Learn PyTorch for Deep Learning
Free course materials (notebooks and slides) for learning PyTorch and deep learning from scratch.mrdbourke/pytorch-deep-learning
- Stars
- 19.3k
- Forks
- 5.2k
- Licence
- MIT
- Language
- Jupyter Notebook
- Last update
- 1 months ago
- Type
- Learning resource
What it is
This repository holds the code notebooks, slides, and exercises for the Zero to Mastery Learn PyTorch for Deep Learning course. It walks through PyTorch basics, neural network classification, computer vision, transfer learning, experiment tracking, paper replication, and model deployment. The full course is also available as an online book at learnpytorch.io.
- Work through the PyTorch Fundamentals notebook to learn tensors and basic operations
- Build and train a neural network classification model following section 02
- Adapt a pre-trained model to your own image problem using the transfer learning section
- Turn notebook code into Python scripts using the going-modular section
Great for
- Beginners in machine learning who want to learn PyTorch in a hands-on, code-first way
- Python programmers with some coding experience who are new to deep learning
- Students who prefer working through notebooks and exercises
- People who want a structured path from basics to a deployed model
Not ideal for
- Experienced ML practitioners, since the README says the course is designed to be beginner-friendly and may offer little new material
- People who need a non-PyTorch framework, because the course focuses only on PyTorch
- Anyone looking for a software library to install and use in a product, since this is teaching material rather than a package
How to get it
- 1
Clone and build from source
You'll need: Python (README suggests 3-6 months of coding experience) · Jupyter Notebook or Google Colab
Clone the repository
git clone https://github.com/mrdbourke/pytorch-deep-learning - 2
Run it
- Clone the repository to your computer
- Open the notebooks in Jupyter Notebook or Google Colab
- Start with section 00 - PyTorch Fundamentals
- Work through the exercises listed for each section
- 3
Get the most out of it
- Read the online book at learnpytorch.io alongside the notebooks for the same material in readable form
- Use the PyTorch Cheatsheet and PyTorch Extra Resources pages linked in the README
- Watch the first five sections on YouTube, as the README links to them
- Ask questions on the GitHub Discussions page
Downloads
Strengths
- Free, with all course materials available online and in this repository under MIT license
- Organized as a numbered sequence of sections with matching slides and exercises
- Code-first, hands-on approach with exercises and extra-curriculum in each section
Trade-offs
- The README is truncated and does not list exact hardware or version requirements beyond the PyTorch 2.0 note
- Content is tied to a specific course and the README states the course is beginner-oriented, so advanced learners may find it too basic
- Relies on Jupyter Notebook or Google Colab familiarity, which the README lists as a prerequisite
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| pytorch-deep-learning (this) | 19.3k | Learning resource | beginner | 1 months ago | Beginners in machine learning who want to learn PyTorch in a hands-on, code-first way |
| annotated_deep_learning_paper_implementations | 67.6k | Learning resource | intermediate | 4 days ago | Students learning how transformers, GANs, or optimizers work internally |
| Made-With-ML | 49.9k | Learning resource | intermediate | 8 months ago | Software engineers who want to learn how ML models get deployed and maintained |
| LaTeX-OCR | 16.6k | AI model / agent | intermediate | 1.8 years ago | Students who want to turn a handwritten or printed formula photo into editable LaTeX |
| onnxruntime | 22.2k | Library | intermediate | 4 days ago | Developers who want to deploy a trained model inside an application or service |
| nni | 14.4k | Library | intermediate | 2.3 years ago | Researchers comparing hyperparameter tuning algorithms such as TPE, Hyperband, or Bayesian methods |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.