---
title: Smolagents2
emoji: ⚡
colorFrom: yellow
colorTo: green
sdk: gradio
sdk_version: 6.3.0
app_file: app.py
pinned: false
---
---
title: Smolagents
emoji: 📚
colorFrom: indigo
colorTo: indigo
sdk: gradio
sdk_version: 6.3.0
app_file: app.py
pinned: false
---
# Alfred AI Assistant - smolagents Framework

A multi-functional AI assistant built with the **smolagents framework** that can:
- 🔍 Search the web for current information
- ⏰ Tell time in any timezone
- 🧮 Perform calculations and currency conversions
- 🌤️ Check weather anywhere
- 📏 Convert between units
- Translator

## 🚀 Features

- **7+ Specialized Tools**: Each optimized for different tasks
- **Real-time Data**: Web search, weather, currency rates
- **Smart Tool Selection**: Agent automatically chooses the right tool
- **User-friendly Interface**: Clean Gradio UI
- **Educational**: Perfect for learning smolagents framework

## 📋 Available Tools

1. **web_search_tool** - Search DuckDuckGo for current information
2. **get_current_time_in_timezone** - Time in any timezone
3. **calculator_tool** - Math + currency conversion
4. **weather_tool** - Current weather worldwide
5. **unit_conversion_tool** - Convert units (length, weight, temp, etc.)

## 🎯 Example Queries

- "What are the latest AI news?"
- "What time is it in Tokyo?"
- "Calculate 15% of 250 + tax"
- "Convert 100 USD to EUR"
- "Weather in Paris"
- "10 km to miles"

## 🛠️ Technical Details

- **Framework**: smolagents
- **LLM**: Qwen/Qwen2.5-Coder-32B-Instruct (via Hugging Face)
- **UI**: Gradio
- **Deployment**: Hugging Face Spaces

## 📁 Files

- `app.py` - Main application with agent and tools
- `prompts.yaml` - System prompts and configuration
- `requirements.txt` - Python dependencies

## 🚀 Deployment

1. Create a new Space on Hugging Face
2. Choose **Gradio** SDK
3. Upload all files
4. The Space will build automatically

## 📚 Course Context

This project is for the HuggingFace AI Agents Course, demonstrating the **smolagents framework**.

## ⚠️ Limitations

- Web search limited to 3 results (for speed)
- Some APIs have rate limits
- Image generation requires specific hardware
- Model may be overloaded during peak times
  
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference