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

[Feature] 可以添stop_token_ids 字段设置么。。。 #3128

Closed
YingchaoX opened this issue Nov 2, 2023 · 8 comments
Closed

[Feature] 可以添stop_token_ids 字段设置么。。。 #3128

YingchaoX opened this issue Nov 2, 2023 · 8 comments

Comments

@YingchaoX
Copy link

使用Qwen-14B-Chat作为backend,没有触发stop_token会一直输出文字。。。

@Issues-translate-bot
Copy link

Bot detected the issue body's language is not English, translate it automatically.


Title: [Feature] Can you add stop_token_ids field setting? . .

Using Qwen-14B-Chat as the backend, text will always be output if stop_token is not triggered. . .

@H0llyW00dzZ
Copy link
Contributor

not available at this moment because it will break client openai LLMApi which is the best already

@Yidadaa
Copy link
Collaborator

Yidadaa commented Nov 2, 2023

你在服务端设置好默认值就可以了,自建模型的优点不就是这样吗

@Yidadaa Yidadaa closed this as completed Nov 2, 2023
@Issues-translate-bot
Copy link

Bot detected the issue body's language is not English, translate it automatically.


You just need to set the default values ​​on the server. Isn't that the advantage of self-built models?

@CoderCore
Copy link

在使用 python -m vllm.entrypoints.openai.api_server 启动API服务器时,若你想为模型推理添加 sample_params(采样参数),通常这些参数不是直接通过命令行传递给API服务器的。相反,sample_params 是在API请求中作为参数传递给模型的。

@Issues-translate-bot
Copy link

Bot detected the issue body's language is not English, translate it automatically.


When using python -m vllm.entrypoints.openai.api_server to start the API server, if you want to add sample_params (sampling parameters) for model inference, usually these parameters are not passed directly to the API server through the command line. Instead, sample_params are passed as parameters to the model in the API request.

@kevinzhao2233
Copy link

在使用 python -m vllm.entrypoints.openai.api_server 启动API服务器时,若你想为模型推理添加 sample_params(采样参数),通常这些参数不是直接通过命令行传递给API服务器的。相反,sample_params 是在API请求中作为参数传递给模型的。

请问你现在是怎么解决这个问题的

@Issues-translate-bot
Copy link

Bot detected the issue body's language is not English, translate it automatically.


When using python -m vllm.entrypoints.openai.api_server to start the API server, if you want to add sample_params (sampling parameters) for model inference, usually these parameters are not passed directly to the API server through the command line. Instead, sample_params are passed as parameters to the model in the API request.

May I ask how you solve this problem now?

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

6 participants