{
  "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": 37.97070444635821,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -0.171875,
      "logits/rejected": -0.267578125,
      "logps/chosen": -270.0,
      "logps/rejected": -274.0,
      "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": 38.19010125711672,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -0.2578125,
      "logits/rejected": -0.19921875,
      "logps/chosen": -258.0,
      "logps/rejected": -268.0,
      "loss": 0.6935,
      "rewards/accuracies": 0.1388888955116272,
      "rewards/chosen": 0.00555419921875,
      "rewards/margins": -0.006927490234375,
      "rewards/rejected": 0.01251220703125,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 37.93417440403385,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -0.30859375,
      "logits/rejected": -0.2578125,
      "logps/chosen": -254.0,
      "logps/rejected": -264.0,
      "loss": 0.695,
      "rewards/accuracies": 0.25,
      "rewards/chosen": 0.00750732421875,
      "rewards/margins": 0.010009765625,
      "rewards/rejected": -0.0024871826171875,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 35.49552063923879,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -0.181640625,
      "logits/rejected": -0.203125,
      "logps/chosen": -256.0,
      "logps/rejected": -270.0,
      "loss": 0.687,
      "rewards/accuracies": 0.25,
      "rewards/chosen": -0.018798828125,
      "rewards/margins": -0.01251220703125,
      "rewards/rejected": -0.0062255859375,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 37.8998353526773,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -0.2412109375,
      "logits/rejected": -0.1923828125,
      "logps/chosen": -260.0,
      "logps/rejected": -268.0,
      "loss": 0.6617,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.00628662109375,
      "rewards/margins": 0.08740234375,
      "rewards/rejected": -0.08154296875,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 34.053751730157884,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -0.236328125,
      "logits/rejected": -0.2138671875,
      "logps/chosen": -262.0,
      "logps/rejected": -270.0,
      "loss": 0.6186,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.0037994384765625,
      "rewards/margins": 0.1728515625,
      "rewards/rejected": -0.1767578125,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 35.07641844235428,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -0.15234375,
      "logits/rejected": -0.25,
      "logps/chosen": -264.0,
      "logps/rejected": -270.0,
      "loss": 0.6035,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.08740234375,
      "rewards/margins": 0.2236328125,
      "rewards/rejected": -0.3125,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 31.268209648722863,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -0.2265625,
      "logits/rejected": -0.2451171875,
      "logps/chosen": -260.0,
      "logps/rejected": -270.0,
      "loss": 0.557,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.11865234375,
      "rewards/margins": 0.390625,
      "rewards/rejected": -0.51171875,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 37.74932690765184,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -0.23828125,
      "logits/rejected": -0.279296875,
      "logps/chosen": -264.0,
      "logps/rejected": -276.0,
      "loss": 0.5534,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.173828125,
      "rewards/margins": 0.6015625,
      "rewards/rejected": -0.77734375,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 34.120567112746684,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -0.341796875,
      "logits/rejected": -0.322265625,
      "logps/chosen": -260.0,
      "logps/rejected": -276.0,
      "loss": 0.5572,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.318359375,
      "rewards/margins": 0.62109375,
      "rewards/rejected": -0.94140625,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 36.951252465221565,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -0.25390625,
      "logits/rejected": -0.1923828125,
      "logps/chosen": -264.0,
      "logps/rejected": -280.0,
      "loss": 0.4991,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.27734375,
      "rewards/margins": 0.734375,
      "rewards/rejected": -1.015625,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 30.02019512528484,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -0.28515625,
      "logits/rejected": -0.1953125,
      "logps/chosen": -262.0,
      "logps/rejected": -276.0,
      "loss": 0.4851,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.3125,
      "rewards/margins": 0.765625,
      "rewards/rejected": -1.078125,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 22.52525022285337,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -0.27734375,
      "logits/rejected": -0.25390625,
      "logps/chosen": -262.0,
      "logps/rejected": -282.0,
      "loss": 0.4861,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.419921875,
      "rewards/margins": 0.80859375,
      "rewards/rejected": -1.2265625,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 26.82147658530774,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -0.240234375,
      "logits/rejected": -0.26953125,
      "logps/chosen": -264.0,
      "logps/rejected": -278.0,
      "loss": 0.4855,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.37109375,
      "rewards/margins": 0.83984375,
