all_results.json

{
    "epoch": 1.9978142076502732,
    "eval_logits/chosen": -1.510839581489563,
    "eval_logits/rejected": -1.4192240238189697,
    "eval_logps/chosen": -0.842383861541748,
    "eval_logps/rejected": -2.170888900756836,
    "eval_loss": 1.4397584199905396,
    "eval_rewards/accuracies": 0.8795180916786194,
    "eval_rewards/chosen": -8.423839569091797,
    "eval_rewards/margins": 13.28504753112793,
    "eval_rewards/rejected": -21.708887100219727,
    "eval_runtime": 33.4155,
    "eval_samples": 1318,
    "eval_samples_per_second": 39.443,
    "eval_semantic_entropy": 0.8327488899230957,
    "eval_steps_per_second": 2.484,
    "total_flos": 0.0,
    "train_loss": 1.6402891297830795,
    "train_runtime": 11806.3913,
    "train_samples": 58558,
    "train_samples_per_second": 9.92,
    "train_steps_per_second": 0.077
}