{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 80000000000,
  "global_step": 757,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 39.25,
      "learning_rate": 6.578947368421053e-08,
      "logits/chosen": 1.1063932180404663,
      "logits/rejected": 1.991615653038025,
      "logps/chosen": -494.3671569824219,
      "logps/rejected": -1003.34521484375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": 0.0,
      "rewards/student_margin": 0.0,
      "rewards/teacher_margin": 0.0,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 37.5,
      "learning_rate": 6.578947368421053e-07,
      "logits/chosen": 2.408324956893921,
      "logits/rejected": 2.1886119842529297,
      "logps/chosen": -703.4932861328125,
      "logps/rejected": -911.1801147460938,
      "loss": 0.7614,
      "rewards/accuracies": 0.48148149251937866,
      "rewards/chosen": -0.01438586413860321,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -0.03394882380962372,
      "rewards/student_margin": 0.019562965258955956,
      "rewards/teacher_margin": 0.0,
      "step": 10
    },
    {
      "epoch": 0.03,
      "grad_norm": 38.0,
      "learning_rate": 1.3157894736842106e-06,
      "logits/chosen": 2.512688159942627,
      "logits/rejected": 2.2939486503601074,
      "logps/chosen": -730.9832763671875,
      "logps/rejected": -981.3518676757812,
      "loss": 0.6732,
      "rewards/accuracies": 0.5666667222976685,
      "rewards/chosen": 0.09581196308135986,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -0.09433543682098389,
      "rewards/student_margin": 0.19014742970466614,
      "rewards/teacher_margin": 0.0,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 34.5,
      "learning_rate": 1.973684210526316e-06,
      "logits/chosen": 2.7114477157592773,
      "logits/rejected": 2.583392858505249,
      "logps/chosen": -763.4949951171875,
      "logps/rejected": -838.5108642578125,
      "loss": 0.5979,
      "rewards/accuracies": 0.8666666746139526,
      "rewards/chosen": 0.19188283383846283,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -0.49494582414627075,
      "rewards/student_margin": 0.68682861328125,
      "rewards/teacher_margin": 0.0,
      "step": 30
    },
    {
      "epoch": 0.05,
      "grad_norm": 26.875,
      "learning_rate": 2.631578947368421e-06,
      "logits/chosen": 2.125535726547241,
      "logits/rejected": 2.1427719593048096,
      "logps/chosen": -681.7012939453125,
      "logps/rejected": -856.2781982421875,
      "loss": 0.4696,
      "rewards/accuracies": 0.8666666746139526,
      "rewards/chosen": 0.25057679414749146,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -0.9931653738021851,
      "rewards/student_margin": 1.2437422275543213,
      "rewards/teacher_margin": 0.0,
      "step": 40
    },
    {
      "epoch": 0.07,
      "grad_norm": 36.25,
      "learning_rate": 3.289473684210527e-06,
      "logits/chosen": 2.4738311767578125,
      "logits/rejected": 2.2791666984558105,
      "logps/chosen": -677.8941040039062,
      "logps/rejected": -804.3392333984375,
      "loss": 0.4189,
      "rewards/accuracies": 0.8333333730697632,
      "rewards/chosen": 0.4866231381893158,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -1.3408163785934448,
      "rewards/student_margin": 1.8274396657943726,
      "rewards/teacher_margin": 0.0,
      "step": 50
    },
    {
      "epoch": 0.08,
      "grad_norm": 24.75,
      "learning_rate": 3.947368421052632e-06,
      "logits/chosen": 2.610652446746826,
      "logits/rejected": 2.885200262069702,
      "logps/chosen": -713.7745361328125,
      "logps/rejected": -946.8289794921875,
      "loss": 0.3675,
      "rewards/accuracies": 0.8666666150093079,
      "rewards/chosen": 0.2676035761833191,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.754769802093506,
      "rewards/student_margin": 3.0223731994628906,
      "rewards/teacher_margin": 0.0,
      "step": 60
    },
    {
      "epoch": 0.09,
      "grad_norm": 17.125,
      "learning_rate": 4.605263157894737e-06,
      "logits/chosen": 2.8368079662323,
      "logits/rejected": 2.663803815841675,
      "logps/chosen": -791.5745239257812,
      "logps/rejected": -844.3741455078125,
      "loss": 0.3444,
      "rewards/accuracies": 0.8666666150093079,
      "rewards/chosen": 0.23359784483909607,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.148716449737549,
      "rewards/student_margin": 2.3823142051696777,
      "rewards/teacher_margin": 0.0,
      "step": 70
    },
    {
      "epoch": 0.11,
      "grad_norm": 22.625,
      "learning_rate": 4.999574378856833e-06,
      "logits/chosen": 2.178849220275879,
      "logits/rejected": 2.505402088165283,
      "logps/chosen": -694.8255615234375,
      "logps/rejected": -889.0216674804688,
      "loss": 0.3232,
      "rewards/accuracies": 0.8666666150093079,
      "rewards/chosen": 0.25273844599723816,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.432995557785034,
      "rewards/student_margin": 2.685734272003174,
      "rewards/teacher_margin": 0.0,
      "step": 80
    },
    {
      "epoch": 0.12,
      "grad_norm": 24.625,
      "learning_rate": 4.994787805186825e-06,
      "logits/chosen": 2.619114875793457,
      "logits/rejected": 2.5830256938934326,
      "logps/chosen": -722.3959350585938,
      "logps/rejected": -835.6867065429688,