      "rewards/rejected": -1.2109375,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 30.662238045996084,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -0.326171875,
      "logits/rejected": -0.296875,
      "logps/chosen": -264.0,
      "logps/rejected": -278.0,
      "loss": 0.4498,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.4765625,
      "rewards/margins": 0.86328125,
      "rewards/rejected": -1.34375,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 28.04279789030741,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -0.259765625,
      "logits/rejected": -0.2158203125,
      "logps/chosen": -256.0,
      "logps/rejected": -284.0,
      "loss": 0.3996,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.1923828125,
      "rewards/margins": 1.2109375,
      "rewards/rejected": -1.40625,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 27.30924134093724,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -0.306640625,
      "logits/rejected": -0.291015625,
      "logps/chosen": -266.0,
      "logps/rejected": -284.0,
      "loss": 0.4284,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.33984375,
      "rewards/margins": 1.2109375,
      "rewards/rejected": -1.546875,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 29.128184986153148,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -0.271484375,
      "logits/rejected": -0.267578125,
      "logps/chosen": -258.0,
      "logps/rejected": -282.0,
      "loss": 0.4041,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.10888671875,
      "rewards/margins": 1.3515625,
      "rewards/rejected": -1.4609375,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 28.44378515275308,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -0.248046875,
      "logits/rejected": -0.232421875,
      "logps/chosen": -264.0,
      "logps/rejected": -280.0,
      "loss": 0.3746,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.2734375,
      "rewards/margins": 1.234375,
      "rewards/rejected": -1.5,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 25.06305758889149,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -0.361328125,
      "logits/rejected": -0.3125,
      "logps/chosen": -260.0,
      "logps/rejected": -280.0,
      "loss": 0.4098,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.2392578125,
      "rewards/margins": 1.4609375,
      "rewards/rejected": -1.703125,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 28.318857151618978,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -0.27734375,
      "logits/rejected": -0.265625,
      "logps/chosen": -264.0,
      "logps/rejected": -290.0,
      "loss": 0.3828,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.44921875,
      "rewards/margins": 1.7265625,
      "rewards/rejected": -2.171875,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 21.974930794627994,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -0.314453125,
      "logits/rejected": -0.296875,
      "logps/chosen": -262.0,
      "logps/rejected": -284.0,
      "loss": 0.3576,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.482421875,
      "rewards/margins": 1.4140625,
      "rewards/rejected": -1.8984375,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 28.53909353728823,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -0.3515625,
      "logits/rejected": -0.2734375,
      "logps/chosen": -262.0,
      "logps/rejected": -282.0,
      "loss": 0.3804,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.27734375,
      "rewards/margins": 1.4921875,
      "rewards/rejected": -1.7734375,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 29.065795336502333,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -0.26171875,
      "logits/rejected": -0.2314453125,
      "logps/chosen": -260.0,
      "logps/rejected": -290.0,
      "loss": 0.3794,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.251953125,
      "rewards/margins": 1.921875,
      "rewards/rejected": -2.171875,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 22.505513504251805,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -0.3671875,
      "logits/rejected": -0.240234375,
      "logps/chosen": -254.0,
      "logps/rejected": -286.0,
      "loss": 0.3013,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.25390625,
      "rewards/margins": 1.7734375,
      "rewards/rejected": -2.03125,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 26.058632092459096,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -0.263671875,
      "logits/rejected": -0.337890625,
      "logps/chosen": -258.0,
      "logps/rejected": -290.0,
      "loss": 0.3555,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.111328125,
      "rewards/margins": 1.9140625,
      "rewards/rejected": -2.015625,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 19.76966492550985,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -0.333984375,
      "logits/rejected": -0.33203125,
      "logps/chosen": -260.0,
      "logps/rejected": -298.0,
      "loss": 0.31,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.28125,
      "rewards/margins": 2.484375,
      "rewards/rejected": -2.765625,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 22.68574773253727,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -0.37890625,
      "logits/rejected": -0.228515625,
      "logps/chosen": -262.0,
      "logps/rejected": -288.0,
      "loss": 0.3331,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.224609375,
      "rewards/margins": 2.109375,
      "rewards/rejected": -2.328125,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 22.56048974682512,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -0.349609375,
      "logits/rejected": -0.33203125,
      "logps/chosen": -262.0,
      "logps/rejected": -286.0,
      "loss": 0.3189,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.0296630859375,
      "rewards/margins": 2.359375,
      "rewards/rejected": -2.390625,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 24.089544696213395,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -0.306640625,
      "logits/rejected": -0.2470703125,
      "logps/chosen": -258.0,
      "logps/rejected": -288.0,
      "loss": 0.3271,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.1337890625,
      "rewards/margins": 1.953125,
      "rewards/rejected": -2.078125,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 35.14348640246449,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -0.330078125,
      "logits/rejected": -0.31640625,
      "logps/chosen": -256.0,
      "logps/rejected": -284.0,
      "loss": 0.3653,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.03125,
      "rewards/margins": 1.890625,
      "rewards/rejected": -1.8671875,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 28.52144192599701,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -0.45703125,
      "logits/rejected": -0.388671875,
      "logps/chosen": -258.0,
      "logps/rejected": -280.0,
      "loss": 0.2972,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.150390625,
      "rewards/margins": 1.5,
      "rewards/rejected": -1.6484375,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -0.224609375,
      "eval_logits/rejected": -0.2490234375,
      "eval_logps/chosen": -262.0,
      "eval_logps/rejected": -294.0,
      "eval_loss": 0.3118359446525574,
      "eval_rewards/accuracies": 0.9107142686843872,
      "eval_rewards/chosen": -0.376953125,
      "eval_rewards/margins": 2.359375,
      "eval_rewards/rejected": -2.734375,
      "eval_runtime": 13.5725,
      "eval_samples_per_second": 14.736,
      "eval_steps_per_second": 0.516,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 20.969230534238307,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -0.29296875,
      "logits/rejected": -0.212890625,
      "logps/chosen": -266.0,
      "logps/rejected": -294.0,
      "loss": 0.2392,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.26171875,
      "rewards/margins": 2.515625,
      "rewards/rejected": -2.78125,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 19.963953384570143,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -0.251953125,
      "logits/rejected": -0.3359375,
      "logps/chosen": -258.0,
      "logps/rejected": -296.0,
      "loss": 0.195,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.134765625,
      "rewards/margins": 2.84375,
      "rewards/rejected": -2.71875,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 20.744405566064483,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -0.3203125,
      "logits/rejected": -0.27734375,
      "logps/chosen": -262.0,
      "logps/rejected": -296.0,
      "loss": 0.2379,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.455078125,
      "rewards/margins": 2.703125,
      "rewards/rejected": -3.15625,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 26.49374323177928,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -0.431640625,
      "logits/rejected": -0.310546875,
      "logps/chosen": -258.0,
      "logps/rejected": -296.0,
      "loss": 0.221,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.10302734375,
      "rewards/margins": 2.875,
      "rewards/rejected": -2.96875,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 21.298179894225086,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -0.3046875,
      "logits/rejected": -0.25390625,
      "logps/chosen": -258.0,
      "logps/rejected": -294.0,
      "loss": 0.1896,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.0078125,
      "rewards/margins": 2.671875,
      "rewards/rejected": -2.6875,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 22.234568787601347,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -0.34765625,
      "logits/rejected": -0.1787109375,
      "logps/chosen": -260.0,
      "logps/rejected": -302.0,
      "loss": 0.1964,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.10400390625,
      "rewards/margins": 2.9375,
      "rewards/rejected": -3.03125,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 18.422273644465452,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -0.3203125,
      "logits/rejected": -0.279296875,
      "logps/chosen": -262.0,
      "logps/rejected": -296.0,
      "loss": 0.2115,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.4765625,
      "rewards/margins": 2.671875,
      "rewards/rejected": -3.15625,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 18.209351064280423,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -0.3828125,
      "logits/rejected": -0.302734375,
      "logps/chosen": -252.0,
      "logps/rejected": -292.0,
      "loss": 0.1728,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.640625,
      "rewards/margins": 3.46875,
      "rewards/rejected": -2.828125,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 18.701508173742003,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -0.390625,
      "logits/rejected": -0.29296875,
      "logps/chosen": -251.0,
      "logps/rejected": -290.0,
      "loss": 0.1735,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.330078125,
