      "loss": 0.0026,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.53125,
      "rewards/margins": 18.0,
      "rewards/rejected": -18.5,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 1.2235240871946855,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -3.1875,
      "logits/rejected": -3.21875,
      "logps/chosen": -116.5,
      "logps/rejected": -266.0,
      "loss": 0.0022,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.0673828125,
      "rewards/margins": 16.25,
      "rewards/rejected": -16.25,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 5.806636810315318,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -3.1875,
      "logits/rejected": -3.171875,
      "logps/chosen": -137.0,
      "logps/rejected": -270.0,
      "loss": 0.0222,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.65625,
      "rewards/margins": 15.625,
      "rewards/rejected": -16.25,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 0.022978299592214306,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.15625,
      "logps/chosen": -147.0,
      "logps/rejected": -284.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.244140625,
      "rewards/margins": 17.75,
      "rewards/rejected": -18.0,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 0.04941514676403275,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.125,
      "logps/chosen": -119.0,
      "logps/rejected": -296.0,
      "loss": 0.0016,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3984375,
      "rewards/margins": 19.0,
      "rewards/rejected": -18.625,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 19.8894352185433,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.15625,
      "logps/chosen": -113.0,
      "logps/rejected": -292.0,
      "loss": 0.0007,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1630859375,
      "rewards/margins": 18.625,
      "rewards/rejected": -18.5,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.13137715719798598,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -3.21875,
      "logits/rejected": -3.1875,
      "logps/chosen": -134.0,
      "logps/rejected": -308.0,
      "loss": 0.0019,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.53125,
      "rewards/margins": 19.75,
      "rewards/rejected": -20.25,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 0.013764483728973553,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.234375,
      "logps/chosen": -114.0,
      "logps/rejected": -320.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.08056640625,
      "rewards/margins": 21.0,
      "rewards/rejected": -21.125,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 3.046295143956079,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -3.203125,
      "logits/rejected": -3.203125,
      "logps/chosen": -103.5,
      "logps/rejected": -320.0,
      "loss": 0.0049,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1064453125,
      "rewards/margins": 21.5,
      "rewards/rejected": -21.625,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.10031611497345358,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.171875,
      "logps/chosen": -130.0,
      "logps/rejected": -332.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.048583984375,
      "rewards/margins": 22.375,
      "rewards/rejected": -22.5,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 1.1092368572457065,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -3.1875,
      "logits/rejected": -3.1875,
      "logps/chosen": -116.5,
      "logps/rejected": -296.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.38671875,
      "rewards/margins": 18.25,
      "rewards/rejected": -18.75,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 0.018829348075951137,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.09375,
      "logps/chosen": -110.5,
      "logps/rejected": -320.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.12890625,
      "rewards/margins": 21.25,
      "rewards/rejected": -21.375,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.011496093293855175,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -3.21875,
      "logits/rejected": -3.1875,
      "logps/chosen": -114.5,
      "logps/rejected": -310.0,
      "loss": 0.0028,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.251953125,
      "rewards/margins": 20.5,
      "rewards/rejected": -20.25,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 0.28952088206305154,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -3.21875,
      "logits/rejected": -3.265625,
      "logps/chosen": -137.0,
      "logps/rejected": -304.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.33203125,
      "rewards/margins": 19.625,
      "rewards/rejected": -19.25,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 0.043962484735603484,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.1875,
      "logps/chosen": -125.5,
      "logps/rejected": -308.0,
      "loss": 0.003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.482421875,
      "rewards/margins": 19.375,
      "rewards/rejected": -19.875,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.03393181975047915,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -3.234375,
      "logits/rejected": -3.171875,
      "logps/chosen": -112.0,
      "logps/rejected": -310.0,
      "loss": 0.0011,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.177734375,
      "rewards/margins": 20.75,
      "rewards/rejected": -20.625,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 0.9663528991851124,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.203125,
      "logps/chosen": -104.5,
      "logps/rejected": -316.0,
      "loss": 0.0009,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1953125,
      "rewards/margins": 20.875,
      "rewards/rejected": -21.0,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 0.09196633155046754,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -3.1875,
      "logits/rejected": -3.171875,
      "logps/chosen": -147.0,
      "logps/rejected": -318.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1640625,
