My First AI Code Agent with Smolagents 🤖

Overview

This project is my first introduction to the world of AI agents, specifically focusing on building a code agent using the smolagents framework. So far, I have installed the necessary libraries, set up access to the Hugging Face API, created a CodeAgent with a specific language model, and attempted to implement and use a custom tool. The initial goal is to create a basic agent that can utilise custom tools to perform specific tasks, starting with a simple tool to get the current time. This project is a learning experience to build a foundational understanding before exploring more advanced agent capabilities.

Smolagents Framework 🤖

Smolagents is a lightweight framework for building AI agents that focus on code generation.

How to Run

1. Open in Google Colab: Click the "Open In Colab" badge above or upload and open the notebook in Google Colab.
2. Install Dependencies: Run the cell containing pip install 'smolagents[toolkit]' to install the required libraries.
3. Set up Hugging Face API Token:
   • Obtain a Hugging Face API token with permissions for Inference Providers.
   • In Google Colab, go to the "🔑" icon (Secrets) in the left sidebar.
   • Add a new secret with the name HF_TOKEN and paste your Hugging Face token as the value.
   • Run the cell that retrieves the HF_TOKEN from userdata.
4. Create the Code Agent: Run the cell that defines and initializes the CodeAgent with the chosen model.
5. Define Custom Tool: Run the cell that defines the get_current_time custom tool.
6. Run the Task: Execute the cell that calls agent.run() with the task you want the agent to perform.