eval_results.json

{
    "epoch": 2.9904761904761905,
    "eval_log_odds_chosen": 0.367882639169693,
    "eval_log_odds_ratio": -0.6760825514793396,
    "eval_logits/chosen": 299.86248779296875,
    "eval_logits/rejected": 270.96282958984375,
    "eval_logps/chosen": -1.1505995988845825,
    "eval_logps/rejected": -1.3976730108261108,
    "eval_loss": 1.59840989112854,
    "eval_nll_loss": 1.5312451124191284,
    "eval_rewards/accuracies": 0.5899280309677124,
    "eval_rewards/chosen": -0.057529982179403305,
    "eval_rewards/margins": 0.012353661470115185,
    "eval_rewards/rejected": -0.06988365203142166,
    "eval_runtime": 112.534,
    "eval_samples": 553,
    "eval_samples_per_second": 4.914,
    "eval_steps_per_second": 1.235
}