🗃️ feat: General File Support for OpenAI, Azure, Custom, Anthropic and Google (RAG) #2143
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ts and successful RAG
…n new conversation
…embedded files to be deleted on new conversation
…n new conversation
danny-avila
changed the title
|
Hello I tryed to setup it with hugingface but when I try to send a pdf file I have this issue :
I did the rag.yml setup with .env and db setup I restarted everything to apply the changes This error happend on two of my endpoints :
in addition this is my logs: |
Try without huggingface first, because I haven't fully tested it aside from the default model listed. Also it would be more helpful to see the RAG API logs in this case. Embedding dimensions might be an issue but you can try |
|
@delta-whiplash Note, you may need to use Also it may be related to your device and I can look into a more flexible setup. This setup works for me: EMBEDDINGS_PROVIDER=huggingface
EMBEDDINGS_MODEL=sentence-transformers/all-MiniLM-L6-v2 |
|
Ok I reviewed my .env (the rag-api one) I changed it and restarted everything but the issue is still there this is the logs from rag-api : the logs from librechat : this is my rag-api .env file : this is my librechat yaml : |
|
I believe you are not making a successful connection to the RAG API. I will add a health check route to verify this. try in your browser to access it: http://rag_api:port/docs |
|
You're right I can't access it BTW I observed multiples restart from the container |
Happens when the API can't access the PSQL/pgvector db. note that |
|
I FOUND IT !!!! now I have new logs : |
Thanks will look into it. I ask that you start a discussion here for further help as I'm locking this PR as merged. |
Repository owner
locked as resolved and limited conversation to collaborators
Mar 20, 2024
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
Implements general file support with a basic implementation of RAG using augmented prompt approach.
This commit will introduce the feature mainly for testing as it will not be part of the default setup for now.
Setup
- Use
rag.ymlcompose file:docker compose -f rag.yml up- More info: https://github.com/danny-avila/rag_api
- Uses PSQL/PGVector + FastAPI
- Uses OpenAI embeddings by default
- Can use HuggingFace Sentence transformers but not thoroughly tested
- Open to contributions
- Current setup relies on sharing volume storage, namely the
uploadsdirectory at project root.- Set
RAG_API_URLaccordinglyRAG_API_URL=http://localhost:8000RAG_API_URL=http://host.docker.internal:8000Current Approach to RAG is subject to change for improvement.
Resend Files
resendImageswithresendFilesand now defaults to true.Other Changes
Change Type
Checklist