all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.769179582595825,
    "eval_logits/rejected": -2.7323365211486816,
    "eval_logps/chosen": -283.5445861816406,
    "eval_logps/rejected": -261.25567626953125,
    "eval_loss": 0.6794171333312988,
    "eval_rewards/accuracies": 0.6209999918937683,
    "eval_rewards/chosen": 0.010488319210708141,
    "eval_rewards/margins": 0.037255968898534775,
    "eval_rewards/margins_max": 0.1949017494916916,
    "eval_rewards/margins_min": -0.09943282604217529,
    "eval_rewards/margins_std": 0.09792803227901459,
    "eval_rewards/rejected": -0.026767650619149208,
    "eval_runtime": 428.8441,
    "eval_samples": 2000,
    "eval_samples_per_second": 4.664,
    "eval_steps_per_second": 0.291,
    "train_loss": 0.6679948979651021,
    "train_runtime": 9131.6092,
    "train_samples": 10941,
    "train_samples_per_second": 1.198,
    "train_steps_per_second": 0.075
}