{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.10276172125883108,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "grad_norm": 6.844920635223389,
      "learning_rate": 5e-05,
      "logits/chosen": -3.09375,
      "logits/rejected": -3.0625,
      "logps/chosen": -258.0,
      "logps/rejected": -322.0,
      "loss": 1.0697,
      "max_memory_allocated (GB)": 94.25,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.30078125,
      "rewards/chosen": 0.67578125,
      "rewards/margins": -0.4140625,
      "rewards/rejected": 1.0859375,
      "step": 10,
      "total_memory_available (GB)": 94.62
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.300484657287598,
      "learning_rate": 0.0001,
      "logits/chosen": -3.046875,
      "logits/rejected": -3.03125,
      "logps/chosen": -286.0,
      "logps/rejected": -336.0,
      "loss": 0.5736,
      "max_memory_allocated (GB)": 94.39,
      "memory_allocated (GB)": 51.7,
      "rewards/accuracies": 0.69140625,
      "rewards/chosen": -0.150390625,
      "rewards/margins": 0.8046875,
      "rewards/rejected": -0.95703125,
      "step": 20,
      "total_memory_available (GB)": 94.62
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 10,
  "total_flos": 1.679841703231488e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}