{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9158878504672896,
  "eval_steps": 50,
  "global_step": 78,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 57.02260636434872,
      "learning_rate": 5e-07,
      "logits/chosen": -2.726503372192383,
      "logits/rejected": -2.7297565937042236,
      "logps/chosen": -246.4221649169922,
      "logps/rejected": -233.8533172607422,
      "loss": 0.6919,
      "rewards/accuracies": 0.2874999940395355,
      "rewards/chosen": 0.01179208792746067,
      "rewards/margins": 0.0005273699061945081,
      "rewards/rejected": 0.011264720931649208,
      "step": 5
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 46.98073487199951,
      "learning_rate": 1e-06,
      "logits/chosen": -2.6742746829986572,
      "logits/rejected": -2.6755404472351074,
      "logps/chosen": -254.0735321044922,
      "logps/rejected": -213.7585906982422,
      "loss": 0.6511,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.4740002751350403,
      "rewards/margins": 0.21410906314849854,
      "rewards/rejected": 0.25989121198654175,
      "step": 10
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 42.555481546926174,
      "learning_rate": 9.867190271803463e-07,
      "logits/chosen": -2.536228895187378,
      "logits/rejected": -2.540681838989258,
      "logps/chosen": -239.7574462890625,
      "logps/rejected": -213.6060791015625,
      "loss": 0.6036,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.3793785572052002,
      "rewards/margins": 0.558377206325531,
      "rewards/rejected": 0.8210013508796692,
      "step": 15
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 54.32236085820561,
      "learning_rate": 9.475816456775312e-07,
      "logits/chosen": -2.464298963546753,
      "logits/rejected": -2.4543182849884033,
      "logps/chosen": -253.5270538330078,
      "logps/rejected": -187.60568237304688,
      "loss": 0.6198,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.651815414428711,
      "rewards/margins": 1.00961172580719,
      "rewards/rejected": 0.6422036290168762,
      "step": 20
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 60.5416226517713,
      "learning_rate": 8.846669854914395e-07,
      "logits/chosen": -2.3736038208007812,
      "logits/rejected": -2.3714096546173096,
      "logps/chosen": -237.95022583007812,
      "logps/rejected": -209.83383178710938,
      "loss": 0.6654,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.2289361953735352,
      "rewards/margins": 0.9275712966918945,
      "rewards/rejected": 0.30136504769325256,
      "step": 25
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 19.1123319245986,
      "learning_rate": 8.013173181896282e-07,
      "logits/chosen": -2.379423141479492,
      "logits/rejected": -2.363684892654419,
      "logps/chosen": -256.1286926269531,
      "logps/rejected": -219.66744995117188,
      "loss": 0.3766,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 2.1505281925201416,
      "rewards/margins": 2.1115310192108154,
      "rewards/rejected": 0.0389973521232605,
      "step": 30
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 17.202654912868127,
      "learning_rate": 7.019605024359474e-07,
      "logits/chosen": -2.384429454803467,
      "logits/rejected": -2.3521230220794678,
      "logps/chosen": -237.07943725585938,
      "logps/rejected": -216.90951538085938,
      "loss": 0.2353,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.1183865070343018,
      "rewards/margins": 2.790973424911499,
      "rewards/rejected": -0.6725870370864868,
      "step": 35
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 20.863387078156986,
      "learning_rate": 5.918747589082852e-07,
      "logits/chosen": -2.4055588245391846,
      "logits/rejected": -2.3711154460906982,
      "logps/chosen": -236.8618927001953,
      "logps/rejected": -210.20046997070312,
      "loss": 0.2308,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.564582109451294,
      "rewards/margins": 2.503974199295044,
      "rewards/rejected": -0.9393919706344604,
      "step": 40
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 19.065414773504852,
      "learning_rate": 4.769082706771303e-07,
      "logits/chosen": -2.3858325481414795,
      "logits/rejected": -2.362159252166748,
