A RAG project with LangChain and Pinecone as a proof of concept for augmenting the capabilities of Meta's LLama-7b-chat model with a tailored data set. This uses LangChain for pipelining, and Pinecone as a vector database.
To run, amend the path.txt to contain the dataset path. Upload the RAG file to Google Collab or notebook of choice. Note that you will need access to T4 or A100 GPUs for ideal performance.