{
  "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": 56.72510549759321,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -3.921875,
      "logits/rejected": -4.03125,
      "logps/chosen": -179.0,
      "logps/rejected": -159.0,
      "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": 56.446808237449936,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -3.75,
      "logits/rejected": -3.765625,
      "logps/chosen": -156.0,
      "logps/rejected": -132.0,
      "loss": 0.6947,
      "rewards/accuracies": 0.125,
      "rewards/chosen": 0.0013885498046875,
      "rewards/margins": 0.0020904541015625,
      "rewards/rejected": -0.000701904296875,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 50.63022580127569,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.65625,
      "logps/chosen": -161.0,
      "logps/rejected": -148.0,
      "loss": 0.6905,
      "rewards/accuracies": 0.21250000596046448,
      "rewards/chosen": -0.0031280517578125,
      "rewards/margins": 0.0093994140625,
      "rewards/rejected": -0.01251220703125,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 53.70156897517662,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.671875,
      "logps/chosen": -172.0,
      "logps/rejected": -145.0,
      "loss": 0.6881,
      "rewards/accuracies": 0.3125,
      "rewards/chosen": -0.01190185546875,
      "rewards/margins": 0.01953125,
      "rewards/rejected": -0.031494140625,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 50.874400763352966,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -3.71875,
      "logits/rejected": -3.734375,
      "logps/chosen": -154.0,
      "logps/rejected": -133.0,
      "loss": 0.6825,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.053466796875,
      "rewards/margins": 0.037841796875,
      "rewards/rejected": -0.09130859375,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 43.47430054265056,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -3.75,
      "logits/rejected": -3.6875,
      "logps/chosen": -153.0,
      "logps/rejected": -144.0,
      "loss": 0.6703,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.0927734375,
      "rewards/margins": 0.05126953125,
      "rewards/rejected": -0.14453125,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 50.8916442938159,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.6875,
      "logps/chosen": -158.0,
      "logps/rejected": -161.0,
      "loss": 0.6756,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.1318359375,
      "rewards/margins": 0.0057373046875,
      "rewards/rejected": -0.1376953125,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 48.38154340571634,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.640625,
      "logps/chosen": -153.0,
      "logps/rejected": -136.0,
      "loss": 0.6529,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.1162109375,
      "rewards/margins": 0.09765625,
      "rewards/rejected": -0.2138671875,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 48.16199178296427,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.734375,
      "logps/chosen": -170.0,
      "logps/rejected": -141.0,
      "loss": 0.647,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.030517578125,
      "rewards/margins": 0.173828125,
      "rewards/rejected": -0.2041015625,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 51.32059362049273,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -3.84375,
      "logits/rejected": -3.796875,
      "logps/chosen": -155.0,
      "logps/rejected": -156.0,
      "loss": 0.6429,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.10693359375,
      "rewards/margins": 0.2490234375,
      "rewards/rejected": -0.35546875,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 45.997248976509255,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.828125,
      "logps/chosen": -159.0,
      "logps/rejected": -160.0,
      "loss": 0.6359,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.3828125,
      "rewards/margins": 0.109375,
      "rewards/rejected": -0.4921875,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 61.80389122802718,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -3.828125,
      "logits/rejected": -3.765625,
      "logps/chosen": -160.0,
      "logps/rejected": -165.0,
      "loss": 0.6107,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.173828125,
      "rewards/margins": 0.248046875,
      "rewards/rejected": -0.421875,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 47.761602240808664,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.734375,
      "logps/chosen": -168.0,
      "logps/rejected": -147.0,
      "loss": 0.5903,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.50390625,
      "rewards/margins": 0.400390625,
      "rewards/rejected": -0.90625,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 55.99986050648511,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.671875,
      "logps/chosen": -179.0,
      "logps/rejected": -156.0,
      "loss": 0.628,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.7578125,
      "rewards/margins": 0.3359375,
