all_results.json

{
    "epoch": 1.0,
    "eval_abs_diff": 0.4453125,
    "eval_all_logps_1": -656.8973388671875,
    "eval_all_logps_1_values": -656.8973388671875,
    "eval_all_logps_2": 434.6329040527344,
    "eval_all_logps_2_values": 434.6329345703125,
    "eval_logits/chosen": 5.71875,
    "eval_logits/rejected": 5.6875,
    "eval_logps/chosen": -1.9453125,
    "eval_logps/rejected": -2.03125,
    "eval_loss": 1.7800946235656738,
    "eval_original_losses": 1.796875,
    "eval_rewards/accuracies": 0.4404762089252472,
    "eval_rewards/chosen": -4.875,
    "eval_rewards/margins": 0.2001953125,
    "eval_rewards/rejected": -5.0625,
    "eval_runtime": 7.6194,
    "eval_samples": 1961,
    "eval_samples_per_second": 257.37,
    "eval_steps_per_second": 0.919,
    "eval_weight": 1.0,
    "total_flos": 0.0,
    "train_loss": 1.8681734525240385,
    "train_runtime": 997.9993,
    "train_samples": 59876,
    "train_samples_per_second": 59.996,
    "train_steps_per_second": 0.026
}