{ "epoch": 2.0, "eval_logits/chosen": 1.4099805355072021, "eval_logits/rejected": 2.3523759841918945, "eval_logps/chosen": -426.9287414550781, "eval_logps/rejected": -553.7540283203125, "eval_loss": 0.5016762614250183, "eval_rewards/accuracies": 0.7890625, "eval_rewards/chosen": -1.6429880857467651, "eval_rewards/margins": 1.267931342124939, "eval_rewards/rejected": -2.910919189453125, "eval_runtime": 92.7226, "eval_samples": 2000, "eval_samples_per_second": 21.57, "eval_steps_per_second": 0.345 }