all_results.json

{
    "epoch": 1.0,
    "eval_dpo_losses": 0.6513049602508545,
    "eval_logits/chosen": -2.831768035888672,
    "eval_logits/rejected": -2.788257598876953,
    "eval_logps/chosen": -274.86163330078125,
    "eval_logps/rejected": -259.8882751464844,
    "eval_loss": 1.0971728563308716,
    "eval_positive_losses": 4.613582134246826,
    "eval_rewards/accuracies": 0.6269841194152832,
    "eval_rewards/chosen": 0.10359582304954529,
    "eval_rewards/margins": 0.11065211147069931,
    "eval_rewards/margins_max": 0.4366241991519928,
    "eval_rewards/margins_min": -0.18976463377475739,
    "eval_rewards/margins_std": 0.27969247102737427,
    "eval_rewards/rejected": -0.007056289818137884,
    "eval_runtime": 283.0495,
    "eval_samples": 2000,
    "eval_samples_per_second": 7.066,
    "eval_steps_per_second": 0.223,
    "train_loss": 0.641208895814349,
    "train_runtime": 6249.598,
    "train_samples": 10941,
    "train_samples_per_second": 1.751,
    "train_steps_per_second": 0.109
}