      "loss": 0.3808,
      "rewards/accuracies": 0.7666666507720947,
      "rewards/chosen": 0.12093061208724976,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.4770522117614746,
      "rewards/student_margin": 2.597982883453369,
      "rewards/teacher_margin": 0.0,
      "step": 90
    },
    {
      "epoch": 0.13,
      "grad_norm": 19.75,
      "learning_rate": 4.984692850202173e-06,
      "logits/chosen": 2.8101232051849365,
      "logits/rejected": 2.5987629890441895,
      "logps/chosen": -783.5296630859375,
      "logps/rejected": -776.9186401367188,
      "loss": 0.2936,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.4671626091003418,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -1.996647834777832,
      "rewards/student_margin": 2.4638102054595947,
      "rewards/teacher_margin": 0.0,
      "step": 100
    },
    {
      "epoch": 0.15,
      "grad_norm": 21.0,
      "learning_rate": 4.9693109938340754e-06,
      "logits/chosen": 2.310434341430664,
      "logits/rejected": 2.1604161262512207,
      "logps/chosen": -729.8660888671875,
      "logps/rejected": -837.9103393554688,
      "loss": 0.36,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.6024355292320251,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.125154972076416,
      "rewards/student_margin": 2.727590322494507,
      "rewards/teacher_margin": 0.0,
      "step": 110
    },
    {
      "epoch": 0.16,
      "grad_norm": 27.25,
      "learning_rate": 4.948674965422784e-06,
      "logits/chosen": 2.2550618648529053,
      "logits/rejected": 2.4227263927459717,
      "logps/chosen": -709.0026245117188,
      "logps/rejected": -861.9893798828125,
      "loss": 0.3055,
      "rewards/accuracies": 0.9000000953674316,
      "rewards/chosen": 0.4605357050895691,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -1.956944465637207,
      "rewards/student_margin": 2.417480230331421,
      "rewards/teacher_margin": 0.0,
      "step": 120
    },
    {
      "epoch": 0.17,
      "grad_norm": 22.75,
      "learning_rate": 4.922828674076479e-06,
      "logits/chosen": 2.638427734375,
      "logits/rejected": 2.6139321327209473,
      "logps/chosen": -738.7854614257812,
      "logps/rejected": -812.3598022460938,
      "loss": 0.299,
      "rewards/accuracies": 0.8666666150093079,
      "rewards/chosen": 0.4465925693511963,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.559539318084717,
      "rewards/student_margin": 3.006131887435913,
      "rewards/teacher_margin": 0.0,
      "step": 130
    },
    {
      "epoch": 0.18,
      "grad_norm": 24.75,
      "learning_rate": 4.891827115241967e-06,
      "logits/chosen": 2.491821765899658,
      "logits/rejected": 2.1934759616851807,
      "logps/chosen": -731.97998046875,
      "logps/rejected": -768.702880859375,
      "loss": 0.3217,
      "rewards/accuracies": 0.8666666150093079,
      "rewards/chosen": 1.0505127906799316,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.0951590538024902,
      "rewards/student_margin": 3.145671844482422,
      "rewards/teacher_margin": 0.0,
      "step": 140
    },
    {
      "epoch": 0.2,
      "grad_norm": 18.75,
      "learning_rate": 4.855736253685992e-06,
      "logits/chosen": 2.818260908126831,
      "logits/rejected": 2.7023606300354004,
      "logps/chosen": -855.1539916992188,
      "logps/rejected": -863.5784912109375,
      "loss": 0.3102,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 0.9310780763626099,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.548107624053955,
      "rewards/student_margin": 3.4791858196258545,
      "rewards/teacher_margin": 0.0,
      "step": 150
    },
    {
      "epoch": 0.21,
      "grad_norm": 15.9375,
      "learning_rate": 4.8146328831361466e-06,
      "logits/chosen": 2.5584826469421387,
      "logits/rejected": 2.5176262855529785,
      "logps/chosen": -710.6329345703125,
      "logps/rejected": -765.013916015625,
      "loss": 0.255,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.6767626404762268,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -1.5031782388687134,
      "rewards/student_margin": 2.179940700531006,
      "rewards/teacher_margin": 0.0,
      "step": 160
    },
    {
      "epoch": 0.22,
      "grad_norm": 21.75,
      "learning_rate": 4.768604462880048e-06,
      "logits/chosen": 2.6777164936065674,
      "logits/rejected": 2.462739944458008,
      "logps/chosen": -709.0971069335938,
      "logps/rejected": -819.5133666992188,
      "loss": 0.2703,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.097994327545166,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.144744873046875,
      "rewards/student_margin": 3.24273943901062,
      "rewards/teacher_margin": 0.0,
      "step": 170
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.25,
      "learning_rate": 4.717748931670449e-06,
      "logits/chosen": 2.2394096851348877,
      "logits/rejected": 2.3399100303649902,
      "logps/chosen": -810.650634765625,
      "logps/rejected": -890.9011840820312,
      "loss": 0.2325,
      "rewards/accuracies": 0.9333332777023315,
      "rewards/chosen": 0.620319128036499,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.6976723670959473,
      "rewards/student_margin": 3.3179917335510254,
      "rewards/teacher_margin": 0.0,
      "step": 180
