{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.992,
  "eval_steps": 500,
  "global_step": 31,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "grad_norm": 40.213084307288156,
      "learning_rate": 4.415111107797445e-07,
      "logits/chosen": -1.3210335969924927,
      "logits/rejected": -1.302152395248413,
      "logps/chosen": -223.4312744140625,
      "logps/rejected": -232.8748321533203,
      "loss": 0.8699,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.016161538660526276,
      "rewards/margins": 0.03614959120750427,
      "rewards/rejected": -0.05231112986803055,
      "step": 10
    },
    {
      "epoch": 0.64,
      "grad_norm": 35.12653259383695,
      "learning_rate": 1.782991918222275e-07,
      "logits/chosen": -1.3020235300064087,
      "logits/rejected": -1.307481050491333,
      "logps/chosen": -261.4952087402344,
      "logps/rejected": -265.2666931152344,
      "loss": 0.7936,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.16055841743946075,
      "rewards/margins": 0.19748952984809875,
      "rewards/rejected": -0.3580479323863983,
      "step": 20
    },
    {
      "epoch": 0.96,
      "grad_norm": 38.8919066173499,
      "learning_rate": 1.690410564514244e-09,
      "logits/chosen": -1.3019598722457886,
      "logits/rejected": -1.3105428218841553,
      "logps/chosen": -253.36685180664062,
      "logps/rejected": -260.6737365722656,
      "loss": 0.7511,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.2009069174528122,
      "rewards/margins": 0.26367321610450745,
      "rewards/rejected": -0.46458011865615845,
      "step": 30
    },
    {
      "epoch": 0.992,
      "step": 31,
      "total_flos": 8638359306240.0,
      "train_loss": 0.8013531161892798,
      "train_runtime": 531.301,
      "train_samples_per_second": 7.529,
      "train_steps_per_second": 0.058
    }
  ],
  "logging_steps": 10,
  "max_steps": 31,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8638359306240.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}