Build a chatbot powered by LlamaIndex that augments Google Gemini with local documents. Based on the chatbot described in this blog post.
- Takes user queries via Streamlit's
st.chat_input
and displays both user queries and model responses withst.chat_message
- Uses LlamaIndex to load and index data and create a chat engine that will retrieve context from that data to respond to each user query
https://sjsu-chatbot-workshop.streamlit.app/
Follow the steps in the workshop slides at https://tiny.sjsu.edu/chatbot-workshop
Caution
Don't commit your secrets file to your GitHub repository. The .gitignore
file in this repo includes .streamlit/secrets.toml
and secrets.toml
.
Once the app is loaded, enter your question about the Streamlit library and wait for a response.
The demo version of the chatbot uses texts from the writer Rabindranath Tagore. These texts are available from Project Gutenberg.