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
short_description: Gemini Agent for GAIA Evaluation

🧠 Gemini Agent for GAIA Evaluation

This project contains a Gemini-powered CodeAgent built with smolagents for use in the GAIA Unit 4 Evaluation on Hugging Face Spaces.

🧑‍🏫 This app is part of the Hugging Face Courses series and was developed as an educational project to showcase how machine learning can be used in real-world event scenarios. Learn more at Hugging Face Courses.

---

🚀 Features

• Uses the Gemini 2.0 Flash model via LiteLLMModel

• Equipped with essential tools:

  • DuckDuckGoSearchTool for quick lookups
  • RunPythonFileTool for executing .py scripts
  • ReverseTextTool for decoding reversed questions
  • download_server for fetching files from URLs
  • Base tools (math, string manipulation, etc.)

📋 Evaluation Strategy

The agent reads questions from the GAIA evaluation endpoint, applies reasoning using a system prompt with strict guidelines, and submits answers back for scoring.

🛠️ Setup

1. Clone this repository or Space

2. Set your environment variables:

   GEMINI_API_KEY=your_api_key_here
   SPACE_ID=your_hf_space_id
   

3. Install dependencies:

   pip install -r requirements.txt
   

4. Run locally:

   python app.py
   

Or launch directly via Hugging Face Spaces.

🧪 Evaluation Flow

1. Log in to Hugging Face through the UI
2. Click “Run Evaluation & Submit All Answers”
3. The agent will fetch tasks, solve them, and submit results

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference