all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.2434139251708984,
    "eval_logits/rejected": -2.2385354042053223,
    "eval_logps/chosen": -34.0224494934082,
    "eval_logps/rejected": -37.51002883911133,
    "eval_loss": 0.9990109801292419,
    "eval_rewards/accuracies": 0.5141196250915527,
    "eval_rewards/chosen": 0.0024205157533288,
    "eval_rewards/margins": 0.0011023369152098894,
    "eval_rewards/rejected": 0.0013181795366108418,
    "eval_runtime": 145.5207,
    "eval_samples": 343,
    "eval_samples_per_second": 2.357,
    "eval_steps_per_second": 0.295,
    "train_loss": 0.998242255619594,
    "train_runtime": 3253.3829,
    "train_samples": 3079,
    "train_samples_per_second": 0.946,
    "train_steps_per_second": 0.118
}