      "logps/chosen": -242.7309112548828,
      "logps/rejected": -240.0617218017578,
      "loss": 0.2041,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.287132740020752,
      "rewards/margins": 3.6341636180877686,
      "rewards/rejected": -1.3470308780670166,
      "step": 45
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 16.23221592573633,
      "learning_rate": 3.6316850496395855e-07,
      "logits/chosen": -2.3492136001586914,
      "logits/rejected": -2.354576587677002,
      "logps/chosen": -222.81600952148438,
      "logps/rejected": -216.98867797851562,
      "loss": 0.219,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.7240461111068726,
      "rewards/margins": 3.0606493949890137,
      "rewards/rejected": -1.336603045463562,
      "step": 50
    },
    {
      "epoch": 1.8691588785046729,
      "eval_logits/chosen": -2.394383430480957,
      "eval_logits/rejected": -2.345257520675659,
      "eval_logps/chosen": -233.8548126220703,
      "eval_logps/rejected": -205.1439971923828,
      "eval_loss": 0.6225951313972473,
      "eval_rewards/accuracies": 0.7395833134651184,
      "eval_rewards/chosen": 1.014007329940796,
      "eval_rewards/margins": 1.1295721530914307,
      "eval_rewards/rejected": -0.11556478589773178,
      "eval_runtime": 51.6397,
      "eval_samples_per_second": 14.717,
      "eval_steps_per_second": 0.232,
      "step": 50
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 23.0307089839972,
      "learning_rate": 2.566977607165719e-07,
      "logits/chosen": -2.359964370727539,
      "logits/rejected": -2.3656833171844482,
      "logps/chosen": -237.4097137451172,
      "logps/rejected": -227.85302734375,
      "loss": 0.2486,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.7794231176376343,
      "rewards/margins": 3.446876049041748,
      "rewards/rejected": -1.6674525737762451,
      "step": 55
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 14.01114954383873,
      "learning_rate": 1.631521781767214e-07,
      "logits/chosen": -2.347280263900757,
      "logits/rejected": -2.336236000061035,
      "logps/chosen": -222.9915008544922,
      "logps/rejected": -242.89633178710938,
      "loss": 0.1282,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.146198034286499,
      "rewards/margins": 3.6815218925476074,
      "rewards/rejected": -1.5353240966796875,
      "step": 60
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 10.13439083687081,
      "learning_rate": 8.75012627008489e-08,
      "logits/chosen": -2.3028604984283447,
      "logits/rejected": -2.2970237731933594,
      "logps/chosen": -230.75216674804688,
      "logps/rejected": -218.7707061767578,
      "loss": 0.1274,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.042752504348755,
      "rewards/margins": 3.5585684776306152,
      "rewards/rejected": -1.5158164501190186,
      "step": 65
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 14.37288741381732,
      "learning_rate": 3.376388529782215e-08,
      "logits/chosen": -2.2832436561584473,
      "logits/rejected": -2.289004325866699,
      "logps/chosen": -238.990966796875,
      "logps/rejected": -225.0963134765625,
      "loss": 0.1198,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 2.3161845207214355,
      "rewards/margins": 3.950852632522583,
      "rewards/rejected": -1.6346677541732788,
      "step": 70
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 10.412412508913112,
      "learning_rate": 4.794784562397458e-09,
      "logits/chosen": -2.3055803775787354,
      "logits/rejected": -2.287616729736328,
      "logps/chosen": -224.54293823242188,
      "logps/rejected": -259.7159423828125,
      "loss": 0.1229,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 2.4232189655303955,
      "rewards/margins": 4.429986000061035,
      "rewards/rejected": -2.0067672729492188,
      "step": 75
    },
    {
      "epoch": 2.9158878504672896,
      "step": 78,
      "total_flos": 919378820333568.0,
      "train_loss": 0.34083553231679475,
      "train_runtime": 2719.7499,
      "train_samples_per_second": 7.543,
      "train_steps_per_second": 0.029
    }
  ],
  "logging_steps": 5,
  "max_steps": 78,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 919378820333568.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}