rag_api
rag_api copied to clipboard
danny-avila
ID-based RAG FastAPI: Integration with Langchain and PostgreSQL/pgvector
Removes debug code accidentally pushed to `main` and switch `images.yaml` workflow to use common `docker-build` workflow. Fixes #40
Add qdrant vector store to improve vectors database
Allows the rag-api to function with Pinecone as a vector store. Configuration follows suit with the PGVector and MongoDB solutions by having details added in the `.env` file which are...
Would it feasible to generate the vector search index upon startup? This approach may simplify deployments, as the API could establish a connection with MongoDB and instantiate the index. Have...
- Created a new rag_api directory to house all the core components of the project. - Inside rag_api, I organized things into a few subdirectories to keep things tidy. Hopefully,...
Hey, For faster processing OpenAI (and Azure) allow to pass a `dimensions` parameter to text-embedding-3 models. Would it be possible to also integrate this as an environment variable? Thanks!
I noticed you say Anthropic is supported for RAG API....But I see no real reference for how to configure that part. All the config references OpenAI and others, not Anthropic....
what follows is the log from the container: `2024-07-18T06:10:51.977995561Z return wrapped(self, *args, **kwargs) 2024-07-18T06:10:51.978002664Z File "/usr/local/lib/python3.10/site-packages/langchain_community/vectorstores/pgvector.py", line 316, in __init__ 2024-07-18T06:10:51.978022051Z self._bind = connection if connection else self._create_engine() 2024-07-18T06:10:51.978050684Z File...
Metadata
Owner
Metadata
ID-based RAG FastAPI: Integration with Langchain and PostgreSQL/pgvector