      "rewards/margins": 2.96875,
      "rewards/rejected": -2.640625,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 18.373395580017903,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -0.384765625,
      "logits/rejected": -0.251953125,
      "logps/chosen": -246.0,
      "logps/rejected": -296.0,
      "loss": 0.1801,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.6171875,
      "rewards/margins": 3.578125,
      "rewards/rejected": -2.96875,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 20.918990517503882,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -0.35546875,
      "logits/rejected": -0.318359375,
      "logps/chosen": -246.0,
      "logps/rejected": -296.0,
      "loss": 0.179,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.55859375,
      "rewards/margins": 3.5625,
      "rewards/rejected": -3.015625,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 17.736489327903804,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -0.349609375,
      "logits/rejected": -0.34765625,
      "logps/chosen": -254.0,
      "logps/rejected": -298.0,
      "loss": 0.1672,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.296875,
      "rewards/margins": 3.515625,
      "rewards/rejected": -3.21875,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 21.942953930447477,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -0.4296875,
      "logits/rejected": -0.353515625,
      "logps/chosen": -252.0,
      "logps/rejected": -298.0,
      "loss": 0.1695,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.44921875,
      "rewards/margins": 3.8125,
      "rewards/rejected": -3.375,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 18.777594197384282,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -0.419921875,
      "logits/rejected": -0.318359375,
      "logps/chosen": -251.0,
      "logps/rejected": -294.0,
      "loss": 0.1777,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.75390625,
      "rewards/margins": 3.75,
      "rewards/rejected": -3.0,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 13.244886312432403,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -0.388671875,
      "logits/rejected": -0.251953125,
      "logps/chosen": -260.0,
      "logps/rejected": -298.0,
      "loss": 0.1566,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.10009765625,
      "rewards/margins": 3.359375,
      "rewards/rejected": -3.25,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 20.776438996599815,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -0.390625,
      "logits/rejected": -0.296875,
      "logps/chosen": -268.0,
      "logps/rejected": -300.0,
      "loss": 0.1738,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.373046875,
      "rewards/margins": 3.03125,
      "rewards/rejected": -3.390625,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 17.15229353424137,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -0.388671875,
      "logits/rejected": -0.375,
      "logps/chosen": -256.0,
      "logps/rejected": -300.0,
      "loss": 0.1766,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.1455078125,
      "rewards/margins": 3.296875,
      "rewards/rejected": -3.15625,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 23.401464498603804,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -0.42578125,
      "logits/rejected": -0.392578125,
      "logps/chosen": -262.0,
      "logps/rejected": -300.0,
      "loss": 0.1758,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.0263671875,
      "rewards/margins": 3.125,
      "rewards/rejected": -3.15625,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 17.822544198091993,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -0.51171875,
      "logits/rejected": -0.419921875,
      "logps/chosen": -260.0,
      "logps/rejected": -300.0,
      "loss": 0.1461,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.2265625,
      "rewards/margins": 3.15625,
      "rewards/rejected": -3.375,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 22.056494041131923,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -0.353515625,
      "logits/rejected": -0.337890625,
      "logps/chosen": -260.0,
      "logps/rejected": -296.0,
      "loss": 0.1813,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.00531005859375,
      "rewards/margins": 3.34375,
      "rewards/rejected": -3.34375,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 18.765876965767607,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -0.375,
      "logits/rejected": -0.2578125,
      "logps/chosen": -258.0,
      "logps/rejected": -302.0,
      "loss": 0.1407,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.1318359375,
      "rewards/margins": 3.796875,
      "rewards/rejected": -3.671875,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 26.180047813644787,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -0.38671875,
      "logits/rejected": -0.283203125,
      "logps/chosen": -255.0,
      "logps/rejected": -296.0,
      "loss": 0.1826,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.0400390625,
      "rewards/margins": 3.25,
      "rewards/rejected": -3.21875,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 18.339809316233776,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -0.3984375,
      "logits/rejected": -0.345703125,
      "logps/chosen": -252.0,
      "logps/rejected": -296.0,
      "loss": 0.1419,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.25,
      "rewards/margins": 3.140625,
      "rewards/rejected": -2.890625,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 14.290075285340114,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -0.427734375,
      "logits/rejected": -0.4765625,
      "logps/chosen": -262.0,
      "logps/rejected": -302.0,
      "loss": 0.1795,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.029052734375,
      "rewards/margins": 3.390625,
      "rewards/rejected": -3.359375,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 15.034355235881119,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -0.384765625,
      "logits/rejected": -0.34765625,
      "logps/chosen": -262.0,
      "logps/rejected": -300.0,
      "loss": 0.1462,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.2001953125,
      "rewards/margins": 3.21875,
      "rewards/rejected": -3.421875,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 19.39500388104877,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -0.38671875,
      "logits/rejected": -0.345703125,
      "logps/chosen": -258.0,
      "logps/rejected": -292.0,
      "loss": 0.1925,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.36328125,
      "rewards/margins": 3.046875,
      "rewards/rejected": -2.6875,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 32.15415516079912,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -0.43359375,
      "logits/rejected": -0.33984375,
      "logps/chosen": -253.0,
      "logps/rejected": -300.0,
      "loss": 0.173,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.2353515625,
      "rewards/margins": 3.3125,
      "rewards/rejected": -3.078125,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 17.02529847359552,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -0.37890625,
      "logits/rejected": -0.380859375,
      "logps/chosen": -258.0,
      "logps/rejected": -304.0,
      "loss": 0.1349,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.3671875,
      "rewards/margins": 3.734375,
      "rewards/rejected": -3.359375,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 13.67342268666293,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -0.4609375,
      "logits/rejected": -0.390625,
      "logps/chosen": -253.0,
      "logps/rejected": -294.0,
      "loss": 0.1815,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.47265625,
      "rewards/margins": 3.640625,
      "rewards/rejected": -3.171875,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 9.486908754976506,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -0.337890625,
      "logits/rejected": -0.40234375,
      "logps/chosen": -266.0,
      "logps/rejected": -302.0,
      "loss": 0.1411,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.32421875,
      "rewards/margins": 3.28125,
      "rewards/rejected": -3.609375,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 14.691533824169545,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -0.51171875,
      "logits/rejected": -0.384765625,
      "logps/chosen": -256.0,
      "logps/rejected": -308.0,
      "loss": 0.1474,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.154296875,
      "rewards/margins": 3.90625,
      "rewards/rejected": -4.0625,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -0.296875,
      "eval_logits/rejected": -0.296875,
      "eval_logps/chosen": -256.0,
      "eval_logps/rejected": -302.0,
      "eval_loss": 0.22533203661441803,
      "eval_rewards/accuracies": 0.8928571343421936,
      "eval_rewards/chosen": 0.1845703125,
      "eval_rewards/margins": 3.75,
      "eval_rewards/rejected": -3.5625,
      "eval_runtime": 13.5854,
      "eval_samples_per_second": 14.722,
      "eval_steps_per_second": 0.515,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 10.933412977760229,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -0.349609375,
      "logits/rejected": -0.22265625,
      "logps/chosen": -262.0,
      "logps/rejected": -310.0,
      "loss": 0.1131,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.24609375,
      "rewards/margins": 3.921875,
      "rewards/rejected": -4.1875,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 7.246474855356494,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -0.4453125,
      "logits/rejected": -0.40234375,
      "logps/chosen": -252.0,
      "logps/rejected": -306.0,
      "loss": 0.0727,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.5859375,
      "rewards/margins": 4.53125,
      "rewards/rejected": -3.953125,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 11.200725866913762,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -0.4453125,
      "logits/rejected": -0.36328125,
      "logps/chosen": -256.0,
      "logps/rejected": -306.0,
      "loss": 0.0838,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.09423828125,
      "rewards/margins": 3.875,
      "rewards/rejected": -3.96875,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 10.183607328693935,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -0.453125,
      "logits/rejected": -0.296875,
      "logps/chosen": -260.0,
      "logps/rejected": -306.0,
      "loss": 0.11,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.2177734375,
      "rewards/margins": 3.765625,
      "rewards/rejected": -3.984375,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 7.486855040157854,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -0.375,
      "logits/rejected": -0.267578125,
      "logps/chosen": -264.0,
      "logps/rejected": -310.0,
