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

Does it support private deployment #8

Open
su-zelong opened this issue Dec 7, 2023 · 6 comments
Open

Does it support private deployment #8

su-zelong opened this issue Dec 7, 2023 · 6 comments

Comments

@su-zelong
Copy link

May I ask if this plugin supports local models packaged in OpenAI format using VLLM

@WissamAntoun
Copy link

Hey, I created a version of this app last week that addresses this https://github.com/WissamAntoun/GalacTex

@su-zelong
Copy link
Author

su-zelong commented Dec 7, 2023

Hey, I created a version of this app last week that addresses this https://github.com/WissamAntoun/GalacTex

Thank you very much for your reply. I have carefully read this app. Is it currently only supported for Galactica models? Can I respond if I use URLs from other local models? For example: chatglm+vllm

@WissamAntoun
Copy link

it was intended for galactica, but it should support any model supported by vllm

@achiyae
Copy link
Member

achiyae commented Dec 7, 2023

Interesting. I was not aware of vLLM/Galactica.

Yesterday, I released version 1.4.0, which should allow connecting any vLLM model.
This version changes the API to the chat completion API and adds the ability to configure the plugin via a JSON editor that was added to the configuration page. Since vLLM supports chat completion API - the plugin will work with vLLM as well. See this image:
image

With this editor, you can change the URL and configure the hyperparameters for the model. You can do that for each command and for all of them. You will still need to set the API key to enable the plugin, but it doesn't have to be a real API key.

@su-zelong
Copy link
Author

Interesting. I was not aware of vLLM/Galactica.

Yesterday, I released version 1.4.0, which should allow connecting any vLLM model. This version changes the API to the chat completion API and adds the ability to configure the plugin via a JSON editor that was added to the configuration page. Since vLLM supports chat completion API - the plugin will work with vLLM as well. See this image: image

With this editor, you can change the URL and configure the hyperparameters for the model. You can do that for each command and for all of them. You will still need to set the API key to enable the plugin, but it doesn't have to be a real API key.

What a great job! This extension will definitely become popular.

@achiyae
Copy link
Member

achiyae commented Dec 8, 2023

Thanks!

Let me know how if it works indeed.
If it does, I would appreciate it if you could add an explanation on how to do that to README.md.

Help us spreading the word by staring the project and adding a review to the extension homepage.

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