      "rewards/rejected": -1.09375,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 44.26545998046523,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.78125,
      "logps/chosen": -150.0,
      "logps/rejected": -158.0,
      "loss": 0.6808,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.76953125,
      "rewards/margins": 0.173828125,
      "rewards/rejected": -0.9453125,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 59.88747780116159,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.71875,
      "logps/chosen": -169.0,
      "logps/rejected": -165.0,
      "loss": 0.6581,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.6796875,
      "rewards/margins": 0.2734375,
      "rewards/rejected": -0.953125,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 46.3881925450707,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.671875,
      "logps/chosen": -169.0,
      "logps/rejected": -156.0,
      "loss": 0.5735,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.68359375,
      "rewards/margins": 0.41015625,
      "rewards/rejected": -1.09375,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 50.59972622792833,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.734375,
      "logps/chosen": -160.0,
      "logps/rejected": -157.0,
      "loss": 0.624,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.93359375,
      "rewards/margins": 0.298828125,
      "rewards/rejected": -1.234375,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 54.54322374285826,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -3.71875,
      "logits/rejected": -3.625,
      "logps/chosen": -171.0,
      "logps/rejected": -149.0,
      "loss": 0.637,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -1.0390625,
      "rewards/margins": 0.5078125,
      "rewards/rejected": -1.546875,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 45.707175747041994,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.734375,
      "logps/chosen": -172.0,
      "logps/rejected": -159.0,
      "loss": 0.6271,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.015625,
      "rewards/margins": 0.4921875,
      "rewards/rejected": -1.515625,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 51.14627583309026,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.65625,
      "logps/chosen": -171.0,
      "logps/rejected": -168.0,
      "loss": 0.6573,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.65625,
      "rewards/margins": 0.447265625,
      "rewards/rejected": -1.1015625,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 46.16132516381053,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.71875,
      "logps/chosen": -170.0,
      "logps/rejected": -154.0,
      "loss": 0.5974,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.56640625,
      "rewards/margins": 0.33984375,
      "rewards/rejected": -0.90625,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 53.323383878395134,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -3.75,
      "logits/rejected": -3.78125,
      "logps/chosen": -155.0,
      "logps/rejected": -158.0,
      "loss": 0.6738,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.703125,
      "rewards/margins": 0.419921875,
      "rewards/rejected": -1.125,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 43.309407700101836,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -3.75,
      "logits/rejected": -3.71875,
      "logps/chosen": -154.0,
      "logps/rejected": -134.0,
      "loss": 0.6227,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.7265625,
      "rewards/margins": 0.431640625,
      "rewards/rejected": -1.15625,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 159.9832633337909,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.78125,
      "logps/chosen": -165.0,
      "logps/rejected": -160.0,
      "loss": 0.5807,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.72265625,
      "rewards/margins": 0.65625,
      "rewards/rejected": -1.3828125,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 51.22221971619883,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -3.84375,
      "logits/rejected": -3.8125,
      "logps/chosen": -176.0,
      "logps/rejected": -161.0,
      "loss": 0.657,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.87109375,
      "rewards/margins": 0.3359375,
      "rewards/rejected": -1.203125,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 47.46463976816363,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.796875,
      "logps/chosen": -150.0,
      "logps/rejected": -135.0,
      "loss": 0.721,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.7109375,
      "rewards/margins": 0.5390625,
      "rewards/rejected": -1.25,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 65.57176539431565,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.71875,
      "logps/chosen": -175.0,
      "logps/rejected": -177.0,
      "loss": 0.5915,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.46875,
      "rewards/margins": 0.64453125,
      "rewards/rejected": -1.1171875,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 50.06111100721229,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -3.703125,
      "logits/rejected": -3.671875,
      "logps/chosen": -194.0,
