Problem/Motivation

It is not clear how to tell which endpoints are appropriate for use, given that there are (May 23) 1.7 million of them.

Proposed resolution

Clearly indicate what type of endpoint to use, give indications in the UI about it for different use cases.

Remaining tasks

Work out what to do.
Do that.

User interface changes

The Embeddings Model "Endpoint" field gives indications as to what sort of model to use, maybe for different use cases.

API changes

TBD

Data model changes

TBD

Comments

damienmckenna created an issue. See original summary.

damienmckenna’s picture

Component: Documentation » User interface
damienmckenna’s picture

Title: Document what the endpoint should be set to » Indicate what the endpoint should be set to
damienmckenna’s picture

Title: Indicate what the endpoint should be set to » More clearly indicate what the endpoint should be set to, or use a select field?

Right now you have to guess what the endpoint value should be - it might be useful to provide a selector?

damienmckenna’s picture

Project: OpenSearch VDB Provider » Huggingface Provider
Version: 1.0.0-alpha1 » 1.0.x-dev
damienmckenna’s picture

Issue summary: View changes

After I connected the correct API key 🤦‍♂️ it still doesn't work and gives this error instead:

Could not load the embeddings engine to get the dimensions. Please check the configuration.Client error: `POST https://api-inference.huggingface.co/models/intfloat/e5-mistral-7b-instruct` resulted in a `404 Not Found` response: Not Found

damienmckenna’s picture

Title: More clearly indicate what the endpoint should be set to, or use a select field? » More clearly indicate what type of endpoint to use
Issue summary: View changes
damienmckenna’s picture