      "rewards/margins": 20.125,
      "rewards/rejected": -21.25,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.07116225512391497,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -3.140625,
      "logits/rejected": -3.125,
      "logps/chosen": -130.0,
      "logps/rejected": -334.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.9296875,
      "rewards/margins": 21.875,
      "rewards/rejected": -22.75,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 0.01618388689873247,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -3.125,
      "logits/rejected": -3.125,
      "logps/chosen": -152.0,
      "logps/rejected": -316.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.3515625,
      "rewards/margins": 20.375,
      "rewards/rejected": -20.75,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -3.234375,
      "eval_logits/rejected": -3.171875,
      "eval_logps/chosen": -172.0,
      "eval_logps/rejected": -310.0,
      "eval_loss": 0.004244150593876839,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.376953125,
      "eval_rewards/margins": 19.25,
      "eval_rewards/rejected": -19.625,
      "eval_runtime": 13.8267,
      "eval_samples_per_second": 14.465,
      "eval_steps_per_second": 0.506,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 0.014031360468919892,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -3.140625,
      "logits/rejected": -3.15625,
      "logps/chosen": -128.0,
      "logps/rejected": -330.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.61328125,
      "rewards/margins": 21.25,
      "rewards/rejected": -22.0,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.01839801735368299,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.0625,
      "logps/chosen": -135.0,
      "logps/rejected": -312.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.5703125,
      "rewards/margins": 19.75,
      "rewards/rejected": -21.375,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 0.016992373953575654,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -3.015625,
      "logits/rejected": -3.0625,
      "logps/chosen": -105.0,
      "logps/rejected": -338.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.859375,
      "rewards/margins": 22.625,
      "rewards/rejected": -23.5,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 0.00648353239139669,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.0625,
      "logps/chosen": -134.0,
      "logps/rejected": -310.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.10009765625,
      "rewards/margins": 20.5,
      "rewards/rejected": -20.625,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.018529336855778705,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.0625,
      "logps/chosen": -121.0,
      "logps/rejected": -316.0,
      "loss": 0.0009,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1298828125,
      "rewards/margins": 21.25,
      "rewards/rejected": -21.125,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 0.0025056934286525165,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -3.15625,
      "logits/rejected": -3.09375,
      "logps/chosen": -116.5,
      "logps/rejected": -314.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.392578125,
      "rewards/margins": 21.0,
      "rewards/rejected": -20.5,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 0.0014546590073500464,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -3.125,
      "logits/rejected": -3.09375,
      "logps/chosen": -114.5,
      "logps/rejected": -318.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2421875,
      "rewards/margins": 21.125,
      "rewards/rejected": -20.875,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.02764778364579726,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -3.078125,
      "logits/rejected": -3.03125,
      "logps/chosen": -108.0,
      "logps/rejected": -334.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1494140625,
      "rewards/margins": 22.875,
      "rewards/rejected": -23.0,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 0.002824886589609134,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -3.171875,
      "logits/rejected": -3.15625,
      "logps/chosen": -147.0,
      "logps/rejected": -322.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.259765625,
      "rewards/margins": 21.0,
      "rewards/rejected": -21.25,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 0.003952105532040841,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -3.125,
      "logits/rejected": -3.0625,
      "logps/chosen": -123.5,
      "logps/rejected": -344.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.201171875,
      "rewards/margins": 23.875,
      "rewards/rejected": -23.625,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.0010212333678410641,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.109375,
      "logps/chosen": -128.0,
      "logps/rejected": -344.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1494140625,
      "rewards/margins": 23.5,
      "rewards/rejected": -23.75,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 0.026417806261673966,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -3.140625,
      "logits/rejected": -3.109375,
      "logps/chosen": -128.0,
      "logps/rejected": -346.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.234375,
      "rewards/margins": 23.75,
      "rewards/rejected": -24.0,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 0.33120053981322833,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.09375,
      "logps/chosen": -128.0,
      "logps/rejected": -360.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.22265625,
      "rewards/margins": 24.875,
      "rewards/rejected": -25.125,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 5.0669153760203136e-05,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -3.03125,
      "logits/rejected": -3.109375,
      "logps/chosen": -113.5,
      "logps/rejected": -348.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.0093994140625,
      "rewards/margins": 23.875,
