all_results.json

{
    "epoch": 4.0,
    "eval_logits/chosen": -2.829094648361206,
    "eval_logits/rejected": -2.8266069889068604,
    "eval_logps/chosen": -33.231201171875,
    "eval_logps/rejected": -37.228851318359375,
    "eval_loss": 0.8066554665565491,
    "eval_rewards/accuracies": 0.5917773842811584,
    "eval_rewards/chosen": -1.1692489385604858,
    "eval_rewards/margins": 0.3491671681404114,
    "eval_rewards/rejected": -1.518416166305542,
    "eval_runtime": 112.7022,
    "eval_samples": 343,
    "eval_samples_per_second": 3.043,
    "eval_steps_per_second": 0.382,
    "train_loss": 0.21652605050763526,
    "train_runtime": 11213.836,
    "train_samples": 3079,
    "train_samples_per_second": 1.098,
    "train_steps_per_second": 0.137
}