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

[Roadmap] Custom OPEN AI Endpoint #323

Closed
gabelul opened this issue Jan 3, 2024 · 10 comments
Closed

[Roadmap] Custom OPEN AI Endpoint #323

gabelul opened this issue Jan 3, 2024 · 10 comments
Milestone

Comments

@gabelul
Copy link

gabelul commented Jan 3, 2024

Allows users to call your custom endpoint in the OpenAI ChatCompletion format. I saw someone mentioning litellm

@enricoros
Copy link
Owner

enricoros commented Jan 3, 2024

@gabelul Is it possible already by going to OpenAI > Advanced Options?

@gabelul
Copy link
Author

gabelul commented Jan 3, 2024

@enricoros, that's what I thought at first, but I tried with the same settings that work with others, and it didn't work

@enricoros
Copy link
Owner

I need some more details to reproduce this. This is a very frequent use case, and we have no reports of malfunction.

Can you attach a screenshot that shows the error or walk through the reproduction steps?

@enricoros
Copy link
Owner

Hi @gabelul please provide more details around the use case or the issue - it's not clear to me. Otherwise I'll have to close the request.

@gabelul
Copy link
Author

gabelul commented Jan 6, 2024

Hey @enricoros, it does seem to work with some of the APIs that way. Not sure how can I send you a PM with the ones that don't so you can take a look ( they are private API )

@enricoros
Copy link
Owner

You can join discord and send me a pm. And attach screenshots of what's not working, maybe I can improve the error messages.

@shiribailem
Copy link

Dropping this here because I think it's related: ideally there would be functionality to have multiple instances of the OpenAI interface, just because someone might want to use both OpenAI directly and one or more systems that share it's API.

@shiribailem
Copy link

I'd imagine putting it in as a separate "OpenAI Compatible" model feature that can be added multiple times (or if that's not worth the effort, maybe just copy it 5 times as "OpenAI Compatible (1)" through "OpenAI Compatible (5)"

@enricoros
Copy link
Owner

@shiribailem You can do this today by changing this 1 line of code: https://github.com/enricoros/big-agi/blob/main/src%2Fmodules%2Fllms%2Fvendors%2Fopenai%2Fopenai.vendor.ts#L42

It's set to a single instance right now just to simplify the UX, but definitely you may create as many OpenAI compatible sources as you'd like.

enricoros added a commit that referenced this issue Jan 9, 2024
@enricoros
Copy link
Owner

@shiribailem added the option for 1-5 OpenAI compatible endpoints (see below).
@shiribailem please let me know what is not working as expected in the OpenAI compatibility (which should be okay).

Screenshot1: adding OpenAI 4
image

Screenshot 2:
image

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