This repo uses langchain to ask pdf and interact with pinecone
The code is modified from
- https://github.com/alphasecio/langchain-examples/tree/main/pinecone-qa
- https://github.com/alejandro-ao/langchain-ask-pdf/tree/main
Please fill in your
- OpenAI API key
- Pinecone API key
- Pinecone environment
- Pinecone index name
The checkbox of reuse pinecone index will not called OpenAI embedding API to embed the documents. If the documents are already embed in the pinecone,
you can check the box to save your credit for OpenAI API. The checkbox will be automatically checked after you enter the first question, because the documents embedding will already be stored in your pinecone vector database.
- Install requirements
pip3 install -r requirements.txt
- Run the code
streamlit run app.py
1, Build docker
docker build -t streamlit .
- Run docker
docker run --rm -p 8501:8501 streamlit