{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.08206811653672548,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008206811653672548,
      "grad_norm": 0.07778492569923401,
      "learning_rate": 4.999451708687114e-06,
      "logits/chosen": -2.053281307220459,
      "logits/rejected": -2.495474338531494,
      "logps/chosen": -0.3126755356788635,
      "logps/rejected": -0.3312620520591736,
      "loss": 7.6211,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -0.4690132737159729,
      "rewards/margins": 0.027879873290657997,
      "rewards/rejected": -0.49689316749572754,
      "step": 10
    },
    {
      "epoch": 0.016413623307345096,
      "grad_norm": 0.07773654907941818,
      "learning_rate": 4.997807075247147e-06,
      "logits/chosen": -2.0624098777770996,
      "logits/rejected": -2.4424185752868652,
      "logps/chosen": -0.26926660537719727,
      "logps/rejected": -0.2978014051914215,
      "loss": 7.5195,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.4038998484611511,
      "rewards/margins": 0.04280223697423935,
      "rewards/rejected": -0.44670209288597107,
      "step": 20
    },
    {
      "epoch": 0.024620434961017644,
      "grad_norm": 0.07357177883386612,
      "learning_rate": 4.9950668210706795e-06,
      "logits/chosen": -2.068427562713623,
      "logits/rejected": -2.486642360687256,
      "logps/chosen": -0.29993391036987305,
      "logps/rejected": -0.34360918402671814,
      "loss": 7.4913,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.4499008059501648,
      "rewards/margins": 0.06551288068294525,
      "rewards/rejected": -0.5154137015342712,
      "step": 30
    },
    {
      "epoch": 0.03282724661469019,
      "grad_norm": 0.14212799072265625,
      "learning_rate": 4.9912321481237616e-06,
      "logits/chosen": -2.015650987625122,
      "logits/rejected": -2.3838727474212646,
      "logps/chosen": -0.2911723852157593,
      "logps/rejected": -0.30521970987319946,
      "loss": 7.5217,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.4367586076259613,
      "rewards/margins": 0.021070968359708786,
      "rewards/rejected": -0.4578295648097992,
      "step": 40
    },
    {
      "epoch": 0.04103405826836274,
      "grad_norm": 0.08107248693704605,
      "learning_rate": 4.986304738420684e-06,
      "logits/chosen": -2.1150989532470703,
      "logits/rejected": -2.4338631629943848,
      "logps/chosen": -0.26249754428863525,
      "logps/rejected": -0.3132360577583313,
      "loss": 7.519,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.3937462866306305,
      "rewards/margins": 0.07610772550106049,
      "rewards/rejected": -0.4698540270328522,
      "step": 50
    },
    {
      "epoch": 0.04103405826836274,
      "eval_logits/chosen": -2.0232737064361572,
      "eval_logits/rejected": -2.4952735900878906,
      "eval_logps/chosen": -0.27974528074264526,
      "eval_logps/rejected": -0.3420677185058594,
      "eval_loss": 0.9291417598724365,
      "eval_rewards/accuracies": 0.49494948983192444,
      "eval_rewards/chosen": -0.41961798071861267,
      "eval_rewards/margins": 0.09348361939191818,
      "eval_rewards/rejected": -0.5131015777587891,
      "eval_runtime": 26.0563,
      "eval_samples_per_second": 30.242,
      "eval_steps_per_second": 3.799,
      "step": 50
    },
    {
      "epoch": 0.04924086992203529,
      "grad_norm": 0.06815352290868759,
      "learning_rate": 4.980286753286196e-06,
      "logits/chosen": -1.9890680313110352,
      "logits/rejected": -2.3848204612731934,
      "logps/chosen": -0.26213228702545166,
      "logps/rejected": -0.31342557072639465,
      "loss": 7.432,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": -0.3931984603404999,
      "rewards/margins": 0.0769399031996727,
      "rewards/rejected": -0.4701383709907532,
      "step": 60
    },
    {
      "epoch": 0.057447681575707836,
      "grad_norm": 0.06748568266630173,
      "learning_rate": 4.973180832407471e-06,
      "logits/chosen": -2.070542812347412,
      "logits/rejected": -2.3977038860321045,
      "logps/chosen": -0.24570491909980774,
      "logps/rejected": -0.3655605912208557,
      "loss": 7.35,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.3685573935508728,
      "rewards/margins": 0.17978355288505554,
      "rewards/rejected": -0.548340916633606,
      "step": 70
    },
    {
      "epoch": 0.06565449322938038,
      "grad_norm": 0.10909309983253479,
      "learning_rate": 4.964990092676263e-06,
      "logits/chosen": -2.2012317180633545,
      "logits/rejected": -2.346029758453369,
      "logps/chosen": -0.2279246598482132,
      "logps/rejected": -0.35396742820739746,
      "loss": 7.5082,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.341886967420578,
      "rewards/margins": 0.18906418979167938,
      "rewards/rejected": -0.5309511423110962,
      "step": 80
    },
    {
      "epoch": 0.07386130488305294,
      "grad_norm": 0.05977805703878403,
      "learning_rate": 4.9557181268217225e-06,
      "logits/chosen": -2.0719449520111084,
      "logits/rejected": -2.4491190910339355,
      "logps/chosen": -0.2503294348716736,
      "logps/rejected": -0.29939892888069153,
      "loss": 7.5129,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.37549418210983276,
      "rewards/margins": 0.07360419631004333,
      "rewards/rejected": -0.4490983486175537,
      "step": 90
    },
    {
      "epoch": 0.08206811653672548,
      "grad_norm": 0.051751479506492615,
      "learning_rate": 4.9453690018345144e-06,
      "logits/chosen": -2.0634045600891113,
      "logits/rejected": -2.458428382873535,
      "logps/chosen": -0.24033495783805847,
      "logps/rejected": -0.29080909490585327,
      "loss": 7.4432,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.3605024516582489,
      "rewards/margins": 0.07571124285459518,
      "rewards/rejected": -0.4362136721611023,
      "step": 100
    },
    {
      "epoch": 0.08206811653672548,
      "eval_logits/chosen": -2.0207154750823975,
      "eval_logits/rejected": -2.486215353012085,
      "eval_logps/chosen": -0.2376101016998291,
      "eval_logps/rejected": -0.32593628764152527,
      "eval_loss": 0.9085211753845215,
      "eval_rewards/accuracies": 0.5353535413742065,
      "eval_rewards/chosen": -0.35641518235206604,
      "eval_rewards/margins": 0.13248924911022186,
      "eval_rewards/rejected": -0.4889043867588043,
      "eval_runtime": 26.0119,
      "eval_samples_per_second": 30.294,
      "eval_steps_per_second": 3.806,
      "step": 100
    }
  ],
  "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": 3.646585074352128e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}