Skip to content

bharatr21/chat-with-your-docs

Repository files navigation

chat-with-your-docs

Chat with your Text Documents (upto 100MB in size)

Tech Stack Used:

  • LlamaIndex for parsing documents
  • ChromaDB for vector database storage, ranking and retrieval
  • Choice of OpenAI/HuggingFace as the LLM (Large Language Model) for both embedding and inference. Obtain an API key from here for OpenAI or here for HuggingFace
    • Embedding models are sentence-transformers/all-MiniLM-L6-v2 for HF and text-embedding-3-small for OpenAI
    • Inference models are meta-llama/Meta-Llama-3.1-8B-Instruct for HF and gpt-4o-mini for OpenAI

Try out the App here