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

genai.vector.encode doesn't allow for changing OpenAI API endpoints #13433

Open
teowz46 opened this issue Mar 27, 2024 · 1 comment
Open

genai.vector.encode doesn't allow for changing OpenAI API endpoints #13433

teowz46 opened this issue Mar 27, 2024 · 1 comment

Comments

@teowz46
Copy link

teowz46 commented Mar 27, 2024

Feature request template

Is your feature request related to a problem? Please describe.
I am unable to use this function because I am on a network which has an internal OpenAI API endpoint. However, it seems like there is no way to change the endpoint to something that is not api.openai.com.

Describe the solution you'd like
Add a configuration option in configuration Map to allow users to pass in custom API endpoints.

Describe alternatives you've considered
Couldn't think of alternatives other than changing the source code.

Additional context
NIL

@parnmatt
Copy link
Contributor

parnmatt commented Mar 27, 2024

The GenAI plugin was written with Aura in mind, and as such we are only allowing known and controlled endpoints and options.

However, we have implemented the plugin with such thoughts of extension in mind, service loading vector embedding providers. Once we finalise the interfaces we will likely expose the interface publically such that on-prem enterprise customers can configure their custom embedding providers, which of course may be private and sensitive.

Due to many exposing an OpenAI-like API endpoint, we may expose an easier way to specifically configure such providers; such as a specific base class to extend, and/or a "simple" provider with a custom URI and pass everything on, etc, as things we are considering specifically for on-prem.

Until then, if it is necessary to use your internal custom endpoint, APOC Extended has https://neo4j.com/labs/apoc/5/ml/openai/ which can be used for this purpose in the mean time.

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

No branches or pull requests

3 participants