avisp / streamlit_test Goto Github PK

This repo shows how to host a streamlit webapp on the community cloud or other webhosting platforms (heroku, ploomber) that makes API call to an AI service (beam.cloud) running on a serverless infrastructure for generative AI

The example shown here will do a sentiment analysis of a given sentence and will display the results provided by a sentiment roberta model.

1. Perform the installation steps for beam
2. Run beam deploy beam_app.py and wait for deplyment to finish
3. Get the url and credentials from the beam dashboard
4. Start streamlit app using streamlit run streamlit_app_git.py
5. Enter the credentials
6. Enter a text and get the sentiment response

A new app streamlit_authenticate.py has been provided that does authentication based on username and password provided. A private google spreadsheet stores the authorized users that can be modified as needed. The details of how to connect streamlit to google sheets through API can be found here.

For local development a folder called .streamlit has to be created in root directory and a secrets.toml inside it where the google API configurations will be stored. For online deployment on streamlit community cloud it will go in Secrets