all_results.json

{
    "epoch": 1.0,
    "eval_logits/chosen": -2.2309982776641846,
    "eval_logits/rejected": -2.2261931896209717,
    "eval_logps/chosen": -33.767173767089844,
    "eval_logps/rejected": -37.282623291015625,
    "eval_loss": 0.49339157342910767,
    "eval_rewards/accuracies": 0.5456810593605042,
    "eval_rewards/chosen": 0.21390400826931,
    "eval_rewards/margins": 0.026708098128437996,
    "eval_rewards/rejected": 0.18719588220119476,
    "eval_runtime": 145.5786,
    "eval_samples": 343,
    "eval_samples_per_second": 2.356,
    "eval_steps_per_second": 0.295,
    "train_loss": 0.4538224170734356,
    "train_runtime": 3252.427,
    "train_samples": 3079,
    "train_samples_per_second": 0.947,
    "train_steps_per_second": 0.118
}