Skip to content

Latest commit

 

History

History
77 lines (52 loc) · 2.23 KB

README.md

File metadata and controls

77 lines (52 loc) · 2.23 KB
Raw

Frontend For Streamlit

This folder contains the frontend code for the OR Assistant using Streamlit. Follow the instructions below to set up the environment, run the application, and perform testing using a mock API.

Preparing the Environment Variables

To configure the application, you'll need to set up the following environment variables in your .env file:

CHAT_ENDPOINT=<your-chat-endpoint> Defaults to http://localhost:8000/chatApp

For Feedback Evaluation (Optional)

To collect feedback, you need to set up a Google Sheet and configure the necessary environment variables:

  1. Create a Google Sheet:

    • Give edit access to the sheet to the service account email.
    • Add the sheet ID in the environment variable.
    FEEDBACK_SHEET_ID=your-google-sheet-id

  2. Using Another Worksheet Inside the Google Sheet:

    • If you are using another worksheet inside the Google Sheet, enter the GID as well.
    FEEDBACK_SHEET_GID=your-gid

  3. Set Up Google Cloud Console:

    • Open Google Cloud Console, enable the Sheets API, and create a service account.
    • Give access to this service account email.
    • Export credentials as JSON and add its path in the environment variable.
    GOOGLE_CREDENTIALS_JSON=<path-to-your-google-credentials-json>

  4. Set the Current Version for Feedback Evaluation:

    • Add the current version of the feedback evaluation to the environment variables.
    • If unset, this defaults to the commit hash of master.
    RAG_VERSION=<current-version>

Running the Application

Install Required Packages

Ensure you have the necessary dependencies installed by running:

pip install -r requirements.txt

Run the Streamlit Application

Start the Streamlit application by running:

streamlit run streamlit_app.py

Testing Using Mock API

To test your application using a mock API, you can run the provided mock endpoint script:

python utils/mock_endpoint.py

This will start a mock API server that simulates responses for testing purposes.

License

This project is licensed under the GNU GENERAL PUBLIC LICENSE - see the LICENSE file for details.