      "loss": 0.0734,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.0712890625,
      "rewards/margins": 4.0625,
      "rewards/rejected": -4.125,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 12.645563891293305,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -0.375,
      "logits/rejected": -0.341796875,
      "logps/chosen": -250.0,
      "logps/rejected": -296.0,
      "loss": 0.0719,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.54296875,
      "rewards/margins": 4.1875,
      "rewards/rejected": -3.640625,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 12.238826672018135,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -0.421875,
      "logits/rejected": -0.328125,
      "logps/chosen": -258.0,
      "logps/rejected": -312.0,
      "loss": 0.0678,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.05810546875,
      "rewards/margins": 4.1875,
      "rewards/rejected": -4.25,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 6.180147320034269,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -0.46484375,
      "logits/rejected": -0.365234375,
      "logps/chosen": -264.0,
      "logps/rejected": -314.0,
      "loss": 0.0899,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.38671875,
      "rewards/margins": 3.953125,
      "rewards/rejected": -4.34375,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 8.726100221119632,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -0.412109375,
      "logits/rejected": -0.267578125,
      "logps/chosen": -258.0,
      "logps/rejected": -308.0,
      "loss": 0.0755,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.44921875,
      "rewards/margins": 3.921875,
      "rewards/rejected": -4.375,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 9.722161213558335,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -0.4140625,
      "logits/rejected": -0.333984375,
      "logps/chosen": -252.0,
      "logps/rejected": -306.0,
      "loss": 0.0791,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.287109375,
      "rewards/margins": 4.625,
      "rewards/rejected": -4.3125,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 17.068386311303975,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -0.48046875,
      "logits/rejected": -0.3203125,
      "logps/chosen": -252.0,
      "logps/rejected": -308.0,
      "loss": 0.0929,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.248046875,
      "rewards/margins": 4.4375,
      "rewards/rejected": -4.1875,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 6.896798736592823,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -0.423828125,
      "logits/rejected": -0.2890625,
      "logps/chosen": -255.0,
      "logps/rejected": -310.0,
      "loss": 0.0822,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.275390625,
      "rewards/margins": 4.53125,
      "rewards/rejected": -4.25,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 2.833897847819323,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -0.4296875,
      "logits/rejected": -0.390625,
      "logps/chosen": -254.0,
      "logps/rejected": -308.0,
      "loss": 0.0617,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.318359375,
      "rewards/margins": 5.125,
      "rewards/rejected": -4.78125,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 10.16078078484658,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -0.404296875,
      "logits/rejected": -0.37109375,
      "logps/chosen": -264.0,
      "logps/rejected": -308.0,
      "loss": 0.0805,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.11474609375,
      "rewards/margins": 4.28125,
      "rewards/rejected": -4.40625,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 19.81860765827931,
      "learning_rate": 1e-07,
      "logits/chosen": -0.47265625,
      "logits/rejected": -0.349609375,
      "logps/chosen": -253.0,
      "logps/rejected": -314.0,
      "loss": 0.0708,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.408203125,
      "rewards/margins": 4.75,
      "rewards/rejected": -4.34375,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 11.372196600797032,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -0.443359375,
      "logits/rejected": -0.392578125,
      "logps/chosen": -256.0,
      "logps/rejected": -312.0,
      "loss": 0.0754,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.392578125,
      "rewards/margins": 4.84375,
      "rewards/rejected": -4.4375,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 14.599761926485138,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -0.470703125,
      "logits/rejected": -0.4453125,
      "logps/chosen": -253.0,
      "logps/rejected": -320.0,
      "loss": 0.0839,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.236328125,
      "rewards/margins": 5.5625,
      "rewards/rejected": -5.3125,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 12.22787699921472,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -0.4609375,
      "logits/rejected": -0.27734375,
      "logps/chosen": -266.0,
      "logps/rejected": -318.0,
      "loss": 0.0813,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.416015625,
      "rewards/margins": 4.28125,
      "rewards/rejected": -4.6875,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 14.388262008568441,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -0.482421875,
      "logits/rejected": -0.3359375,
      "logps/chosen": -245.0,
      "logps/rejected": -306.0,
      "loss": 0.0699,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.8515625,
