2025年11月8日
初识PPO
摘要： for batch_prompt in prompt_dataset: batch_response = active_model.generate(batch_prompt) batch_data = concat(batch_prompt, batch_response) batch_score 阅读全文
posted @ 2025-11-08 18:46 leleleocc 阅读(18) 评论(0) 推荐(0)