Neura Market
Agents Course logo

Agents Course

Free

This repository contains the Hugging Face Agents Course.

FreeFree tier
Type
Open Source
Company
Hugging Face
Visit Website GitHub (28.8k stars)

About Agents Course

The Hugging Face Agents Course is a free, open-source, structured curriculum that teaches learners how to build AI agents. Hosted on GitHub under the Apache-2.0 license, the course is divided into 4 main units covering the fundamentals of agents, large language models (LLMs), and practical frameworks such as smolagents, LangGraph, and LlamaIndex. It includes bonus units on fine-tuning LLMs for function calling, observability, evaluation, and even a playful Pokémon module. The course culminates in a final project where students create, test, and certify their own agent, with automated evaluation and a leaderboard. Prerequisites are basic knowledge of Python and LLMs. The course encourages community contributions and discussions via Discord.

Key Features

Structured 4-unit curriculum covering agents from basics to production
Hands-on with multiple frameworks: smolagents, LlamaIndex, LangGraph
Bonus units on fine-tuning LLMs for function calling, observability, and evaluation
Final project with automated evaluation and leaderboard
Community contributions and discussions via Discord
Available for free on GitHub with Apache-2.0 license

Pros & Cons

Pros
  • Completely free and open-source with Apache-2.0 license
  • Covers multiple leading agent frameworks (smolagents, LangGraph, LlamaIndex)
  • Includes practical bonus topics like function-calling fine-tuning and observability
  • Final project with automated evaluation and leaderboard provides tangible outcome
  • Active community and contribution opportunities via GitHub and Discord
Cons
  • Requires basic Python and LLM knowledge; not suitable for complete beginners
  • Content is primarily text-based on GitHub (not video or interactive), which may not suit all learning styles
  • Limited to Hugging Face ecosystem focus; may not cover other agent libraries in depth

Best For

Learning AI agent development from fundamentals to deploymentExploring practical frameworks for building LLM-powered agentsStudying agentic RAG and observability/evaluation techniquesGaining hands-on experience through a final certification project

FAQ

What is the Hugging Face Agents Course?
It is a free, open-source course that teaches how to build AI agents, covering fundamentals, LLMs, and frameworks like smolagents, LangGraph, and LlamaIndex.
Who is the course intended for?
Learners with basic knowledge of Python and LLMs who want to understand and build AI agents.
What topics are covered in the course?
The course includes units on introduction to agents, fine-tuning LLMs for function calling, frameworks (smolagents, LlamaIndex, LangGraph), observability and evaluation, agentic RAG, and a final project.
Is the course really free?
Yes, the course is entirely free and hosted on GitHub under the Apache-2.0 license. You can sign up at https://bit.ly/hf-learn-agents.
How can I contribute to the course?
You can contribute by fixing typos/grammar via pull requests, or by proposing a new unit through an issue on the GitHub repository.
What is the final project?
Unit 4 is a final project where you create, test, and certify your own agent, with automated evaluation and a leaderboard.