all_results.json

{
    "epoch": 0.9989701338825953,
    "eval_logits/chosen": -2.0243144035339355,
    "eval_logits/rejected": -1.930372714996338,
    "eval_logps/chosen": -330.1051025390625,
    "eval_logps/rejected": -416.3554382324219,
    "eval_loss": 0.5171983242034912,
    "eval_rewards/accuracies": 0.75,
    "eval_rewards/chosen": -1.077383041381836,
    "eval_rewards/margins": 0.8083283305168152,
    "eval_rewards/rejected": -1.885711431503296,
    "eval_runtime": 65.4398,
    "eval_samples_per_second": 15.281,
    "eval_steps_per_second": 0.244
}