all_results.json

{
    "epoch": 1.0,
    "eval_dpo_losses": 0.6446115970611572,
    "eval_logits/chosen": -2.713839530944824,
    "eval_logits/rejected": -2.6717445850372314,
    "eval_logps/chosen": -284.5663146972656,
    "eval_logps/rejected": -271.5415954589844,
    "eval_loss": 1.616790771484375,
    "eval_positive_losses": 8.828282356262207,
    "eval_rewards/accuracies": 0.6499999761581421,
    "eval_rewards/chosen": 0.0002709717955440283,
    "eval_rewards/margins": 0.12989762425422668,
    "eval_rewards/margins_max": 0.6433730721473694,
    "eval_rewards/margins_min": -0.34936216473579407,
    "eval_rewards/margins_std": 0.33273470401763916,
    "eval_rewards/rejected": -0.12962664663791656,
    "eval_runtime": 428.5272,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.667,
    "eval_steps_per_second": 0.292,
    "train_loss": 0.5743894765074824,
    "train_runtime": 4311.1014,
    "train_samples": 5678,
    "train_samples_per_second": 1.317,
    "train_steps_per_second": 0.082
}