{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12310217480508823,
  "eval_steps": 50,
  "global_step": 150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008206811653672548,
      "grad_norm": 0.06318386644124985,
      "learning_rate": 4.999451708687114e-06,
      "logits/chosen": -2.1367907524108887,
      "logits/rejected": -2.4948182106018066,
      "logps/chosen": -0.291498601436615,
      "logps/rejected": -0.3196522295475006,
      "loss": 7.5728,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.4372479021549225,
      "rewards/margins": 0.04223042353987694,
      "rewards/rejected": -0.47947829961776733,
      "step": 10
    },
    {
      "epoch": 0.016413623307345096,
      "grad_norm": 0.07310314476490021,
      "learning_rate": 4.997807075247147e-06,
      "logits/chosen": -2.1456007957458496,
      "logits/rejected": -2.4455342292785645,
      "logps/chosen": -0.26213544607162476,
      "logps/rejected": -0.32332050800323486,
      "loss": 7.5298,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.3932031989097595,
      "rewards/margins": 0.09177760779857635,
      "rewards/rejected": -0.4849807620048523,
      "step": 20
    },
    {
      "epoch": 0.024620434961017644,
      "grad_norm": 0.05936102196574211,
      "learning_rate": 4.9950668210706795e-06,
      "logits/chosen": -2.0765950679779053,
      "logits/rejected": -2.485799789428711,
      "logps/chosen": -0.26631081104278564,
      "logps/rejected": -0.32647624611854553,
      "loss": 7.5208,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.39946624636650085,
      "rewards/margins": 0.09024813771247864,
      "rewards/rejected": -0.4897143840789795,
      "step": 30
    },
    {
      "epoch": 0.03282724661469019,
      "grad_norm": 0.08499134331941605,
      "learning_rate": 4.9912321481237616e-06,
      "logits/chosen": -2.0753884315490723,
      "logits/rejected": -2.441580295562744,
      "logps/chosen": -0.2749950885772705,
      "logps/rejected": -0.30180150270462036,
      "loss": 7.4229,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.41249266266822815,
      "rewards/margins": 0.04020959883928299,
      "rewards/rejected": -0.45270222425460815,
      "step": 40
    },
    {
      "epoch": 0.04103405826836274,
      "grad_norm": 0.07681389898061752,
      "learning_rate": 4.986304738420684e-06,
      "logits/chosen": -2.145660877227783,
      "logits/rejected": -2.465946912765503,
      "logps/chosen": -0.24909739196300507,
      "logps/rejected": -0.2796121835708618,
      "loss": 7.4811,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.373646080493927,
      "rewards/margins": 0.045772187411785126,
      "rewards/rejected": -0.41941824555397034,
      "step": 50
    },
    {
      "epoch": 0.04103405826836274,
      "eval_logits/chosen": -2.012000799179077,
      "eval_logits/rejected": -2.5381252765655518,
      "eval_logps/chosen": -0.24157460033893585,
      "eval_logps/rejected": -0.2957758605480194,
      "eval_loss": 0.9317650198936462,
      "eval_rewards/accuracies": 0.5252525210380554,
      "eval_rewards/chosen": -0.3623619079589844,
      "eval_rewards/margins": 0.08130191266536713,
      "eval_rewards/rejected": -0.4436637759208679,
      "eval_runtime": 26.0809,
      "eval_samples_per_second": 30.214,
      "eval_steps_per_second": 3.796,
      "step": 50
    },
    {
      "epoch": 0.04924086992203529,
      "grad_norm": 0.06638535112142563,
      "learning_rate": 4.980286753286196e-06,
      "logits/chosen": -2.145846128463745,
      "logits/rejected": -2.4077115058898926,
      "logps/chosen": -0.22265203297138214,
      "logps/rejected": -0.30774614214897156,
      "loss": 7.4605,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.3339780271053314,
      "rewards/margins": 0.1276412308216095,
      "rewards/rejected": -0.4616192877292633,
      "step": 60
    },
    {
      "epoch": 0.057447681575707836,
      "grad_norm": 0.057281140238046646,
      "learning_rate": 4.973180832407471e-06,
      "logits/chosen": -2.0021350383758545,
      "logits/rejected": -2.4299912452697754,
      "logps/chosen": -0.23488977551460266,
      "logps/rejected": -0.33270469307899475,
      "loss": 7.4257,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.3523346781730652,
      "rewards/margins": 0.14672236144542694,
      "rewards/rejected": -0.4990570545196533,
      "step": 70
    },
    {
      "epoch": 0.06565449322938038,
      "grad_norm": 0.07725922018289566,
      "learning_rate": 4.964990092676263e-06,
      "logits/chosen": -2.117995023727417,
      "logits/rejected": -2.359265089035034,
      "logps/chosen": -0.21598832309246063,
      "logps/rejected": -0.300583153963089,
      "loss": 7.4384,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.32398244738578796,
      "rewards/margins": 0.12689228355884552,
      "rewards/rejected": -0.4508747458457947,
      "step": 80
    },
    {
      "epoch": 0.07386130488305294,
      "grad_norm": 0.0598183274269104,
      "learning_rate": 4.9557181268217225e-06,
      "logits/chosen": -2.282627582550049,
      "logits/rejected": -2.441333532333374,
      "logps/chosen": -0.23655852675437927,
      "logps/rejected": -0.3246815800666809,
      "loss": 7.4584,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.3548378050327301,
      "rewards/margins": 0.13218457996845245,
      "rewards/rejected": -0.48702239990234375,
      "step": 90
    },
    {
      "epoch": 0.08206811653672548,
      "grad_norm": 0.058213479816913605,
      "learning_rate": 4.9453690018345144e-06,
      "logits/chosen": -2.1114468574523926,
      "logits/rejected": -2.5035691261291504,
      "logps/chosen": -0.23073866963386536,
      "logps/rejected": -0.29445192217826843,
      "loss": 7.4116,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.34610801935195923,
      "rewards/margins": 0.09556989371776581,
      "rewards/rejected": -0.44167789816856384,
      "step": 100
    },
    {
      "epoch": 0.08206811653672548,
      "eval_logits/chosen": -2.0183491706848145,
      "eval_logits/rejected": -2.5400593280792236,
      "eval_logps/chosen": -0.20393377542495728,
      "eval_logps/rejected": -0.2818409502506256,
      "eval_loss": 0.9129964113235474,
      "eval_rewards/accuracies": 0.5656565427780151,
      "eval_rewards/chosen": -0.3059006631374359,
      "eval_rewards/margins": 0.11686072498559952,
      "eval_rewards/rejected": -0.4227614104747772,
      "eval_runtime": 26.0825,
      "eval_samples_per_second": 30.212,
      "eval_steps_per_second": 3.796,
      "step": 100
    },
    {
      "epoch": 0.09027492819039803,
      "grad_norm": 0.06249881908297539,
      "learning_rate": 4.933947257182901e-06,
      "logits/chosen": -2.1324548721313477,
      "logits/rejected": -2.434319019317627,
      "logps/chosen": -0.22180762887001038,
      "logps/rejected": -0.28862181305885315,
      "loss": 7.3604,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.332711398601532,
      "rewards/margins": 0.10022131353616714,
      "rewards/rejected": -0.43293270468711853,
      "step": 110
    },
    {
      "epoch": 0.09848173984407058,
      "grad_norm": 0.061758093535900116,
      "learning_rate": 4.921457902821578e-06,
      "logits/chosen": -2.0597169399261475,
      "logits/rejected": -2.4386391639709473,
      "logps/chosen": -0.22720107436180115,
      "logps/rejected": -0.303659051656723,
      "loss": 7.3624,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.3408016264438629,
      "rewards/margins": 0.1146869882941246,
      "rewards/rejected": -0.4554885923862457,
      "step": 120
    },
    {
      "epoch": 0.10668855149774313,
      "grad_norm": 0.08368540555238724,
      "learning_rate": 4.907906416994146e-06,
      "logits/chosen": -2.0944437980651855,
      "logits/rejected": -2.4157254695892334,
      "logps/chosen": -0.19590887427330017,
      "logps/rejected": -0.3365771770477295,
      "loss": 7.3464,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.29386329650878906,
      "rewards/margins": 0.2110024392604828,
      "rewards/rejected": -0.5048657655715942,
      "step": 130
    },
    {
      "epoch": 0.11489536315141567,
      "grad_norm": 0.060954928398132324,
      "learning_rate": 4.893298743830168e-06,
      "logits/chosen": -2.1551766395568848,
      "logits/rejected": -2.5695576667785645,
      "logps/chosen": -0.19875812530517578,
      "logps/rejected": -0.2967599928379059,
      "loss": 7.3179,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.29813718795776367,
      "rewards/margins": 0.14700281620025635,
      "rewards/rejected": -0.4451400339603424,
      "step": 140
    },
    {
      "epoch": 0.12310217480508823,
      "grad_norm": 0.05665091797709465,
      "learning_rate": 4.8776412907378845e-06,
      "logits/chosen": -2.1410365104675293,
      "logits/rejected": -2.4798667430877686,
      "logps/chosen": -0.19316771626472473,
      "logps/rejected": -0.2972142696380615,
      "loss": 7.2384,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.2897515296936035,
      "rewards/margins": 0.15606984496116638,
      "rewards/rejected": -0.4458213746547699,
      "step": 150
    },
    {
      "epoch": 0.12310217480508823,
      "eval_logits/chosen": -2.0654072761535645,
      "eval_logits/rejected": -2.596571207046509,
      "eval_logps/chosen": -0.17970335483551025,
      "eval_logps/rejected": -0.2767573893070221,
      "eval_loss": 0.8982937335968018,
      "eval_rewards/accuracies": 0.5858585834503174,
      "eval_rewards/chosen": -0.269555002450943,
      "eval_rewards/margins": 0.14558106660842896,
      "eval_rewards/rejected": -0.41513609886169434,
      "eval_runtime": 26.0741,
      "eval_samples_per_second": 30.222,
      "eval_steps_per_second": 3.797,
      "step": 150
    }
  ],
  "logging_steps": 10,
  "max_steps": 1500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.5967069864617574e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}