README.md

metadata

title: Template Final Assignment
emoji: 🕵🏻‍♂️
colorFrom: indigo
colorTo: indigo
sdk: gradio
sdk_version: 5.25.2
app_file: app.py
pinned: false
hf_oauth: true
hf_oauth_expiration_minutes: 480

GAIA Agent Evaluation Runner

Small Gradio app for the Hugging Face Agents Course final assignment.

The app logs in with Hugging Face OAuth, downloads the GAIA evaluation questions, runs a smolagents agent on each question, and submits the answers to the scoring endpoint.

Setup

Create a .env file with the secrets needed by the model provider and by Hugging Face Spaces:

GROQ_API_KEY=your_groq_key_here
SPACE_ID=your-username/your-space-name

Install the dependencies:

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Run locally:

python app.py

Notes

• The app uses https://agents-course-unit4-scoring.hf.space as the scoring API.
• Text answers use Groq llama-3.3-70b-versatile; audio files use Groq Whisper; image files use a Groq vision model.
• The Gradio SDK version is pinned in this README frontmatter and dependencies are pinned in requirements.txt.
• OAuth must be enabled on the Hugging Face Space for the login flow to work.