      "rewards/rejected": -23.875,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 0.002235783476143995,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.015625,
      "logps/chosen": -122.5,
      "logps/rejected": -346.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.42578125,
      "rewards/margins": 23.625,
      "rewards/rejected": -24.0,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 0.0021115625525813,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.046875,
      "logps/chosen": -127.5,
      "logps/rejected": -352.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.2490234375,
      "rewards/margins": 24.5,
      "rewards/rejected": -24.75,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.0005420939322641936,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.015625,
      "logps/chosen": -148.0,
      "logps/rejected": -358.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8828125,
      "rewards/margins": 24.125,
      "rewards/rejected": -25.0,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 0.007358687931687971,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.046875,
      "logps/chosen": -125.0,
      "logps/rejected": -360.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.625,
      "rewards/margins": 24.75,
      "rewards/rejected": -25.375,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 0.0654742270230356,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.046875,
      "logps/chosen": -124.0,
      "logps/rejected": -334.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.046875,
      "rewards/margins": 21.875,
      "rewards/rejected": -22.875,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.0008641901527854177,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.078125,
      "logps/chosen": -131.0,
      "logps/rejected": -342.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.69140625,
      "rewards/margins": 22.875,
      "rewards/rejected": -23.625,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 0.015004697710408908,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -3.015625,
      "logits/rejected": -3.0625,
      "logps/chosen": -100.5,
      "logps/rejected": -374.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.0927734375,
      "rewards/margins": 26.5,
      "rewards/rejected": -26.625,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 0.01615930039253711,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.0,
      "logps/chosen": -130.0,
      "logps/rejected": -336.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.010009765625,
      "rewards/margins": 23.25,
      "rewards/rejected": -23.25,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.005625013393470251,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -3.015625,
      "logits/rejected": -3.015625,
      "logps/chosen": -123.0,
      "logps/rejected": -338.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.482421875,
      "rewards/margins": 23.375,
      "rewards/rejected": -22.875,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 0.006628504166462446,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -3.109375,
      "logits/rejected": -3.0625,
      "logps/chosen": -128.0,
      "logps/rejected": -324.0,
      "loss": 0.0008,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1650390625,
      "rewards/margins": 21.75,
      "rewards/rejected": -21.5,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 0.0051543594641130655,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -3.0625,
      "logits/rejected": -3.0,
      "logps/chosen": -102.0,
      "logps/rejected": -336.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.283203125,
      "rewards/margins": 23.25,
      "rewards/rejected": -23.0,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.0019828665244342454,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -2.890625,
      "logits/rejected": -2.9375,
      "logps/chosen": -112.0,
      "logps/rejected": -328.0,
      "loss": 0.001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.051513671875,
      "rewards/margins": 22.375,
      "rewards/rejected": -22.375,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 0.0010506656519779034,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -3.03125,
      "logits/rejected": -2.953125,
      "logps/chosen": -133.0,
      "logps/rejected": -342.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.11767578125,
      "rewards/margins": 23.125,
      "rewards/rejected": -23.25,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 0.026254564073093664,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -2.9375,
      "logits/rejected": -2.921875,
      "logps/chosen": -126.0,
      "logps/rejected": -332.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1484375,
      "rewards/margins": 22.0,
      "rewards/rejected": -22.125,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.0011075869246766719,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -2.921875,
      "logits/rejected": -2.875,
      "logps/chosen": -97.0,
      "logps/rejected": -326.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.12109375,
      "rewards/margins": 22.375,
      "rewards/rejected": -22.25,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 3.965796587839269,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -2.9375,
      "logits/rejected": -2.890625,
      "logps/chosen": -140.0,
      "logps/rejected": -338.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.404296875,
      "rewards/margins": 22.75,
      "rewards/rejected": -23.125,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 0.14930442211812853,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -2.921875,
      "logits/rejected": -2.890625,
      "logps/chosen": -116.5,
      "logps/rejected": -370.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.9765625,
      "rewards/margins": 25.25,
      "rewards/rejected": -26.25,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -3.03125,
      "eval_logits/rejected": -2.953125,
      "eval_logps/chosen": -175.0,
      "eval_logps/rejected": -336.0,