      "rewards/margins": 4.9375,
      "rewards/rejected": -4.09375,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 6.077591657503602,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -0.3828125,
      "logits/rejected": -0.384765625,
      "logps/chosen": -260.0,
      "logps/rejected": -318.0,
      "loss": 0.0878,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.050537109375,
      "rewards/margins": 5.09375,
      "rewards/rejected": -5.03125,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 16.251661055271924,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -0.3984375,
      "logits/rejected": -0.33203125,
      "logps/chosen": -260.0,
      "logps/rejected": -316.0,
      "loss": 0.0657,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.365234375,
      "rewards/margins": 5.375,
      "rewards/rejected": -5.03125,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 15.630444293807114,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -0.46875,
      "logits/rejected": -0.3984375,
      "logps/chosen": -258.0,
      "logps/rejected": -316.0,
      "loss": 0.0814,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.478515625,
      "rewards/margins": 4.84375,
      "rewards/rejected": -4.34375,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 13.755671589382665,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -0.49609375,
      "logits/rejected": -0.37109375,
      "logps/chosen": -258.0,
      "logps/rejected": -310.0,
      "loss": 0.0794,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -0.1494140625,
      "rewards/margins": 3.90625,
      "rewards/rejected": -4.0625,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 8.011897998800759,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -0.37109375,
      "logits/rejected": -0.333984375,
      "logps/chosen": -258.0,
      "logps/rejected": -310.0,
      "loss": 0.072,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.310546875,
      "rewards/margins": 4.6875,
      "rewards/rejected": -4.375,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 19.550720618587036,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -0.5234375,
      "logits/rejected": -0.38671875,
      "logps/chosen": -254.0,
      "logps/rejected": -308.0,
      "loss": 0.0663,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.494140625,
      "rewards/margins": 4.875,
      "rewards/rejected": -4.375,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 16.019939323270826,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -0.484375,
      "logits/rejected": -0.45703125,
      "logps/chosen": -260.0,
      "logps/rejected": -312.0,
      "loss": 0.0868,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.38671875,
      "rewards/margins": 4.53125,
      "rewards/rejected": -4.90625,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 11.168391698626577,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -0.3671875,
      "logits/rejected": -0.2890625,
      "logps/chosen": -250.0,
      "logps/rejected": -316.0,
      "loss": 0.0799,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.478515625,
      "rewards/margins": 5.25,
      "rewards/rejected": -4.78125,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 7.149272204979997,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -0.54296875,
      "logits/rejected": -0.37890625,
      "logps/chosen": -254.0,
      "logps/rejected": -306.0,
      "loss": 0.0712,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.490234375,
      "rewards/margins": 4.84375,
      "rewards/rejected": -4.34375,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 23.608238784265787,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -0.310546875,
      "logits/rejected": -0.2578125,
      "logps/chosen": -258.0,
      "logps/rejected": -310.0,
      "loss": 0.0894,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.201171875,
      "rewards/margins": 4.40625,
      "rewards/rejected": -4.21875,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 7.420883092053745,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -0.4765625,
      "logits/rejected": -0.296875,
      "logps/chosen": -251.0,
      "logps/rejected": -310.0,
      "loss": 0.0616,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.828125,
      "rewards/margins": 5.5,
      "rewards/rejected": -4.6875,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 28.670086263134607,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -0.51171875,
      "logits/rejected": -0.390625,
      "logps/chosen": -245.0,
      "logps/rejected": -308.0,
      "loss": 0.0729,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.94140625,
      "rewards/margins": 5.15625,
      "rewards/rejected": -4.1875,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -0.318359375,
      "eval_logits/rejected": -0.30859375,
      "eval_logps/chosen": -255.0,
      "eval_logps/rejected": -306.0,
      "eval_loss": 0.22841796278953552,
      "eval_rewards/accuracies": 0.8928571343421936,
      "eval_rewards/chosen": 0.384765625,
      "eval_rewards/margins": 4.34375,
      "eval_rewards/rejected": -3.953125,
      "eval_runtime": 13.784,
      "eval_samples_per_second": 14.51,
      "eval_steps_per_second": 0.508,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.23579829137800093,
      "train_runtime": 5562.5353,
      "train_samples_per_second": 5.391,
      "train_steps_per_second": 0.169
    }
  ],
  "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
}