all_results.json

{
    "epoch": 2.76,
    "eval_logits/chosen": -2.274653911590576,
    "eval_logits/rejected": -2.2999308109283447,
    "eval_logps/chosen": -190.57220458984375,
    "eval_logps/rejected": -132.8740234375,
    "eval_loss": 0.68434077501297,
    "eval_rewards/accuracies": 0.5,
    "eval_rewards/chosen": 0.04404526203870773,
    "eval_rewards/margins": 0.03693275526165962,
    "eval_rewards/rejected": 0.0071125030517578125,
    "eval_runtime": 6.6846,
    "eval_samples": 30,
    "eval_samples_per_second": 4.488,
    "eval_steps_per_second": 0.15,
    "train_loss": 0.6927925229072571,
    "train_runtime": 943.7212,
    "train_samples": 922,
    "train_samples_per_second": 4.885,
    "train_steps_per_second": 0.005
}