{ "epoch": 1.97, "eval_logits/chosen": 94.06119537353516, "eval_logits/rejected": 88.1175308227539, "eval_logps/chosen": -418.7874755859375, "eval_logps/rejected": -448.4524230957031, "eval_loss": 0.46823105216026306, "eval_rewards/accuracies": 0.7395833134651184, "eval_rewards/chosen": -2.838282823562622, "eval_rewards/margins": 1.6033902168273926, "eval_rewards/rejected": -4.4416728019714355, "eval_runtime": 57.2277, "eval_samples": 750, "eval_samples_per_second": 13.106, "eval_steps_per_second": 0.419 }