Getting Started with Smolagents: Build Your First Code Agent in 15 Minutes
In the rapidly evolving world of artificial intelligence, the ability to create autonomous agents has become increasingly accessible. Hugging Face’s smolagents library is at the forefront of this movement, allowing developers and enthusiasts to build intelligent agents with ease. In this article, we will guide you through the process of creating a simple AI weather agent in just 40 lines of Python code. This hands-on tutorial will cover essential concepts such as creating tools, connecting language models, and running autonomous tasks.
What Are Smolagents?
Smolagents are lightweight, flexible code agents designed to interact with the world and perform tasks autonomously. They can be utilized in various applications, from simple data retrieval to complex decision-making processes. The smolagents library simplifies the integration of AI models and tools, allowing developers to focus on creating effective solutions rather than dealing with complex frameworks.
Prerequisites
Before diving into the code, ensure you have the following:
- Python 3.7 or higher installed on your machine
- A basic understanding of Python programming
- A Hugging Face account to access the smolagents library
- Internet access to retrieve weather data
Setting Up Your Environment
Start by installing the smolagents library along with any dependencies. You can do this using pip:
pip install smolagentsOnce the installation is complete, you are ready to start coding your weather agent.
Building the Weather Agent
In this section, we will create a simple weather agent that fetches current weather information based on user input. Below is the complete code:
from smolagents import Agent, Tool
class WeatherTool(Tool):
def __init__(self):
super().__init__(name="WeatherTool", description="Fetches current weather data.")
def run(self, location):
import requests
api_key = "your_api_key_here"
url = f"http://api.openweathermap.org/data/2.5/weather?q={location}&appid={api_key}&units=metric"
response = requests.get(url)
data = response.json()
if response.status_code == 200:
return f"The current temperature in {location} is {data['main']['temp']}°C."
else:
return "Sorry, I couldn't fetch the weather data."
agent = Agent(tools=[WeatherTool()])
response = agent.run("London")
print(response)
Explanation of the Code
The code consists of two main components: the WeatherTool class and the Agent instance. The WeatherTool class is responsible for fetching weather data using the OpenWeatherMap API. You need to replace your_api_key_here with your actual API key.
The Agent instance is created with the WeatherTool as its tool. When the agent runs with a specified location, it returns the current temperature in that area.
Running Your Agent
To execute your weather agent, simply run the Python script. You should see the current temperature of the specified location printed in your console. Feel free to modify the location to test it with different cities.
Conclusion
Congratulations! You have successfully built your first AI weather agent using the smolagents library in just 15 minutes. This simple project demonstrates the power of smolagents in creating autonomous code agents that can perform useful tasks. With this foundation, you can explore more complex applications and enhance your skills in AI development.