all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.498661994934082,
    "eval_logits/rejected": -2.4008209705352783,
    "eval_logps/chosen": -269.65570068359375,
    "eval_logps/rejected": -257.6377868652344,
    "eval_loss": 0.6512373089790344,
    "eval_rewards/accuracies": 0.6639999747276306,
    "eval_rewards/chosen": -0.06204665079712868,
    "eval_rewards/margins": 0.10243026912212372,
    "eval_rewards/rejected": -0.1644769161939621,
    "eval_runtime": 542.9483,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.684,
    "eval_steps_per_second": 0.46,
    "train_loss": 0.0,
    "train_runtime": 0.0278,
    "train_samples": 10000,
    "train_samples_per_second": 359461.447,
    "train_steps_per_second": 1401.9
}