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.8125,
      "learning_rate": 4.662174499332275e-06,
      "logits/chosen": 2.627516269683838,
      "logits/rejected": 2.6304800510406494,
      "logps/chosen": -743.6221923828125,
      "logps/rejected": -846.0401611328125,
      "loss": 0.2124,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.11163077503442764,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.5682520866394043,
      "rewards/student_margin": 2.6798830032348633,
      "rewards/teacher_margin": 0.0,
      "step": 190
    },
    {
      "epoch": 0.26,
      "grad_norm": 16.75,
      "learning_rate": 4.601999416514973e-06,
      "logits/chosen": 2.425690174102783,
      "logits/rejected": 2.57381272315979,
      "logps/chosen": -764.7454833984375,
      "logps/rejected": -895.3714599609375,
      "loss": 0.2577,
      "rewards/accuracies": 0.9000000953674316,
      "rewards/chosen": 1.1225372552871704,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.3252832889556885,
      "rewards/student_margin": 3.4478206634521484,
      "rewards/teacher_margin": 0.0,
      "step": 200
    },
    {
      "epoch": 0.28,
      "grad_norm": 23.375,
      "learning_rate": 4.537351723080117e-06,
      "logits/chosen": 2.468726634979248,
      "logits/rejected": 2.464418888092041,
      "logps/chosen": -720.8771362304688,
      "logps/rejected": -906.501953125,
      "loss": 0.1797,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5130088329315186,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.8423776626586914,
      "rewards/student_margin": 4.355386734008789,
      "rewards/teacher_margin": 0.0,
      "step": 210
    },
    {
      "epoch": 0.29,
      "grad_norm": 19.875,
      "learning_rate": 4.468368975659638e-06,
      "logits/chosen": 2.875859498977661,
      "logits/rejected": 2.759331226348877,
      "logps/chosen": -804.3804321289062,
      "logps/rejected": -869.57958984375,
      "loss": 0.1659,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.109999179840088,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.0473203659057617,
      "rewards/student_margin": 4.157319068908691,
      "rewards/teacher_margin": 0.0,
      "step": 220
    },
    {
      "epoch": 0.3,
      "grad_norm": 24.0,
      "learning_rate": 4.395197954964377e-06,
      "logits/chosen": 2.299351692199707,
      "logits/rejected": 2.2807610034942627,
      "logps/chosen": -668.815673828125,
      "logps/rejected": -830.9869384765625,
      "loss": 0.2658,
      "rewards/accuracies": 0.9333332777023315,
      "rewards/chosen": 1.0605835914611816,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.628436326980591,
      "rewards/student_margin": 3.6890196800231934,
      "rewards/teacher_margin": 0.0,
      "step": 230
    },
    {
      "epoch": 0.32,
      "grad_norm": 17.0,
      "learning_rate": 4.317994353465743e-06,
      "logits/chosen": 2.4613468647003174,
      "logits/rejected": 2.313901424407959,
      "logps/chosen": -764.72119140625,
      "logps/rejected": -858.3145751953125,
      "loss": 0.2166,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.183833360671997,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.823739767074585,
      "rewards/student_margin": 4.007573127746582,
      "rewards/teacher_margin": 0.0,
      "step": 240
    },
    {
      "epoch": 0.33,
      "grad_norm": 13.125,
      "learning_rate": 4.236922444115033e-06,
      "logits/chosen": 2.9010465145111084,
      "logits/rejected": 2.975884199142456,
      "logps/chosen": -766.8567504882812,
      "logps/rejected": -921.1828002929688,
      "loss": 0.1874,
      "rewards/accuracies": 0.8333333730697632,
      "rewards/chosen": 0.8090434074401855,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -1.870889663696289,
      "rewards/student_margin": 2.6799330711364746,
      "rewards/teacher_margin": 0.0,
      "step": 250
    },
    {
      "epoch": 0.34,
      "grad_norm": 28.875,
      "learning_rate": 4.152154730805301e-06,
      "logits/chosen": 2.6264891624450684,
      "logits/rejected": 2.745565891265869,
      "logps/chosen": -748.6024169921875,
      "logps/rejected": -891.1272583007812,
      "loss": 0.2406,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.5763810873031616,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.169581890106201,
      "rewards/student_margin": 3.7459628582000732,
      "rewards/teacher_margin": 0.0,
      "step": 260
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.0625,
      "learning_rate": 4.063871581319526e-06,
      "logits/chosen": 2.5747060775756836,
      "logits/rejected": 2.6211907863616943,
      "logps/chosen": -729.8704833984375,
      "logps/rejected": -893.0794067382812,
      "loss": 0.1921,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 0.9890060424804688,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.523777961730957,
      "rewards/student_margin": 3.5127837657928467,
      "rewards/teacher_margin": 0.0,
      "step": 270
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.875,
      "learning_rate": 3.9722608435460716e-06,
      "logits/chosen": 2.8572492599487305,
      "logits/rejected": 2.787008762359619,
      "logps/chosen": -744.4260864257812,
      "logps/rejected": -903.74560546875,
      "loss": 0.2131,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.8562577962875366,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.471896171569824,
      "rewards/student_margin": 4.32815408706665,
      "rewards/teacher_margin": 0.0,
      "step": 280
    },
    {
      "epoch": 0.38,
      "grad_norm": 18.25,
      "learning_rate": 3.877517445778084e-06,
      "logits/chosen": 2.7553367614746094,
      "logits/rejected": 2.84175705909729,
      "logps/chosen": -851.8234252929688,
      "logps/rejected": -869.7921142578125,
      "loss": 0.1889,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.2772903442382812,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.959285259246826,
      "rewards/student_margin": 4.236575603485107,
      "rewards/teacher_margin": 0.0,
      "step": 290
    },
    {
      "epoch": 0.4,
      "grad_norm": 20.0,
      "learning_rate": 3.779842981947257e-06,
      "logits/chosen": 2.5193684101104736,
      "logits/rejected": 2.8250792026519775,
      "logps/chosen": -777.2213134765625,
      "logps/rejected": -962.1539916992188,
      "loss": 0.1268,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.30393385887146,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.1054177284240723,
      "rewards/student_margin": 4.409350872039795,
      "rewards/teacher_margin": 0.0,
      "step": 300
    },
    {
      "epoch": 0.41,
      "grad_norm": 19.625,
      "learning_rate": 3.679445282674553e-06,
      "logits/chosen": 2.3319785594940186,
      "logits/rejected": 2.200749635696411,
      "logps/chosen": -736.1558837890625,
      "logps/rejected": -828.9473876953125,
      "loss": 0.2129,
      "rewards/accuracies": 0.9333332777023315,
      "rewards/chosen": 1.365607738494873,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.0576624870300293,
      "rewards/student_margin": 4.423270225524902,
      "rewards/teacher_margin": 0.0,
      "step": 310
    },
    {
      "epoch": 0.42,
      "grad_norm": 14.125,
      "learning_rate": 3.5765379730505488e-06,
      "logits/chosen": 2.6987969875335693,
      "logits/rejected": 2.7594046592712402,
      "logps/chosen": -770.1749877929688,
      "logps/rejected": -893.2551879882812,
      "loss": 0.1738,
      "rewards/accuracies": 0.9333332777023315,
      "rewards/chosen": 1.076539158821106,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.4383862018585205,
      "rewards/student_margin": 4.5149245262146,
      "rewards/teacher_margin": 0.0,
      "step": 320
    },
    {
      "epoch": 0.44,
      "grad_norm": 23.5,
      "learning_rate": 3.471340018086381e-06,
      "logits/chosen": 2.5061514377593994,
      "logits/rejected": 2.6357250213623047,
      "logps/chosen": -764.2349243164062,
      "logps/rejected": -1048.420166015625,
      "loss": 0.2055,
      "rewards/accuracies": 0.9666666984558105,
      "rewards/chosen": 1.4593900442123413,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.759369373321533,
      "rewards/student_margin": 4.218760013580322,
      "rewards/teacher_margin": 0.0,
      "step": 330
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.625,
      "learning_rate": 3.364075256802471e-06,
      "logits/chosen": 2.7208733558654785,
      "logits/rejected": 2.7141761779785156,
      "logps/chosen": -805.8480834960938,
      "logps/rejected": -931.7926025390625,
      "loss": 0.2091,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.5951147079467773,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.8993289470672607,
      "rewards/student_margin": 4.494443893432617,
      "rewards/teacher_margin": 0.0,
      "step": 340
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.8125,
      "learning_rate": 3.254971925946383e-06,
      "logits/chosen": 2.7437257766723633,
      "logits/rejected": 2.5028164386749268,
      "logps/chosen": -769.4174194335938,
      "logps/rejected": -856.7975463867188,
      "loss": 0.1375,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.8919599056243896,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.4549193382263184,
      "rewards/student_margin": 4.346879005432129,
      "rewards/teacher_margin": 0.0,
      "step": 350
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.3125,
      "learning_rate": 3.1442621743532516e-06,
      "logits/chosen": 2.987748622894287,
      "logits/rejected": 2.8533196449279785,
      "logps/chosen": -763.906494140625,
      "logps/rejected": -766.51708984375,
      "loss": 0.2198,
      "rewards/accuracies": 0.8666666746139526,
      "rewards/chosen": 1.3865398168563843,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.3393778800964355,
      "rewards/student_margin": 3.725917339324951,
      "rewards/teacher_margin": 0.0,
      "step": 360
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.28125,
      "learning_rate": 3.0321815689821127e-06,
      "logits/chosen": 2.4668517112731934,
      "logits/rejected": 2.6435983180999756,
      "logps/chosen": -738.1652221679688,
      "logps/rejected": -1032.048583984375,
      "loss": 0.1364,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.38614821434021,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.3692269325256348,
      "rewards/student_margin": 4.755375385284424,
      "rewards/teacher_margin": 0.0,
      "step": 370
    },
    {
      "epoch": 0.5,
      "grad_norm": 26.875,
      "learning_rate": 2.9189685936791856e-06,
      "logits/chosen": 2.366330146789551,
