Spaces:
Sleeping
Sleeping
A newer version of the Gradio SDK is available: 6.26.0
metadata
title: GAIA Agent
emoji: 🤖
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 5.23.1
app_file: app.py
pinned: false
hf_oauth: true
tags:
- smolagents
- agent
- smolagent
- tool
- agent-course
- gaia
GAIA Benchmark Agent
A smolagents-based CodeAgent designed to pass the GAIA Level 1 benchmark for the HuggingFace Agents Course (Unit 4 Final Assignment).
Features
- DuckDuckGo Web Search — Real-time information retrieval
- Webpage Visitor — Deep-reads any URL as markdown
- Code Execution — Runs Python code for computation and reasoning
- Planning — Periodic re-planning for multi-step tasks
- Exact Answer Extraction — Outputs only the final answer value
Usage
- Log in with your HuggingFace account.
- Click Run Evaluation & Submit All Answers.
- Wait for the agent to process all 20 questions.
- View your score!
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference