{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.992,
  "eval_steps": 200,
  "global_step": 62,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016,
      "grad_norm": 198.25669918398648,
      "learning_rate": 7.142857142857142e-08,
      "logits/generated": -3.0665292739868164,
      "logits/real": -3.1044487953186035,
      "logps/generated": -260.97100830078125,
      "logps/real": -222.56707763671875,
      "loss": 0.9049,
      "rewards/accuracies": 0.0,
      "rewards/generated": 0.0,
      "rewards/margins": 0.0,
      "rewards/real": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "grad_norm": 191.14682428587082,
      "learning_rate": 4.727272727272727e-07,
      "logits/generated": -2.957974910736084,
      "logits/real": -3.005094051361084,
      "logps/generated": -255.78170776367188,
      "logps/real": -211.8086700439453,
      "loss": 0.7861,
      "rewards/accuracies": 0.6805555820465088,
      "rewards/generated": -1.2839804887771606,
      "rewards/margins": 1.258236289024353,
      "rewards/real": -0.025744358077645302,
      "step": 10
    },
    {
      "epoch": 0.32,
      "grad_norm": 127.1618329581298,
      "learning_rate": 3.818181818181818e-07,
      "logits/generated": -3.0288896560668945,
      "logits/real": -3.092313289642334,
      "logps/generated": -240.3448028564453,
      "logps/real": -194.00479125976562,
      "loss": 0.6416,
      "rewards/accuracies": 0.762499988079071,
      "rewards/generated": -0.07217645645141602,
      "rewards/margins": 1.4789001941680908,
      "rewards/real": 1.4067238569259644,
      "step": 20
    },
    {
      "epoch": 0.48,
      "grad_norm": 158.65956873160914,
      "learning_rate": 2.909090909090909e-07,
      "logits/generated": -2.827226161956787,
      "logits/real": -2.930786609649658,
      "logps/generated": -263.2423400878906,
      "logps/real": -203.39224243164062,
      "loss": 0.5109,
      "rewards/accuracies": 0.824999988079071,
      "rewards/generated": -1.4480739831924438,
      "rewards/margins": 2.398531675338745,
      "rewards/real": 0.9504578709602356,
      "step": 30
    },
    {
      "epoch": 0.64,
      "grad_norm": 160.69468296728675,
      "learning_rate": 2e-07,
      "logits/generated": -2.819359064102173,
      "logits/real": -2.861755847930908,
      "logps/generated": -266.3091735839844,
      "logps/real": -194.5775146484375,
      "loss": 0.5326,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/generated": -2.1509218215942383,
      "rewards/margins": 3.369136095046997,
      "rewards/real": 1.2182139158248901,
      "step": 40
    },
    {
      "epoch": 0.8,
      "grad_norm": 94.40293067720134,
      "learning_rate": 1.0909090909090908e-07,
      "logits/generated": -2.8005573749542236,
      "logits/real": -2.8787262439727783,
      "logps/generated": -261.2699279785156,
      "logps/real": -196.7130889892578,
      "loss": 0.4681,
      "rewards/accuracies": 0.824999988079071,
      "rewards/generated": -2.203646659851074,
      "rewards/margins": 3.2588391304016113,
      "rewards/real": 1.055192232131958,
      "step": 50
    },
    {
      "epoch": 0.96,
      "grad_norm": 75.79186500073192,
      "learning_rate": 1.818181818181818e-08,
      "logits/generated": -2.8064606189727783,
      "logits/real": -2.833669900894165,
      "logps/generated": -264.3329162597656,
      "logps/real": -198.6341094970703,
      "loss": 0.4321,
      "rewards/accuracies": 0.8125,
      "rewards/generated": -2.0550496578216553,
      "rewards/margins": 3.226325511932373,
      "rewards/real": 1.1712758541107178,
      "step": 60
    },
    {
      "epoch": 0.992,
      "step": 62,
      "total_flos": 0.0,
      "train_loss": 0.5572480242098531,
      "train_runtime": 732.2453,
      "train_samples_per_second": 2.731,
      "train_steps_per_second": 0.085
    }
  ],
  "logging_steps": 10,
  "max_steps": 62,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}