all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.04455828666687,
    "eval_logits/rejected": -2.031615972518921,
    "eval_logps/chosen": -434.86834716796875,
    "eval_logps/rejected": -446.6819152832031,
    "eval_loss": 0.6363195180892944,
    "eval_rewards/accuracies": 0.6587926745414734,
    "eval_rewards/chosen": -2.122913122177124,
    "eval_rewards/margins": 0.4067881107330322,
    "eval_rewards/mix_margin": 0.14258870482444763,
    "eval_rewards/rejected": -2.529700994491577,
    "eval_runtime": 719.4785,
    "eval_samples": 1522,
    "eval_samples_per_second": 2.115,
    "eval_steps_per_second": 0.53,
    "train_loss": 0.5550967540059771,
    "train_runtime": 33981.529,
    "train_samples": 29585,
    "train_samples_per_second": 0.871,
    "train_steps_per_second": 0.014
}