      "logps/rejected": -156.0,
      "loss": 0.5965,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.392578125,
      "rewards/margins": 0.734375,
      "rewards/rejected": -1.125,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 60.19864413150602,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -3.71875,
      "logits/rejected": -3.765625,
      "logps/chosen": -162.0,
      "logps/rejected": -152.0,
      "loss": 0.6536,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.74609375,
      "rewards/margins": 0.154296875,
      "rewards/rejected": -0.8984375,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 49.15565959758055,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.765625,
      "logps/chosen": -184.0,
      "logps/rejected": -172.0,
      "loss": 0.6022,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.64453125,
      "rewards/margins": 0.482421875,
      "rewards/rejected": -1.125,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 49.85370609782219,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -3.6875,
      "logits/rejected": -3.71875,
      "logps/chosen": -198.0,
      "logps/rejected": -175.0,
      "loss": 0.6208,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.80859375,
      "rewards/margins": 0.578125,
      "rewards/rejected": -1.3828125,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -3.828125,
      "eval_logits/rejected": -3.84375,
      "eval_logps/chosen": -155.0,
      "eval_logps/rejected": -145.0,
      "eval_loss": 0.5767968893051147,
      "eval_rewards/accuracies": 0.7142857313156128,
      "eval_rewards/chosen": -0.419921875,
      "eval_rewards/margins": 0.59375,
      "eval_rewards/rejected": -1.015625,
      "eval_runtime": 12.5477,
      "eval_samples_per_second": 15.939,
      "eval_steps_per_second": 0.558,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 32.8934140269683,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -3.671875,
      "logits/rejected": -3.640625,
      "logps/chosen": -180.0,
      "logps/rejected": -154.0,
      "loss": 0.3807,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.423828125,
      "rewards/margins": 1.359375,
      "rewards/rejected": -1.78125,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 26.293321659480707,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.734375,
      "logps/chosen": -165.0,
      "logps/rejected": -166.0,
      "loss": 0.2736,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.240234375,
      "rewards/margins": 1.8125,
      "rewards/rejected": -2.0625,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 18.530023934222726,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.796875,
      "logps/chosen": -159.0,
      "logps/rejected": -181.0,
      "loss": 0.2379,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.63671875,
      "rewards/margins": 1.9140625,
      "rewards/rejected": -2.546875,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 26.920543277352525,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.78125,
      "logps/chosen": -156.0,
      "logps/rejected": -165.0,
      "loss": 0.235,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.003143310546875,
      "rewards/margins": 2.34375,
      "rewards/rejected": -2.34375,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 16.543384890487943,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -3.75,
      "logits/rejected": -3.78125,
      "logps/chosen": -175.0,
      "logps/rejected": -178.0,
      "loss": 0.284,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.01123046875,
      "rewards/margins": 2.125,
      "rewards/rejected": -2.140625,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 19.580709479779234,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -3.75,
      "logits/rejected": -3.703125,
      "logps/chosen": -148.0,
      "logps/rejected": -154.0,
      "loss": 0.2602,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.2197265625,
      "rewards/margins": 1.78125,
      "rewards/rejected": -2.0,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 44.22442899275912,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.734375,
      "logps/chosen": -159.0,
      "logps/rejected": -177.0,
      "loss": 0.2864,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.2421875,
      "rewards/margins": 2.3125,
      "rewards/rejected": -2.546875,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 24.31645593739508,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.75,
      "logps/chosen": -136.0,
      "logps/rejected": -155.0,
      "loss": 0.2402,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.17578125,
      "rewards/margins": 1.8046875,
      "rewards/rejected": -1.984375,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 19.57297161002994,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -3.71875,
      "logits/rejected": -3.734375,
      "logps/chosen": -141.0,
      "logps/rejected": -166.0,
      "loss": 0.2417,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.37890625,
      "rewards/margins": 2.15625,
      "rewards/rejected": -2.53125,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 25.945863470074855,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.75,
