{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9230769230769231,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 324.0,
      "learning_rate": 2.8032770122273748e-05,
      "logits/chosen": -4.008616924285889,
      "logits/rejected": -3.9074325561523438,
      "logps/chosen": -6.578675746917725,
      "logps/rejected": -6.709109306335449,
      "loss": 25.12,
      "rewards/accuracies": 0.515625,
      "rewards/chosen": -0.014975759200751781,
      "rewards/margins": 0.00037566880928352475,
      "rewards/rejected": -0.015351429581642151,
      "step": 5
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 326.0,
      "learning_rate": 6.307373277511592e-05,
      "logits/chosen": -3.9267916679382324,
      "logits/rejected": -3.806602954864502,
      "logps/chosen": -6.459668159484863,
      "logps/rejected": -7.284135341644287,
      "loss": 21.4875,
      "rewards/accuracies": 0.621874988079071,
      "rewards/chosen": -0.009770961478352547,
      "rewards/margins": 0.058142371475696564,
      "rewards/rejected": -0.06791333854198456,
      "step": 10
    },
    {
      "epoch": 0.9230769230769231,
      "eval_logits/chosen": -3.4977667331695557,
      "eval_logits/rejected": -3.4567360877990723,
      "eval_logps/chosen": -7.072432041168213,
      "eval_logps/rejected": -7.649731636047363,
      "eval_loss": 22.01285171508789,
      "eval_rewards/accuracies": 0.6650000214576721,
      "eval_rewards/chosen": -0.05798153579235077,
      "eval_rewards/margins": 0.05125926434993744,
      "eval_rewards/rejected": -0.10924080014228821,
      "eval_runtime": 2.3364,
      "eval_samples_per_second": 85.6,
      "eval_steps_per_second": 85.6,
      "step": 12
    }
  ],
  "logging_steps": 5,
  "max_steps": 65,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}