all_results.json

{
    "epoch": 0.9989071038251366,
    "eval_logits/chosen": -1.200439691543579,
    "eval_logits/rejected": -1.150753140449524,
    "eval_logps/chosen": -2.5561068058013916,
    "eval_logps/rejected": -3.309354782104492,
    "eval_loss": 1.7538212537765503,
    "eval_rewards/accuracies": 0.8433734774589539,
    "eval_rewards/chosen": -25.561067581176758,
    "eval_rewards/margins": 7.532484531402588,
    "eval_rewards/rejected": -33.09355163574219,
    "eval_runtime": 33.6933,
    "eval_samples": 1318,
    "eval_samples_per_second": 39.118,
    "eval_semantic_entropy": 0.36752766370773315,
    "eval_steps_per_second": 2.463,
    "total_flos": 0.0,
    "train_loss": 2.585477126766347,
    "train_runtime": 5989.402,
    "train_samples": 58558,
    "train_samples_per_second": 9.777,
    "train_steps_per_second": 0.076
}