      "logps/chosen": -154.0,
      "logps/rejected": -161.0,
      "loss": 0.2371,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.455078125,
      "rewards/margins": 2.1875,
      "rewards/rejected": -2.640625,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 31.912265169645966,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.75,
      "logps/chosen": -169.0,
      "logps/rejected": -179.0,
      "loss": 0.2455,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.2734375,
      "rewards/margins": 2.34375,
      "rewards/rejected": -2.625,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 38.2641259412187,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -3.84375,
      "logits/rejected": -3.84375,
      "logps/chosen": -159.0,
      "logps/rejected": -147.0,
      "loss": 0.2957,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.40234375,
      "rewards/margins": 1.984375,
      "rewards/rejected": -2.390625,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 123.86824660564758,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.859375,
      "logps/chosen": -165.0,
      "logps/rejected": -157.0,
      "loss": 0.238,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.255859375,
      "rewards/margins": 2.015625,
      "rewards/rejected": -2.265625,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 29.128423677932044,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.71875,
      "logps/chosen": -168.0,
      "logps/rejected": -168.0,
      "loss": 0.2587,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.2412109375,
      "rewards/margins": 2.4375,
      "rewards/rejected": -2.671875,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 71.83483460230082,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.71875,
      "logps/chosen": -158.0,
      "logps/rejected": -160.0,
      "loss": 0.3002,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.671875,
      "rewards/margins": 1.8828125,
      "rewards/rejected": -2.5625,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 49.670427143932685,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.734375,
      "logps/chosen": -163.0,
      "logps/rejected": -163.0,
      "loss": 0.2729,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.1748046875,
      "rewards/margins": 2.4375,
      "rewards/rejected": -2.625,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 19.060615412082008,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.6875,
      "logps/chosen": -158.0,
      "logps/rejected": -166.0,
      "loss": 0.251,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.1796875,
      "rewards/margins": 2.4375,
      "rewards/rejected": -2.609375,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 39.892097574872004,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -3.828125,
      "logits/rejected": -3.765625,
      "logps/chosen": -166.0,
      "logps/rejected": -164.0,
      "loss": 0.2573,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.271484375,
      "rewards/margins": 2.46875,
      "rewards/rejected": -2.734375,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 21.65359611700786,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -3.828125,
      "logits/rejected": -3.828125,
      "logps/chosen": -162.0,
      "logps/rejected": -178.0,
      "loss": 0.2344,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.189453125,
      "rewards/margins": 2.171875,
      "rewards/rejected": -2.359375,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 75.00568195550674,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -3.75,
      "logits/rejected": -3.6875,
      "logps/chosen": -164.0,
      "logps/rejected": -167.0,
      "loss": 0.3566,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.34375,
      "rewards/margins": 2.453125,
      "rewards/rejected": -2.796875,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 44.248430598525125,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.75,
      "logps/chosen": -146.0,
      "logps/rejected": -168.0,
      "loss": 0.2997,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.609375,
      "rewards/margins": 2.375,
      "rewards/rejected": -2.984375,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 19.428223623591357,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.78125,
      "logps/chosen": -177.0,
      "logps/rejected": -179.0,
      "loss": 0.2518,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.59765625,
      "rewards/margins": 2.390625,
      "rewards/rejected": -3.0,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 26.98351153586123,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -3.734375,
      "logits/rejected": -3.703125,
      "logps/chosen": -180.0,
      "logps/rejected": -173.0,
      "loss": 0.2538,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.1943359375,
      "rewards/margins": 2.71875,
      "rewards/rejected": -2.90625,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 39.469369075577596,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.75,
      "logps/chosen": -182.0,
      "logps/rejected": -186.0,
      "loss": 0.2687,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.267578125,