      "logits/rejected": 2.6206741333007812,
      "logps/chosen": -717.1395263671875,
      "logps/rejected": -890.14990234375,
      "loss": 0.2086,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.2325283288955688,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.064573049545288,
      "rewards/student_margin": 4.297101020812988,
      "rewards/teacher_margin": 0.0,
      "step": 380
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.75,
      "learning_rate": 2.8048641417346447e-06,
      "logits/chosen": 2.639984130859375,
      "logits/rejected": 2.6091980934143066,
      "logps/chosen": -689.2213134765625,
      "logps/rejected": -898.9273681640625,
      "loss": 0.1774,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.432947039604187,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.152224063873291,
      "rewards/student_margin": 4.585171222686768,
      "rewards/teacher_margin": 0.0,
      "step": 390
    },
    {
      "epoch": 0.53,
      "grad_norm": 17.25,
      "learning_rate": 2.69011100331259e-06,
      "logits/chosen": 2.7592196464538574,
      "logits/rejected": 2.9880573749542236,
      "logps/chosen": -680.7615966796875,
      "logps/rejected": -911.51708984375,
      "loss": 0.1216,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.259261131286621,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.410494565963745,
      "rewards/student_margin": 4.669755458831787,
      "rewards/teacher_margin": 0.0,
      "step": 400
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.15625,
      "learning_rate": 2.574953348844884e-06,
      "logits/chosen": 2.678232192993164,
      "logits/rejected": 2.359405994415283,
      "logps/chosen": -791.8428955078125,
      "logps/rejected": -739.9708251953125,
      "loss": 0.1245,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.0176723003387451,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.4821863174438477,
      "rewards/student_margin": 3.4998583793640137,
      "rewards/teacher_margin": 0.0,
      "step": 410
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.5625,
      "learning_rate": 2.4596362094880617e-06,
      "logits/chosen": 2.144169569015503,
      "logits/rejected": 1.9014570713043213,
      "logps/chosen": -629.1177978515625,
      "logps/rejected": -703.9340209960938,
      "loss": 0.1786,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.782714605331421,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.654998302459717,
      "rewards/student_margin": 4.437712669372559,
      "rewards/teacher_margin": 0.0,
      "step": 420
    },
    {
      "epoch": 0.57,
      "grad_norm": 15.625,
      "learning_rate": 2.344404955748805e-06,
      "logits/chosen": 2.4446656703948975,
      "logits/rejected": 2.495633840560913,
      "logps/chosen": -737.5274047851562,
      "logps/rejected": -865.40771484375,
      "loss": 0.2011,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.6026709079742432,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.152886152267456,
      "rewards/student_margin": 4.755557060241699,
      "rewards/teacher_margin": 0.0,
      "step": 430
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.0625,
      "learning_rate": 2.2295047753873465e-06,
      "logits/chosen": 2.3097705841064453,
      "logits/rejected": 2.4514598846435547,
      "logps/chosen": -707.8262939453125,
      "logps/rejected": -975.5931396484375,
      "loss": 0.0938,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.9523252248764038,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.6663551330566406,
      "rewards/student_margin": 5.618680000305176,
      "rewards/teacher_margin": 0.0,
      "step": 440
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.1875,
      "learning_rate": 2.1151801517097266e-06,
      "logits/chosen": 2.646815776824951,
      "logits/rejected": 2.6606507301330566,
      "logps/chosen": -785.2379150390625,
      "logps/rejected": -882.7388916015625,
      "loss": 0.1304,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.0314725637435913,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.4571354389190674,
      "rewards/student_margin": 4.488607883453369,
      "rewards/teacher_margin": 0.0,
      "step": 450
    },
    {
      "epoch": 0.61,
      "grad_norm": 12.1875,
      "learning_rate": 2.001674343358975e-06,
      "logits/chosen": 2.504572629928589,
      "logits/rejected": 2.6332459449768066,
      "logps/chosen": -711.09375,
      "logps/rejected": -883.10107421875,
      "loss": 0.1243,
      "rewards/accuracies": 0.9333332777023315,
      "rewards/chosen": 1.1015676259994507,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.6316211223602295,
      "rewards/student_margin": 4.733188629150391,
      "rewards/teacher_margin": 0.0,
      "step": 460
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.3125,
      "learning_rate": 1.8892288667121139e-06,
      "logits/chosen": 2.414062738418579,
      "logits/rejected": 2.228271484375,
      "logps/chosen": -739.9596557617188,
      "logps/rejected": -847.4876098632812,
      "loss": 0.1183,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.0411046743392944,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.962963581085205,
      "rewards/student_margin": 4.004067897796631,
      "rewards/teacher_margin": 0.0,
