This is a Starter Repository for RAG. In this we will be building out a PDF RAG Agent to chat with. The documents in pdf format will be the contect corpus to use for embeddings that we will index in the pinecone graph index.
- fork and clone this repository
- Create a pinecone account here
- setup an index and call it
pineidx
- create a Virtual Env
python3 -m venv .venv
- install the requirements
pip install -r requirements.txt
- look over the python files and read the comments
- get a pinecone key from your account
- export the openaikey
export OPENAI_API_KEY=pk_your_key_here
- export the pinecone key
export PINECONE_API_KEY=d0_your_pinecone_key_here
- get started writing out the upload logic
- add some pdf documents in to the docs directory
- run the upload.py
python upload.py
- if it was successful then we can move on to building out the main file logic
- once we have the main file done try it out
python main.py
- if it is working we can build a repl and make the prompt be more useful maybe even choose more pdf files to upload