Learn Hugging Face Bootcamp review

What it is

Learn Hugging Face Bootcamp is Jose Portilla's hands-on course on the Hugging Face ecosystem, the open-source hub where most public ML models actually live. It runs 9.5 hours across 8 sections and 49 lectures, and holds a 4.3 rating from 764 ratings. On AcademyGems's Udemy courses list it fills a specific gap: every other pick teaches you to use AI through somebody's API, and this one teaches you to run the models yourself.

Who teaches it

Jose Portilla is Head of Data Science at Pierian Training and one of the longest-running, highest-volume data science instructors on Udemy, with a large catalogue spanning Python, machine learning and data analysis. The teaching style is the one his Python courses are known for: notebook-driven, build-as-you-go, light on theory.

What is inside

The course is organised by model type rather than by tool, which is what makes it useful:

• Introduction to Hugging Face: models, datasets and spaces, plus environment setup
• NLP with the Transformers library: text classification, entity recognition and related tasks
• Image generation with the Diffusers library
• Video models
• Audio models, the section most competing courses skip entirely
• Building machine learning GUIs with Gradio, so a model becomes something a person can actually use
• An appendix on Git

The requirements are stated plainly on the course page and are worth taking seriously: Python and pip, PyTorch and CUDA, and Jupyter notebooks. This is not a no-code course.

How it compares

• vs AI Engineer Core Track: Ed Donner's course is five times longer and aimed at shipping LLM products, with RAG, QLoRA and agents. This one is narrower and lower-level, focused on the open-source model ecosystem itself, including image, video and audio. If you want to build an AI product, take the Core Track; if you want to understand and run open models, take this.
• vs ChatGPT & Generative AI: The Complete Guide: Academind's course touches running open LLMs locally as one section among many. This is that idea expanded into a full course, across every modality.
• vs Zero To Mastery: ZTM's Machine Learning with Hugging Face Bootcamp covers similar ground for roughly $25/month as part of a subscription. If you only want this one topic, the standalone Udemy purchase is cheaper; if you want the wider catalogue, ZTM wins.

Who it is for

Python developers and data scientists who are comfortable in a notebook and want to move past API calls into open-source models they can host, fine-tune and wrap in an interface. The Gradio section is the practical payoff: it turns a model into something you can hand to someone.

Who should skip it

Non-technical learners should skip it outright. The prerequisites are real, and PyTorch plus CUDA in particular means you need to be comfortable with the ML tooling stack before you start. The bigger caveat is the date: this course was last updated in May 2024. The Hugging Face libraries move quickly, so expect some API drift and dated model choices, and budget time for reading current docs alongside the videos. If you want a course that reflects the state of the field this year, a more recently refreshed pick will serve you better.

Pricing, and the one rule

It was listed around €80 and discounted to roughly €18 when AcademyGems last checked, with regional pricing and near-constant sales, so treat both figures as approximate. The usual Udemy rule applies: never pay list price, and lean on the 30-day refund. The extra rule here, given the 2024 update date, is to check the course page's own "last updated" stamp before buying, because if it has not moved, the drift will only have grown.

First-look verdict

The most technically distinctive course on this list, and the only one that treats image, video and audio models as first-class rather than an afterthought. On sale it is worth it for a Python developer who wants the open-source side of AI. The stale update date is the honest catch: buy it for the concepts and the ecosystem map, expect to patch the specifics yourself.