Gemma Fine-Tuning Frontend

This is the Streamlit frontend for the Gemma Fine-Tuning Platform. It provides a user-friendly interface for managing the fine-tuning workflow.

Features

• Dataset selection and upload
• Data preprocessing configuration
• Model configuration settings
• Training monitoring and results

Installation

Please note that though the commands given use uv, it's because we use uv to develop the application. You can use equivalent pip commands and the project will work just fine.

uv venv

Running the App

uv run streamlit run src/main.py

Deployment

This frontend is designed to be deployed on Hugging Face Spaces. Simply push this directory to a Hugging Face Space repository.

Configuration

The app connects to backend services for:

• Data preprocessing (CPU-based Cloud Run service)
• Model fine-tuning (GPU-based Cloud Run service)

Backend endpoints can be configured via environment variables or Streamlit secrets.