      "rewards/margins": 2.46875,
      "rewards/rejected": -2.734375,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 17.754718906595105,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.703125,
      "logps/chosen": -158.0,
      "logps/rejected": -160.0,
      "loss": 0.234,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.63671875,
      "rewards/margins": 2.5,
      "rewards/rejected": -3.125,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 25.664313503068346,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.84375,
      "logps/chosen": -164.0,
      "logps/rejected": -177.0,
      "loss": 0.2881,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.4140625,
      "rewards/margins": 2.3125,
      "rewards/rejected": -2.734375,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 24.645641316868154,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -3.9375,
      "logits/rejected": -3.890625,
      "logps/chosen": -162.0,
      "logps/rejected": -169.0,
      "loss": 0.2672,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.283203125,
      "rewards/margins": 2.296875,
      "rewards/rejected": -2.59375,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 23.663661235688544,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -3.875,
      "logits/rejected": -3.796875,
      "logps/chosen": -160.0,
      "logps/rejected": -167.0,
      "loss": 0.3127,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.06494140625,
      "rewards/margins": 2.234375,
      "rewards/rejected": -2.296875,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 32.49659843416573,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -3.828125,
      "logits/rejected": -3.828125,
      "logps/chosen": -161.0,
      "logps/rejected": -174.0,
      "loss": 0.2824,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.51171875,
      "rewards/margins": 2.078125,
      "rewards/rejected": -2.59375,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 30.1838550783329,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.78125,
      "logps/chosen": -162.0,
      "logps/rejected": -179.0,
      "loss": 0.3086,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.41796875,
      "rewards/margins": 2.453125,
      "rewards/rejected": -2.875,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 44.78498783548206,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -3.78125,
      "logits/rejected": -3.765625,
      "logps/chosen": -150.0,
      "logps/rejected": -161.0,
      "loss": 0.2827,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.27734375,
      "rewards/margins": 2.53125,
      "rewards/rejected": -2.8125,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -3.921875,
      "eval_logits/rejected": -3.9375,
      "eval_logps/chosen": -160.0,
      "eval_logps/rejected": -152.0,
      "eval_loss": 0.59765625,
      "eval_rewards/accuracies": 0.7142857313156128,
      "eval_rewards/chosen": -0.91015625,
      "eval_rewards/margins": 0.7890625,
      "eval_rewards/rejected": -1.703125,
      "eval_runtime": 12.5642,
      "eval_samples_per_second": 15.918,
      "eval_steps_per_second": 0.557,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 17.2477751746402,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -3.859375,
      "logits/rejected": -3.84375,
      "logps/chosen": -176.0,
      "logps/rejected": -186.0,
      "loss": 0.2298,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.4140625,
      "rewards/margins": 2.65625,
      "rewards/rejected": -3.0625,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 14.384728716810327,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -3.890625,
      "logits/rejected": -3.875,
      "logps/chosen": -146.0,
      "logps/rejected": -169.0,
      "loss": 0.1122,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.0400390625,
      "rewards/margins": 3.15625,
      "rewards/rejected": -3.109375,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 19.29457533448998,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.734375,
      "logps/chosen": -166.0,
      "logps/rejected": -191.0,
      "loss": 0.1055,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.322265625,
      "rewards/margins": 3.375,
      "rewards/rejected": -3.703125,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 24.032579929334997,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.796875,
      "logps/chosen": -172.0,
      "logps/rejected": -185.0,
      "loss": 0.125,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.2138671875,
      "rewards/margins": 3.40625,
      "rewards/rejected": -3.625,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 18.284931179318995,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.828125,
      "logps/chosen": -163.0,
      "logps/rejected": -199.0,
      "loss": 0.1143,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.29296875,
      "rewards/margins": 3.78125,
      "rewards/rejected": -4.0625,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 17.109017415204463,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -3.84375,
      "logits/rejected": -3.796875,
      "logps/chosen": -162.0,
      "logps/rejected": -190.0,
