eval_results.json

{
    "epoch": 10.0,
    "eval_logits/chosen": -0.3184443712234497,
    "eval_logits/rejected": -0.36903008818626404,
    "eval_logps/chosen": -0.1208958849310875,
    "eval_logps/rejected": -0.6922038793563843,
    "eval_loss": 0.1529119908809662,
    "eval_odds_ratio_loss": 1.3675727844238281,
    "eval_rewards/accuracies": 0.8125,
    "eval_rewards/chosen": -0.012089588679373264,
    "eval_rewards/margins": 0.05713079497218132,
    "eval_rewards/rejected": -0.06922038644552231,
    "eval_runtime": 3.3385,
    "eval_samples_per_second": 23.963,
    "eval_sft_loss": 0.016154715791344643,
    "eval_steps_per_second": 11.982
}