      "step": 470
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.8125,
      "learning_rate": 1.7780829819843458e-06,
      "logits/chosen": 2.6683709621429443,
      "logits/rejected": 2.7936007976531982,
      "logps/chosen": -721.1678466796875,
      "logps/rejected": -834.6106567382812,
      "loss": 0.1543,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.5083043575286865,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.2512154579162598,
      "rewards/student_margin": 4.759519100189209,
      "rewards/teacher_margin": 0.0,
      "step": 480
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.0,
      "learning_rate": 1.668473184133862e-06,
      "logits/chosen": 2.3921468257904053,
      "logits/rejected": 2.376159429550171,
      "logps/chosen": -738.181884765625,
      "logps/rejected": -903.42919921875,
      "loss": 0.1793,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.5232467651367188,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.4825570583343506,
      "rewards/student_margin": 4.00580358505249,
      "rewards/teacher_margin": 0.0,
      "step": 490
    },
    {
      "epoch": 0.66,
      "grad_norm": 23.125,
      "learning_rate": 1.5606326996505478e-06,
      "logits/chosen": 2.4245734214782715,
      "logits/rejected": 2.066218614578247,
      "logps/chosen": -789.9022827148438,
      "logps/rejected": -760.2530517578125,
      "loss": 0.1997,
      "rewards/accuracies": 0.9666666984558105,
      "rewards/chosen": 1.2094615697860718,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.621572256088257,
      "rewards/student_margin": 3.831033706665039,
      "rewards/teacher_margin": 0.0,
      "step": 500
    },
    {
      "epoch": 0.67,
      "grad_norm": 20.625,
      "learning_rate": 1.45479099029929e-06,
      "logits/chosen": 2.312058687210083,
      "logits/rejected": 2.6048178672790527,
      "logps/chosen": -724.6693115234375,
      "logps/rejected": -881.6843872070312,
      "loss": 0.1882,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.326018214225769,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.731388568878174,
      "rewards/student_margin": 4.057406425476074,
      "rewards/teacher_margin": 0.0,
      "step": 510
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.875,
      "learning_rate": 1.3511732648738213e-06,
      "logits/chosen": 2.2725682258605957,
      "logits/rejected": 2.4900736808776855,
      "logps/chosen": -779.0091552734375,
      "logps/rejected": -983.3134765625,
      "loss": 0.1265,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.638450026512146,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.900458812713623,
      "rewards/student_margin": 4.538908958435059,
      "rewards/teacher_margin": 0.0,
      "step": 520
    },
    {
      "epoch": 0.7,
      "grad_norm": 20.0,
      "learning_rate": 1.2500000000000007e-06,
      "logits/chosen": 2.5716114044189453,
      "logits/rejected": 2.5023512840270996,
      "logps/chosen": -705.2183837890625,
      "logps/rejected": -907.6337890625,
      "loss": 0.1224,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.617611289024353,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.396939992904663,
      "rewards/student_margin": 4.014551639556885,
      "rewards/teacher_margin": 0.0,
      "step": 530
    },
    {
      "epoch": 0.71,
      "grad_norm": 20.125,
      "learning_rate": 1.1514864710081295e-06,
      "logits/chosen": 2.1336722373962402,
      "logits/rejected": 2.49604868888855,
      "logps/chosen": -596.4373779296875,
      "logps/rejected": -815.4148559570312,
      "loss": 0.1487,
      "rewards/accuracies": 0.9666666984558105,
      "rewards/chosen": 1.7167946100234985,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.7153236865997314,
      "rewards/student_margin": 4.4321184158325195,
      "rewards/teacher_margin": 0.0,
      "step": 540
    },
    {
      "epoch": 0.73,
      "grad_norm": 14.0,
      "learning_rate": 1.0558422938725404e-06,
      "logits/chosen": 2.081679344177246,
      "logits/rejected": 2.593083620071411,
      "logps/chosen": -605.888427734375,
      "logps/rejected": -937.4814453125,
      "loss": 0.1279,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.3643019199371338,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.542567491531372,
      "rewards/student_margin": 4.906869411468506,
      "rewards/teacher_margin": 0.0,
      "step": 550
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.3125,
      "learning_rate": 9.632709791930977e-07,
      "logits/chosen": 2.3516314029693604,
      "logits/rejected": 2.1533358097076416,
      "logps/chosen": -748.89892578125,
      "logps/rejected": -780.47998046875,
      "loss": 0.2078,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.812962293624878,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.3793187141418457,
      "rewards/student_margin": 4.1922807693481445,
      "rewards/teacher_margin": 0.0,
      "step": 560
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.1875,
      "learning_rate": 8.739694991676328e-07,
      "logits/chosen": 2.1663920879364014,
      "logits/rejected": 2.4739956855773926,
      "logps/chosen": -654.0028076171875,
      "logps/rejected": -846.7198486328125,
      "loss": 0.153,
      "rewards/accuracies": 0.9666666984558105,
