关于CogVLM多轮对话训练的问题 #317
Replies: 1 comment
-
|
是的,开源版本的模型是这么做,仅对最后一轮的response进行损失计算 |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
问题描述:
您好,我正在研究CogVLM的微调代码,通过观察utils/utils/language.py文件中的_history_to_prompt和llama2_text_processor,并不支持类似这里Firefly方法提到的的高效充分的多轮对话训练方式。
具体疑虑:
因此我想了解一下,CogVLM的训练(如cogagent-chat、cogvlm-chat-v1.1)都是采用了拼接完多轮对话(chat_history_to_prompt与chat_old_history_to_prompt)后,仅对最后一轮的response进行有效的损失计算的方式?
非常期待得到社区的回答和建议,谢谢大家的时间和帮助!
Beta Was this translation helpful? Give feedback.
All reactions