| import streamlit as st | |
| import os | |
| # Set up page | |
| st.set_page_config( | |
| page_title="Aerospace Chatbot: AMS", | |
| ) | |
| st.title("Aerospace Chatbot Homepage") | |
| st.markdown("Code base: https://github.com/dsmueller3760/aerospace_chatbot/tree/rag_study") | |
| st.markdown('---') | |
| st.title("Chatbots") | |
| st.markdown(""" | |
| Chatbots for aerospace mechanisms symposia, using all available papers published since 2000 | |
| * Modular version meant to study retrieval methods | |
| """) | |
| st.subheader("AMS") | |
| ''' | |
| It is recommended to run this streamlit app locally for improved performance. The hosted hugging face version is for proof of concept. | |
| You must have poetry installed locally to manage depdenencies. To run locally, clone the repository and run the following commands. | |
| ```poetry config virtualenvs.in-project true | |
| ```poetry install | |
| ```source .venv/bin/activate | |
| ```cd ./scripts | |
| ```streamlit run Start.py | |
| ''' | |
| st.subheader("AMS") | |
| ''' | |
| This chatbot will look up from all Aerospace Mechanism Symposia in the following location: https://github.com/dsmueller3760/aerospace_chatbot/tree/main/data/AMS | |
| * Available models: https://platform.openai.com/docs/models | |
| * Model parameters: https://platform.openai.com/docs/api-reference/chat/create | |
| * Pinecone: https://docs.pinecone.io/docs/projects#api-keys | |
| * OpenAI API: https://platform.openai.com/api-keys | |
| ''' |