PyTorch Deep Learning Course Materials

Note: This course was originally recorded for Wanmen University, which is no longer available. The materials are open-sourced here for public benefit. The course content is adapted and rewritten from Dive into Deep Learning, with a focus on PyTorch implementations and hands-on projects.

📚 Course Introduction

The PyTorch Deep Learning Course focuses on using the PyTorch framework, avoiding heavy mathematical derivations, and instead emphasizes hands-on projects to build intuition and skills.

Key Features

• Theory and Practice Combined | Combines theoretical knowledge with practical implementation
• Practical Coding Focus | Emphasizes real-world code applications
• Python Prerequisites | Designed for learners with basic Python knowledge

🎥 Video Playlist: Bilibili - PyTorch Deep Learning Course

🗂️ Course Outline

Lecture 1: Introduction to Deep Learning

• Historical background and applications
• Key features of deep learning

Lecture 2: PyTorch Basics

• Environment setup and execution
• Tensors and automatic differentiation

Lecture 3: Deep Learning Fundamentals

• Linear regression and Softmax regression
• Hands-on Practice: Image classification (Fashion-MNIST)

Lecture 4: Building Models in PyTorch

• Custom model creation and parameter management
• GPU utilization methods

Lecture 5: Convolutional Neural Networks

• Convolution operations and network architectures
• Hands-on Practice: CIFAR-10 image classification

Lecture 6: Recurrent Neural Networks

• RNN, LSTM, GRU architectures
• Hands-on Practice: Lyric generation and text sequence modeling

Lecture 7: Optimization Algorithms

• Gradient descent and improved methods
• Learning rate scheduling strategies

Lecture 8: Performance Optimization

• Imperative and symbolic programming integration
• Multi-GPU parallel computing

Lecture 9: Computer Vision

• Image augmentation and transfer learning
• Hands-on Practice: Pikachu object detection

Lecture 10: Natural Language Processing

• Word embedding (Word2Vec)
• Hands-on Practice: Text sentiment classification

Lecture 11: Generative Adversarial Networks

• GAN principles and variants
• Hands-on Practice: Pokemon dataset generation

📂 Repository Content

• ppt/ — Lecture slides and presentations
• notebooks/ — Jupyter notebooks with code and experiments

🛠️ Technical Stack

• Framework: PyTorch
• Language: Python
• Notebooks: Jupyter
• Topics: Deep Learning, Machine Learning, Computer Vision, NLP, GANs

🎯 Learning Objectives

By completing this course, students will:

• Master PyTorch framework fundamentals
• Understand core deep learning concepts
• Implement various neural network architectures
• Apply deep learning to real-world problems
• Gain hands-on experience with computer vision and NLP tasks

📊 Course Statistics

• Total Lectures: 11 comprehensive sessions
• Hands-on Projects: 6 practical implementations
• Coverage: From basics to advanced topics
• Language: Primarily English with Chinese support

🙏 Acknowledgments

Thanks to everyone interested in and supporting this course. We hope this repository helps more learners quickly get started and go deeper with PyTorch deep learning.

🔗 Links

• GitHub Repository: wanmen-pytorch-course-materials
• Video Course: Bilibili Playlist
• Original Source: Dive into Deep Learning

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.