Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

基于fastchat部署,推理异常 #46

Open
ruifengma opened this issue Dec 1, 2023 · 3 comments
Open

基于fastchat部署,推理异常 #46

ruifengma opened this issue Dec 1, 2023 · 3 comments

Comments

@ruifengma
Copy link

用fastchat部署AgentLM-13B,推理的时候格式是乱的,尤其是streaming的模式,每行只有几个字符就切换到下一行了,一个单词被切成了好几个字母或字母组合。如果手动用transformer加载并用gradio展示的话就没有这个问题,用fastchat的debug模式看了一下,用的是LlamaForCausalLM加载的模型,应该没错

@Btlmd
Copy link
Member

Btlmd commented Dec 18, 2023

我们之前没有在 FastChat 上进行过推理实验。能否给出更详细的报错、不符合预期的场景或者配置?

@heimy2000
Copy link

用fastchat部署AgentLM-13B,推理的时候格式是乱的,尤其是streaming的模式,每行只有几个字符就切换到下一行了,一个单词被切成了好几个字母或字母组合。如果手动用transformer加载并用gradio展示的话就没有这个问题,用fastchat的debug模式看了一下,用的是LlamaForCausalLM加载的模型,应该没错

老师,请问您是怎么操作的呢

@ruifengma
Copy link
Author

hi @Btlmd @heimy2000 , 其实就是简单的使用 fastchat.serve.cli 来加载模型进行基础对话,其实没有报错,只是输出的chunk就很奇怪并且不成句,其实你们可以试试看,操作很简单
python3 -m fastchat.serve.cli --model-path AgentTuning/AgentLM
--model-path后面就是加的实际存储的AgentLM权重

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants