{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9936305732484076,
  "eval_steps": 500,
  "global_step": 39,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025477707006369428,
      "grad_norm": 11.445121344316327,
      "learning_rate": 1.25e-07,
      "logits/chosen": -2.9487781524658203,
      "logits/rejected": -2.858518123626709,
      "logps/chosen": -304.58709716796875,
      "logps/pi_response": -66.34120178222656,
      "logps/ref_response": -66.34120178222656,
      "logps/rejected": -133.11962890625,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.25477707006369427,
      "grad_norm": 9.326893955932823,
      "learning_rate": 4.646121984004665e-07,
      "logits/chosen": -2.7599236965179443,
      "logits/rejected": -2.730475902557373,
      "logps/chosen": -261.4306945800781,
      "logps/pi_response": -78.0402603149414,
      "logps/ref_response": -72.56087493896484,
      "logps/rejected": -178.43402099609375,
      "loss": 0.6826,
      "rewards/accuracies": 0.5555555820465088,
      "rewards/chosen": -0.026773735880851746,
      "rewards/margins": 0.01566091552376747,
      "rewards/rejected": -0.04243465140461922,
      "step": 10
    },
    {
      "epoch": 0.5095541401273885,
      "grad_norm": 7.686707777817671,
      "learning_rate": 2.8355831645441387e-07,
      "logits/chosen": -2.631629467010498,
      "logits/rejected": -2.6079890727996826,
      "logps/chosen": -261.0021057128906,
      "logps/pi_response": -111.87821960449219,
      "logps/ref_response": -70.75907897949219,
      "logps/rejected": -207.450439453125,
      "loss": 0.6217,
      "rewards/accuracies": 0.6656249761581421,
      "rewards/chosen": -0.24031546711921692,
      "rewards/margins": 0.06851284205913544,
      "rewards/rejected": -0.30882829427719116,
      "step": 20
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 9.468696729808302,
      "learning_rate": 7.723433775328384e-08,
      "logits/chosen": -2.556835651397705,
      "logits/rejected": -2.5192666053771973,
      "logps/chosen": -305.44903564453125,
      "logps/pi_response": -155.01785278320312,
      "logps/ref_response": -70.76898193359375,
      "logps/rejected": -243.7476043701172,
      "loss": 0.5527,
      "rewards/accuracies": 0.6656249761581421,
      "rewards/chosen": -0.5297790765762329,
      "rewards/margins": 0.16432583332061768,
      "rewards/rejected": -0.694104790687561,
      "step": 30
    },
    {
      "epoch": 0.9936305732484076,
      "step": 39,
      "total_flos": 0.0,
      "train_loss": 0.5961279991345528,
      "train_runtime": 1787.1353,
      "train_samples_per_second": 5.596,
      "train_steps_per_second": 0.022
    }
  ],
  "logging_steps": 10,
  "max_steps": 39,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}