This repo is an implementation of a locally hosted chatbot specifically focused on question answering over the LangChain documentation. Built with LangChain and FastAPI.

The app leverages LangChain's streaming support and async API to update the page in real time for multiple users.

1. Install dependencies: pip install -r requirements.txt
2. Create a copy of .env.template, call it .env and update with your unique OpenAI API Key after the =, without any quotes or spaces.
3. Run ingest.sh to ingest LangChain docs data into the vectorstore (only needs to be done once).
   1. If on Windows, Run ingest.bat instead. Must have wget for windows installed and updated (instructions here).
   2. You can use other Document Loaders to load your own data into the vectorstore.
4. Run the app: make start
   1. If on Windows, Run python start.py.
   2. To enable tracing, make sure langchain-server is running locally and pass tracing=True to get_chain in main.py. You can find more documentation here.
5. Open localhost:9000 in your browser.

Deployed version (to be updated soon): chat.langchain.dev

Hugging Face Space (to be updated soon): huggingface.co/spaces/hwchase17/chat-langchain

Blog Posts:

• Initial Launch
• Streaming Support

There are two components: ingestion and question-answering.

Ingestion has the following steps:

1. Pull html from documentation site
2. Load html with LangChain's ReadTheDocs Loader
3. Split documents with LangChain's TextSplitter
4. Create a vectorstore of embeddings, using LangChain's vectorstore wrapper (with OpenAI's embeddings and FAISS vectorstore).

Question-Answering has the following steps, all handled by ChatVectorDBChain:

1. Given the chat history and new user input, determine what a standalone question would be (using GPT-3).
2. Given that standalone question, look up relevant documents from the vectorstore.
3. Pass the standalone question and relevant documents to GPT-3 to generate a final answer.