all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -0.8573879599571228,
    "eval_logits/rejected": -0.7623493671417236,
    "eval_logps/chosen": -399.9150085449219,
    "eval_logps/rejected": -371.0875549316406,
    "eval_loss": 0.05358276888728142,
    "eval_rewards/accuracies": 0.46950000524520874,
    "eval_rewards/chosen": -0.003609743434935808,
    "eval_rewards/margins": 0.0002423849218757823,
    "eval_rewards/rejected": -0.0038521282840520144,
    "eval_runtime": 545.4002,
    "eval_samples": 2000,
    "eval_samples_per_second": 3.667,
    "eval_steps_per_second": 0.917,
    "train_loss": 0.14686254965953338,
    "train_runtime": 4173.6713,
    "train_samples": 61135,
    "train_samples_per_second": 1.198,
    "train_steps_per_second": 0.075
}