🎃[Provider Request]: Implement Chroma as an Vector DB

[shabbyjoon]

Description

A Vector Database provider is designed to facilitate nearest-neighbor lookups.

Chroma is a popular open-source vector database that stores, indexes, and manages vector embeddings.

Featureform supports vector databases like Pinecone, Redis, and Weaviate.

In Featureform, Vector DBs share several similarities with an inference store but are distinguished by their support for the client.nearest API. Configuration is typically done when registering an embedding associated with an entity. This setup enables efficient retrieval of nearest neighbors based on feature vectors.

Current Behavior

No response

Desired Behavior

Implement Chroma as a vector database so that it can be registered as a provider in Featureform.

Benefits

With Chroma provider support, users can manage and version their ML embeddings with Featureform while leveraging Chroma's indexing, storage, and nearest-neighbor lookups.

Possible Implementation

It should implement the https://github.com/featureform/featureform/blob/main/provider/online.go#L36 interface.

Look at https://github.com/featureform/featureform/blob/main/provider/pinecone.go for inspiration

Additional Context

No response

[error9098x]

I would like to try this one please :) thanks

[error9098x]

@shabbyjoon

[Nupoor10]

Hello, Can I work on this? @shabbyjoon