A basic gradio app to chat with local documents
LangChain used for LLM tooling
[https://python.langchain.com/docs/get_started/introduction.html]
Chroma used for vector database
[https://www.trychroma.com/]
This will ingest the pdf file in the data directory, store it as vectors in the .chromadb directory, then provide the most relevant vectors as context in the chat questions.
- Install dependencies:
pip install -r requirements.txt
- Create local .env for configuration secrets:
mv .env.example .env
{update values in .env}
- Start gradio app:
python chat.py
- Open browser:
- Ask away!
"What are the main motivations for the author to climb?"