      "loss": 0.116,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.015625,
      "rewards/margins": 3.46875,
      "rewards/rejected": -4.46875,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 14.430120339337325,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.765625,
      "logps/chosen": -158.0,
      "logps/rejected": -176.0,
      "loss": 0.0811,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.69140625,
      "rewards/margins": 3.4375,
      "rewards/rejected": -4.125,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 36.84699798238744,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -3.796875,
      "logits/rejected": -3.828125,
      "logps/chosen": -176.0,
      "logps/rejected": -207.0,
      "loss": 0.1149,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.5703125,
      "rewards/margins": 3.765625,
      "rewards/rejected": -4.3125,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 38.70415298863847,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -3.859375,
      "logits/rejected": -3.8125,
      "logps/chosen": -139.0,
      "logps/rejected": -150.0,
      "loss": 0.1411,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.7265625,
      "rewards/margins": 3.03125,
      "rewards/rejected": -3.75,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 34.52111152142296,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -3.875,
      "logits/rejected": -3.828125,
      "logps/chosen": -172.0,
      "logps/rejected": -174.0,
      "loss": 0.1185,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.73046875,
      "rewards/margins": 3.578125,
      "rewards/rejected": -4.3125,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 24.993624258157183,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -3.921875,
      "logits/rejected": -3.890625,
      "logps/chosen": -174.0,
      "logps/rejected": -203.0,
      "loss": 0.0959,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.373046875,
      "rewards/margins": 4.03125,
      "rewards/rejected": -4.40625,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 56.97777145014865,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -3.875,
      "logits/rejected": -3.828125,
      "logps/chosen": -165.0,
      "logps/rejected": -205.0,
      "loss": 0.1172,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.6953125,
      "rewards/margins": 4.0625,
      "rewards/rejected": -4.75,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 15.424016825011353,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -3.9375,
      "logits/rejected": -3.890625,
      "logps/chosen": -170.0,
      "logps/rejected": -186.0,
      "loss": 0.177,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.251953125,
      "rewards/margins": 4.125,
      "rewards/rejected": -4.375,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 27.60867638158319,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -3.890625,
      "logits/rejected": -3.859375,
      "logps/chosen": -156.0,
      "logps/rejected": -182.0,
      "loss": 0.1404,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.59375,
      "rewards/margins": 3.78125,
      "rewards/rejected": -4.375,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 32.6927091889918,
      "learning_rate": 1e-07,
      "logits/chosen": -3.90625,
      "logits/rejected": -3.890625,
      "logps/chosen": -145.0,
      "logps/rejected": -167.0,
      "loss": 0.1322,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.439453125,
      "rewards/margins": 3.625,
      "rewards/rejected": -4.0625,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 50.32075234217129,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -3.859375,
      "logits/rejected": -3.84375,
      "logps/chosen": -166.0,
      "logps/rejected": -205.0,
      "loss": 0.1506,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.0118408203125,
      "rewards/margins": 4.0,
      "rewards/rejected": -4.0,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 25.726617616913085,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -3.765625,
      "logits/rejected": -3.703125,
      "logps/chosen": -164.0,
      "logps/rejected": -184.0,
      "loss": 0.0632,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.455078125,
      "rewards/margins": 4.25,
      "rewards/rejected": -4.6875,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 59.79014353161108,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -3.84375,
      "logits/rejected": -3.78125,
      "logps/chosen": -176.0,
      "logps/rejected": -197.0,
      "loss": 0.1917,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.51171875,
      "rewards/margins": 4.0625,
      "rewards/rejected": -4.5625,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 20.883232904842796,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -3.859375,
      "logits/rejected": -3.75,
      "logps/chosen": -153.0,
      "logps/rejected": -202.0,
      "loss": 0.1078,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.98046875,
      "rewards/margins": 4.125,
      "rewards/rejected": -5.125,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 33.82809230023863,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -3.84375,
