all_results.json

{
    "epoch": 4.695652173913043,
    "eval_logits/chosen": -0.107421875,
    "eval_logits/rejected": 0.0263671875,
    "eval_logps/chosen": -129.0,
    "eval_logps/rejected": -171.0,
    "eval_loss": 0.5796619057655334,
    "eval_rewards/accuracies": 1.0,
    "eval_rewards/chosen": -0.007049560546875,
    "eval_rewards/margins": 0.255859375,
    "eval_rewards/rejected": -0.26171875,
    "eval_runtime": 5.9126,
    "eval_samples_per_second": 41.268,
    "eval_steps_per_second": 1.353
}