      "eval_loss": 0.0031948494724929333,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.7578125,
      "eval_rewards/margins": 21.375,
      "eval_rewards/rejected": -22.125,
      "eval_runtime": 13.9279,
      "eval_samples_per_second": 14.36,
      "eval_steps_per_second": 0.503,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.0032974843044556753,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -2.859375,
      "logits/rejected": -2.84375,
      "logps/chosen": -133.0,
      "logps/rejected": -346.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.486328125,
      "rewards/margins": 23.5,
      "rewards/rejected": -24.0,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 0.1432508053581138,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -2.96875,
      "logits/rejected": -2.9375,
      "logps/chosen": -135.0,
      "logps/rejected": -342.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.39453125,
      "rewards/margins": 22.875,
      "rewards/rejected": -23.25,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 0.0001731085607822556,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -2.765625,
      "logits/rejected": -2.796875,
      "logps/chosen": -109.0,
      "logps/rejected": -374.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.19921875,
      "rewards/margins": 26.625,
      "rewards/rejected": -26.75,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 0.0019185996024299748,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -2.859375,
      "logits/rejected": -2.8125,
      "logps/chosen": -122.0,
      "logps/rejected": -342.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.77734375,
      "rewards/margins": 23.0,
      "rewards/rejected": -23.75,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 0.0008784455055030455,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -2.859375,
      "logits/rejected": -2.859375,
      "logps/chosen": -125.0,
      "logps/rejected": -356.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.478515625,
      "rewards/margins": 24.0,
      "rewards/rejected": -24.5,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 0.009325179732468034,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -2.828125,
      "logits/rejected": -2.828125,
      "logps/chosen": -126.5,
      "logps/rejected": -358.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.5234375,
      "rewards/margins": 24.5,
      "rewards/rejected": -25.0,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 0.0007194651213122703,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -2.9375,
      "logits/rejected": -2.859375,
      "logps/chosen": -158.0,
      "logps/rejected": -362.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.90625,
      "rewards/margins": 23.875,
      "rewards/rejected": -24.875,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 0.008732155259144234,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -2.78125,
      "logits/rejected": -2.765625,
      "logps/chosen": -126.5,
      "logps/rejected": -368.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.765625,
      "rewards/margins": 25.625,
      "rewards/rejected": -26.375,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 0.0005971006192250428,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -2.8125,
      "logits/rejected": -2.875,
      "logps/chosen": -114.5,
      "logps/rejected": -344.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.46875,
      "rewards/margins": 23.125,
      "rewards/rejected": -23.625,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.00013779482104364402,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -2.828125,
      "logits/rejected": -2.859375,
      "logps/chosen": -134.0,
      "logps/rejected": -364.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.5234375,
      "rewards/margins": 24.875,
      "rewards/rejected": -25.375,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 6.075224878814938e-05,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -2.796875,
      "logits/rejected": -2.828125,
      "logps/chosen": -121.0,
      "logps/rejected": -358.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8046875,
      "rewards/margins": 24.125,
      "rewards/rejected": -25.0,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 0.0007165926429755341,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -2.921875,
      "logits/rejected": -2.8125,
      "logps/chosen": -137.0,
      "logps/rejected": -374.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.498046875,
      "rewards/margins": 26.125,
      "rewards/rejected": -26.625,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 0.00026410738853004405,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -2.796875,
      "logits/rejected": -2.8125,
      "logps/chosen": -146.0,
      "logps/rejected": -366.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.96875,
      "rewards/margins": 24.75,
      "rewards/rejected": -25.75,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 0.02405558435283539,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -2.78125,
      "logits/rejected": -2.8125,
      "logps/chosen": -143.0,
      "logps/rejected": -368.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.82421875,
      "rewards/margins": 24.875,
      "rewards/rejected": -25.75,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 0.00034653212382964406,
      "learning_rate": 1e-07,
      "logits/chosen": -2.84375,
      "logits/rejected": -2.84375,
      "logps/chosen": -126.0,
      "logps/rejected": -370.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.69921875,
      "rewards/margins": 25.625,
      "rewards/rejected": -26.25,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.0015560065245145238,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -2.78125,
      "logits/rejected": -2.796875,
      "logps/chosen": -128.0,
      "logps/rejected": -344.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.51171875,
      "rewards/margins": 23.375,