      "logits/rejected": -3.890625,
      "logps/chosen": -166.0,
      "logps/rejected": -185.0,
      "loss": 0.1577,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.330078125,
      "rewards/margins": 3.71875,
      "rewards/rejected": -4.0625,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 27.231276769636622,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -3.875,
      "logits/rejected": -3.8125,
      "logps/chosen": -174.0,
      "logps/rejected": -180.0,
      "loss": 0.1474,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.6171875,
      "rewards/margins": 3.96875,
      "rewards/rejected": -4.59375,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 19.897144457067984,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -3.921875,
      "logits/rejected": -3.90625,
      "logps/chosen": -163.0,
      "logps/rejected": -190.0,
      "loss": 0.1056,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.4609375,
      "rewards/margins": 3.875,
      "rewards/rejected": -4.34375,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 16.832809273135723,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -3.90625,
      "logits/rejected": -3.84375,
      "logps/chosen": -185.0,
      "logps/rejected": -192.0,
      "loss": 0.1505,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.4921875,
      "rewards/margins": 3.890625,
      "rewards/rejected": -4.375,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 25.5826383013988,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -3.875,
      "logits/rejected": -3.84375,
      "logps/chosen": -170.0,
      "logps/rejected": -198.0,
      "loss": 0.189,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.578125,
      "rewards/margins": 3.984375,
      "rewards/rejected": -4.5625,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 7.961103175945574,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -3.984375,
      "logits/rejected": -3.859375,
      "logps/chosen": -155.0,
      "logps/rejected": -199.0,
      "loss": 0.0914,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.6796875,
      "rewards/margins": 3.90625,
      "rewards/rejected": -4.59375,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 43.21277166372785,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -3.9375,
      "logits/rejected": -3.796875,
      "logps/chosen": -144.0,
      "logps/rejected": -177.0,
      "loss": 0.1453,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.71484375,
      "rewards/margins": 3.875,
      "rewards/rejected": -4.59375,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 32.395806124933166,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -3.890625,
      "logits/rejected": -3.875,
      "logps/chosen": -180.0,
      "logps/rejected": -191.0,
      "loss": 0.1552,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.74609375,
      "rewards/margins": 3.84375,
      "rewards/rejected": -4.59375,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 15.20514124558027,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -3.875,
      "logits/rejected": -3.828125,
      "logps/chosen": -164.0,
      "logps/rejected": -183.0,
      "loss": 0.1864,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.62109375,
      "rewards/margins": 3.5625,
      "rewards/rejected": -4.1875,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 21.901769137176068,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -3.8125,
      "logits/rejected": -3.828125,
      "logps/chosen": -175.0,
      "logps/rejected": -191.0,
      "loss": 0.1521,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.7578125,
      "rewards/margins": 3.859375,
      "rewards/rejected": -4.625,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 11.054400660722038,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -3.859375,
      "logits/rejected": -3.765625,
      "logps/chosen": -159.0,
      "logps/rejected": -185.0,
      "loss": 0.1066,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.9296875,
      "rewards/margins": 3.890625,
      "rewards/rejected": -4.8125,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 19.191663313068442,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -3.9375,
      "logits/rejected": -3.890625,
      "logps/chosen": -162.0,
      "logps/rejected": -178.0,
      "loss": 0.1137,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.74609375,
      "rewards/margins": 3.78125,
      "rewards/rejected": -4.53125,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -3.984375,
      "eval_logits/rejected": -4.0,
      "eval_logps/chosen": -168.0,
      "eval_logps/rejected": -161.0,
      "eval_loss": 0.6615625023841858,
      "eval_rewards/accuracies": 0.6785714030265808,
      "eval_rewards/chosen": -1.7109375,
      "eval_rewards/margins": 0.94140625,
      "eval_rewards/rejected": -2.640625,
      "eval_runtime": 15.5574,
      "eval_samples_per_second": 12.856,
      "eval_steps_per_second": 0.45,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.3473652442550253,
      "train_runtime": 4873.7931,
      "train_samples_per_second": 6.155,
      "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
}