all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -1.4588701725006104,
    "eval_logits/rejected": -1.4782198667526245,
    "eval_logps/chosen": -299.7252197265625,
    "eval_logps/rejected": -1401.1368408203125,
    "eval_loss": 0.13985097408294678,
    "eval_rewards/accuracies": 0.9375,
    "eval_rewards/chosen": -4.069133281707764,
    "eval_rewards/margins": 58.453983306884766,
    "eval_rewards/rejected": -62.52311706542969,
    "eval_runtime": 21.8141,
    "eval_samples_per_second": 17.42,
    "eval_steps_per_second": 2.2,
    "total_flos": 4121088491520.0,
    "train_loss": 0.18359575779349716,
    "train_runtime": 542.9915,
    "train_samples_per_second": 6.298,
    "train_steps_per_second": 0.099
}