Playing with smolagents

Este repositorio contiene un script de Python que implementa un sistema multiagente. Utiliza varios modelos y herramientas El script aprovecha los modelos de Hugging Face y el modelo LiteLLM de OpenAI para crear agentes capaces de ejecutar codigo creado por el mismo agente.

Features

• Hugging Face Integration: Utilizes Hugging Face's CodeAgent and HfApiModel for executing code-related tasks.
• OpenAI Integration: Implements a proxy for OpenAI's LiteLLM model to handle requests.
• Custom Tool Creation: Demonstrates how to create and use custom tools with the @tool decorator.
• Web Scraping: Includes an example of an agent that can scrape and return data from a webpage in JSON format.
• CodeAgent: Test
• ToolCallingAgent: Test
• Search Functionality: Integrates a search tool to fetch information from DuckDuckGo.

How?

1. UV package manager: UV install

2. Prepare environment:

   uv venv --python 3.11 && source .venv/bin/activate && uv pip install -r requirements.txt

3. Install the required packages:

   uv pip install -r requirements.txt

4. Create a .env file in the root directory and add your API tokens:

   HF_API_TOKEN=your_hugging_face_api_token
   OPENAI_API_KEY=your_openai_api_key
   

Usage

First Telemetry:

uv run python -m phoenix.server.main serve

then

Run the script using:

uv run main.py

memory loop is wrong driven .