all_results.json

{
    "epoch": 1.99,
    "eval_logits/chosen": 0.8372604250907898,
    "eval_logits/rejected": 0.8945639133453369,
    "eval_logps/chosen": -256.5552978515625,
    "eval_logps/rejected": -233.49595642089844,
    "eval_loss": 0.01026910264045,
    "eval_rewards/accuracies": 0.4945000112056732,
    "eval_rewards/chosen": 0.0006274624029174447,
    "eval_rewards/margins": 0.00010108990682056174,
    "eval_rewards/rejected": 0.0005263725761324167,
    "eval_runtime": 709.1992,
    "eval_samples": 2000,
    "eval_samples_per_second": 2.82,
    "eval_steps_per_second": 0.705,
    "train_loss": 0.001956978782771095,
    "train_runtime": 290.9594,
    "train_samples": 61135,
    "train_samples_per_second": 6.874,
    "train_steps_per_second": 0.426
}