all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.6051297187805176,
    "eval_logits/rejected": -2.574681043624878,
    "eval_logps/chosen": -304.7639465332031,
    "eval_logps/rejected": -317.67572021484375,
    "eval_loss": 0.5925807952880859,
    "eval_rewards/accuracies": 0.7064999938011169,
    "eval_rewards/chosen": -0.2030859887599945,
    "eval_rewards/margins": 0.3151388466358185,
    "eval_rewards/margins_max": 1.1834423542022705,
    "eval_rewards/margins_min": -0.5405970811843872,
    "eval_rewards/margins_std": 0.5821082592010498,
    "eval_rewards/rejected": -0.518224835395813,
    "eval_runtime": 859.7381,
    "eval_samples": 4000,
    "eval_samples_per_second": 4.653,
    "eval_steps_per_second": 0.291,
    "train_loss": 0.6106676421631342,
    "train_runtime": 67977.0297,
    "train_samples": 66812,
    "train_samples_per_second": 0.983,
    "train_steps_per_second": 0.061
}