Lab — Publish a Gemma model API on HuggingFace Spaces with Google Auth#

Objective#

Deploy a text generation API using a local LLM framework (like Ollama or transformers) to Hugging Face Spaces, and protect the endpoint using Google OAuth.

Requirements#

1. Create a FastAPI application that serves a generation endpoint (e.g., /generate).
2. Integrate Google OAuth 2.0 so that only authenticated users can access the endpoint.
3. Containerize the application using Docker.
4. Deploy the Docker container to a Hugging Face Space.

Deliverables#

• Link to your Hugging Face Space.
• The Dockerfile and main.py code.
• A brief explanation of how you configured the OAuth credentials.