      "rewards/chosen": 1.5124599933624268,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.6132845878601074,
      "rewards/student_margin": 4.125744819641113,
      "rewards/teacher_margin": 0.0,
      "step": 570
    },
    {
      "epoch": 0.77,
      "grad_norm": 20.125,
      "learning_rate": 7.881278684767354e-07,
      "logits/chosen": 2.249051809310913,
      "logits/rejected": 2.377701997756958,
      "logps/chosen": -744.0155029296875,
      "logps/rejected": -837.7652587890625,
      "loss": 0.1775,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.166346788406372,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.4967403411865234,
      "rewards/student_margin": 3.6630866527557373,
      "rewards/teacher_margin": 0.0,
      "step": 580
    },
    {
      "epoch": 0.78,
      "grad_norm": 19.5,
      "learning_rate": 7.059287399726558e-07,
      "logits/chosen": 2.0429625511169434,
      "logits/rejected": 2.2974820137023926,
      "logps/chosen": -675.0552368164062,
      "logps/rejected": -815.6538696289062,
      "loss": 0.1259,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.4241259098052979,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.6004066467285156,
      "rewards/student_margin": 4.024532794952393,
      "rewards/teacher_margin": 0.0,
      "step": 590
    },
    {
      "epoch": 0.79,
      "grad_norm": 15.25,
      "learning_rate": 6.275470160326308e-07,
      "logits/chosen": 2.4218051433563232,
      "logits/rejected": 2.626558780670166,
      "logps/chosen": -694.3748168945312,
      "logps/rejected": -825.705078125,
      "loss": 0.1295,
      "rewards/accuracies": 0.8333333134651184,
      "rewards/chosen": 1.1830830574035645,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.594542980194092,
      "rewards/student_margin": 3.7776265144348145,
      "rewards/teacher_margin": 0.0,
      "step": 600
    },
    {
      "epoch": 0.81,
      "grad_norm": 14.0625,
      "learning_rate": 5.531494764035842e-07,
      "logits/chosen": 2.403853416442871,
      "logits/rejected": 2.4846677780151367,
      "logps/chosen": -745.2608642578125,
      "logps/rejected": -816.1573486328125,
      "loss": 0.1549,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.9859026670455933,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.085474729537964,
      "rewards/student_margin": 3.0713775157928467,
      "rewards/teacher_margin": 0.0,
      "step": 610
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.1875,
      "learning_rate": 4.828944233300581e-07,
      "logits/chosen": 2.5432651042938232,
      "logits/rejected": 2.3120391368865967,
      "logps/chosen": -758.4690551757812,
      "logps/rejected": -820.4471435546875,
      "loss": 0.0827,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.7055679559707642,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.691899538040161,
      "rewards/student_margin": 4.397467613220215,
      "rewards/teacher_margin": 0.0,
      "step": 620
    },
    {
      "epoch": 0.83,
      "grad_norm": 18.875,
      "learning_rate": 4.1693134472049254e-07,
      "logits/chosen": 2.4983761310577393,
      "logits/rejected": 2.4922070503234863,
      "logps/chosen": -759.421875,
      "logps/rejected": -880.2047119140625,
      "loss": 0.1595,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.5521160364151,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.921255588531494,
      "rewards/student_margin": 4.473371505737305,
      "rewards/teacher_margin": 0.0,
      "step": 630
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.875,
      "learning_rate": 3.554005960685364e-07,
      "logits/chosen": 2.4154155254364014,
      "logits/rejected": 2.3739418983459473,
      "logps/chosen": -722.2140502929688,
      "logps/rejected": -797.0501708984375,
      "loss": 0.1201,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 2.029052495956421,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.8493266105651855,
      "rewards/student_margin": 4.8783793449401855,
      "rewards/teacher_margin": 0.0,
      "step": 640
    },
    {
      "epoch": 0.86,
      "grad_norm": 28.75,
      "learning_rate": 2.984331018062148e-07,
      "logits/chosen": 2.273947238922119,
      "logits/rejected": 2.346890926361084,
      "logps/chosen": -617.5139770507812,
      "logps/rejected": -829.6735229492188,
      "loss": 0.1981,
      "rewards/accuracies": 0.9000000953674316,
      "rewards/chosen": 1.2198824882507324,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.615978717803955,
      "rewards/student_margin": 3.8358612060546875,
      "rewards/teacher_margin": 0.0,
      "step": 650
    },
    {
      "epoch": 0.87,
      "grad_norm": 3.953125,
      "learning_rate": 2.461500767244049e-07,
      "logits/chosen": 2.328996181488037,
      "logits/rejected": 2.3885445594787598,
      "logps/chosen": -684.5684814453125,
      "logps/rejected": -871.5986328125,
      "loss": 0.1078,
      "rewards/accuracies": 0.9666666984558105,
      "rewards/chosen": 1.9273674488067627,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.5571177005767822,
      "rewards/student_margin": 4.484485149383545,
      "rewards/teacher_margin": 0.0,
      "step": 660
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.9375,
      "learning_rate": 1.9866276805336733e-07,
