LangChain Sand Box
This is a working repository for various experiments that I am conducting with LangChain.
I use the the python-dotenv
library to maintain the necessary API keys.
The files that I am using reside in the data subdirectory.
A collection of results from different experiments can be found in the screenshots subdirectory.
The summarize.py script
is based on https://medium.com/mlearning-ai/building-a-custom-summarization-app-with-streamlit-and-langchain-11ab19099822
The create_vectordb.py
script can be used to create a Chroma vector database from a source directory. Type python create_vectordb.py -h
for more information.
The chat_with_vectordb.py
script takes a path to a directory containing a Chroma vector database and initiates a chat. You can quit the chat by typing quit
and start a new conversation by typing clear
. The sources relied upon for each answer are included.