- Website Interaction: The chatbot uses LangChain to interact with and extract information from various websites.
- Large Language Model Integration: Powered by GPT-4.
- Streamlit GUI: A clean and intuitive user interface built with Streamlit.
- Python-based: Entirely coded in Python.
- CSS: Used CSS to make the GUI more attractive and user-friendly.
- User Paste's any website link to the input box.
- All the website text is scrapped using BeautifulSoup.
- Text is split into small chunks so that it can be fed to the Embeddings Model.
- Embeddings model performs vectorization and stores it in vector data base.
- The user query is also embeded and top matching text present in the vector data base is found out through semantic search.
- The Top ranked text found from our webpage data and the user query both are feeded to the LLM model.
- LLM model produces the answer to the query and displays it to the user.
- A RAG chatbot is an AI-powered chatbot that uses Retrieval Augmented Generation (RAG) to provide more accurate and relevant responses than traditional chatbots.
- For the Demo we are using Wikipedia Page of "INDIA IN OLYMPICS 2024".
- Paste the link of Wikipedia website in the input box.
- After the link is pasted, the bot automatically start.
- Question 1
- Question 2
- Question 3. The bot also remembers the chat history and answers accordingly.
- Question 4. It also summarizes the webpage information.
