all_results.json

{
    "epoch": 3.97,
    "eval_logits/chosen": -0.06500499695539474,
    "eval_logits/rejected": 0.03410104289650917,
    "eval_logps/chosen": -307.4422302246094,
    "eval_logps/rejected": -279.5290832519531,
    "eval_loss": 0.0018018082482740283,
    "eval_rewards/accuracies": 0.49549999833106995,
    "eval_rewards/chosen": -0.006038154941052198,
    "eval_rewards/margins": -0.00034422188764438033,
    "eval_rewards/rejected": -0.0056939334608614445,
    "eval_runtime": 512.3455,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.904,
    "eval_steps_per_second": 0.976,
    "train_loss": 0.0010581601813863663,
    "train_runtime": 3353.9187,
    "train_samples": 61135,
    "train_samples_per_second": 1.193,
    "train_steps_per_second": 0.074
}