all_results.json

{
    "epoch": 3.0,
    "eval_logits/chosen": -1.4396553039550781,
    "eval_logits/rejected": -1.4439030885696411,
    "eval_logps/chosen": -151.50469970703125,
    "eval_logps/rejected": -180.52976989746094,
    "eval_loss": 2.0568761825561523,
    "eval_rewards/accuracies": 0.6666666865348816,
    "eval_rewards/chosen": 9.014835357666016,
    "eval_rewards/margins": 1.1928315162658691,
    "eval_rewards/rejected": 7.822003364562988,
    "eval_runtime": 280.3771,
    "eval_samples_per_second": 2.514,
    "eval_steps_per_second": 2.514,
    "train_loss": 0.8950637202828078,
    "train_runtime": 14615.0375,
    "train_samples_per_second": 0.985,
    "train_steps_per_second": 0.123
}