      "logits/chosen": 2.427218198776245,
      "logits/rejected": 2.4423775672912598,
      "logps/chosen": -671.7244262695312,
      "logps/rejected": -919.7501220703125,
      "loss": 0.1608,
      "rewards/accuracies": 0.8666666746139526,
      "rewards/chosen": 1.4714218378067017,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.4355459213256836,
      "rewards/student_margin": 3.9069676399230957,
      "rewards/teacher_margin": 0.0,
      "step": 670
    },
    {
      "epoch": 0.9,
      "grad_norm": 13.375,
      "learning_rate": 1.560722187521496e-07,
      "logits/chosen": 2.2423176765441895,
      "logits/rejected": 1.9140275716781616,
      "logps/chosen": -724.7633056640625,
      "logps/rejected": -754.9696655273438,
      "loss": 0.1739,
      "rewards/accuracies": 0.9666666984558105,
      "rewards/chosen": 1.4883036613464355,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.9442994594573975,
      "rewards/student_margin": 4.432602882385254,
      "rewards/teacher_margin": 0.0,
      "step": 680
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.5625,
      "learning_rate": 1.1846905251051894e-07,
      "logits/chosen": 2.494938850402832,
      "logits/rejected": 2.532592296600342,
      "logps/chosen": -734.0770263671875,
      "logps/rejected": -898.4133911132812,
      "loss": 0.156,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.3606359958648682,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.778499126434326,
      "rewards/student_margin": 4.139134883880615,
      "rewards/teacher_margin": 0.0,
      "step": 690
    },
    {
      "epoch": 0.92,
      "grad_norm": 19.125,
      "learning_rate": 8.593328092089737e-08,
      "logits/chosen": 2.027740001678467,
      "logits/rejected": 1.8771311044692993,
      "logps/chosen": -752.6876831054688,
      "logps/rejected": -762.588134765625,
      "loss": 0.1369,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.0343955755233765,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.3003334999084473,
      "rewards/student_margin": 3.3347294330596924,
      "rewards/teacher_margin": 0.0,
      "step": 700
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.125,
      "learning_rate": 5.8534133230599974e-08,
      "logits/chosen": 2.5199155807495117,
      "logits/rejected": 2.8889317512512207,
      "logps/chosen": -701.169921875,
      "logps/rejected": -910.6904296875,
      "loss": 0.1635,
      "rewards/accuracies": 0.966666579246521,
      "rewards/chosen": 1.4893516302108765,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.9409499168395996,
      "rewards/student_margin": 4.430301189422607,
      "rewards/teacher_margin": 0.0,
      "step": 710
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.5,
      "learning_rate": 3.632990903662398e-08,
      "logits/chosen": 2.757976770401001,
      "logits/rejected": 2.7887349128723145,
      "logps/chosen": -732.608642578125,
      "logps/rejected": -873.4423828125,
      "loss": 0.1751,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.6327383518218994,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -3.2143402099609375,
      "rewards/student_margin": 4.847078800201416,
      "rewards/teacher_margin": 0.0,
      "step": 720
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.875,
      "learning_rate": 1.9367854236425065e-08,
      "logits/chosen": 2.469646453857422,
      "logits/rejected": 2.3219096660614014,
      "logps/chosen": -724.3168334960938,
      "logps/rejected": -875.6312255859375,
      "loss": 0.1225,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 1.6782375574111938,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.8410890102386475,
      "rewards/student_margin": 4.519326686859131,
      "rewards/teacher_margin": 0.0,
      "step": 730
    },
    {
      "epoch": 0.98,
      "grad_norm": 3.140625,
      "learning_rate": 7.684060498630575e-09,
      "logits/chosen": 2.5574631690979004,
      "logits/rejected": 2.5168144702911377,
      "logps/chosen": -690.5845336914062,
      "logps/rejected": -850.1302490234375,
      "loss": 0.1444,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 2.156329393386841,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.7730555534362793,
      "rewards/student_margin": 4.929385185241699,
      "rewards/teacher_margin": 0.0,
      "step": 740
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.34375,
      "learning_rate": 1.3033884675933695e-09,
      "logits/chosen": 2.7296438217163086,
      "logits/rejected": 2.305525779724121,
      "logps/chosen": -714.5927734375,
      "logps/rejected": -803.5744018554688,
      "loss": 0.1035,
      "rewards/accuracies": 0.9333333969116211,
      "rewards/chosen": 1.8840978145599365,
      "rewards/diff": 0.0,
      "rewards/diff_abs": 0.0,
      "rewards/rejected": -2.6910955905914307,
      "rewards/student_margin": 4.575193405151367,
      "rewards/teacher_margin": 0.0,
      "step": 750
    },
    {
      "epoch": 1.0,
      "step": 757,
      "total_flos": 0.0,
      "train_loss": 0.2174788669361936,
      "train_runtime": 882.7004,
      "train_samples_per_second": 20.572,
      "train_steps_per_second": 0.858
    }
  ],
  "logging_steps": 10,
  "max_steps": 757,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "total_flos": 0.0,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}