eval_results.json

{
    "epoch": 0.9880609304240429,
    "eval_logits/chosen": 9.272286415100098,
    "eval_logits/rejected": 10.088170051574707,
    "eval_logps/chosen": -94.22303771972656,
    "eval_logps/ref_chosen": -106.16099548339844,
    "eval_logps/ref_rejected": -119.78473663330078,
    "eval_logps/rejected": -108.40995025634766,
    "eval_loss": 0.6188845634460449,
    "eval_rewards/accuracies": 0.5546296238899231,
    "eval_rewards/chosen": 0.11937955021858215,
    "eval_rewards/margins": 0.005631651729345322,
    "eval_rewards/rejected": 0.11374790966510773,
    "eval_runtime": 154.55,
    "eval_samples": 1080,
    "eval_samples_per_second": 6.988,
    "eval_sft_loss": 0.3315518796443939,
    "eval_steps_per_second": 1.747
}