      "rewards/rejected": -23.875,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 0.00033309018788403783,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -2.921875,
      "logits/rejected": -2.859375,
      "logps/chosen": -134.0,
      "logps/rejected": -364.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.357421875,
      "rewards/margins": 25.75,
      "rewards/rejected": -26.0,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 0.0025148503211950596,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -2.78125,
      "logits/rejected": -2.78125,
      "logps/chosen": -104.5,
      "logps/rejected": -382.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.470703125,
      "rewards/margins": 27.0,
      "rewards/rejected": -27.5,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 0.00048763191845162235,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -2.90625,
      "logits/rejected": -2.84375,
      "logps/chosen": -139.0,
      "logps/rejected": -370.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.53125,
      "rewards/margins": 25.875,
      "rewards/rejected": -26.375,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 0.0010336427914530199,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -2.796875,
      "logits/rejected": -2.84375,
      "logps/chosen": -119.0,
      "logps/rejected": -372.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.435546875,
      "rewards/margins": 25.75,
      "rewards/rejected": -26.25,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 9.273196336389341e-05,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -2.765625,
      "logits/rejected": -2.765625,
      "logps/chosen": -118.5,
      "logps/rejected": -366.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.271484375,
      "rewards/margins": 25.625,
      "rewards/rejected": -25.875,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 0.0004326121515741712,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -2.828125,
      "logits/rejected": -2.75,
      "logps/chosen": -139.0,
      "logps/rejected": -356.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.5703125,
      "rewards/margins": 24.375,
      "rewards/rejected": -25.0,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 0.0013592593922840503,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -2.859375,
      "logits/rejected": -2.78125,
      "logps/chosen": -145.0,
      "logps/rejected": -360.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.55859375,
      "rewards/margins": 25.0,
      "rewards/rejected": -25.5,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 0.00597818443333345,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -2.953125,
      "logits/rejected": -2.859375,
      "logps/chosen": -135.0,
      "logps/rejected": -380.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.6484375,
      "rewards/margins": 26.375,
      "rewards/rejected": -27.0,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 0.0013466028912008678,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -2.765625,
      "logits/rejected": -2.796875,
      "logps/chosen": -107.0,
      "logps/rejected": -366.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.37109375,
      "rewards/margins": 25.75,
      "rewards/rejected": -26.125,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 2.7058882274462487e-05,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -2.78125,
      "logits/rejected": -2.78125,
      "logps/chosen": -99.0,
      "logps/rejected": -376.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1787109375,
      "rewards/margins": 27.125,
      "rewards/rejected": -27.25,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 0.002179423905695273,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -2.84375,
      "logits/rejected": -2.84375,
      "logps/chosen": -118.5,
      "logps/rejected": -376.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.5703125,
      "rewards/margins": 26.375,
      "rewards/rejected": -26.875,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.0006890554421005211,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -2.765625,
      "logits/rejected": -2.75,
      "logps/chosen": -112.0,
      "logps/rejected": -364.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.42578125,
      "rewards/margins": 25.75,
      "rewards/rejected": -26.25,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 0.0017781976127544882,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -2.8125,
      "logits/rejected": -2.84375,
      "logps/chosen": -121.5,
      "logps/rejected": -350.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.67578125,
      "rewards/margins": 23.875,
      "rewards/rejected": -24.5,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 0.00157882752297583,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -2.765625,
      "logits/rejected": -2.78125,
      "logps/chosen": -105.5,
      "logps/rejected": -388.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.1259765625,
      "rewards/margins": 28.0,
      "rewards/rejected": -28.125,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 0.0018180020601003572,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -2.828125,
      "logits/rejected": -2.8125,
      "logps/chosen": -139.0,
      "logps/rejected": -370.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.4609375,
      "rewards/margins": 25.875,
      "rewards/rejected": -26.25,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -2.96875,
      "eval_logits/rejected": -2.90625,
      "eval_logps/chosen": -173.0,
      "eval_logps/rejected": -350.0,
      "eval_loss": 0.004805178381502628,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.515625,
      "eval_rewards/margins": 23.0,
      "eval_rewards/rejected": -23.625,
      "eval_runtime": 16.1771,
      "eval_samples_per_second": 12.363,
      "eval_steps_per_second": 0.433,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.01815998419456569,
      "train_runtime": 5884.866,
      "train_samples_per_second": 5.098,
      "train_steps_per_second": 0.16
    }
  ],
  "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
}