This repository provides a comprehensive collection of Jupyter notebooks and code examples for deep learning courses, specifically tailored for the "dataflowr" curriculum at École Polytechnique. It aims to offer hands-on, practical experience for students learning deep learning concepts, from foundational PyTorch mechanics to advanced architectures like Transformers and Diffusion Models.

How It Works

The project is structured around a series of modules, each covering a specific deep learning topic. Notebooks demonstrate key concepts, provide implementations from scratch (e.g., MLPs, backpropagation), and showcase the application of pre-trained models and advanced techniques. The approach emphasizes practical coding alongside theoretical explanations, using PyTorch as the primary framework.

Quick Start & Requirements

• Installation: Follow instructions in "Module 0 - Running the notebooks locally".
• Prerequisites: PyTorch, Python. Specific modules may require additional libraries as indicated within the notebooks.
• Resources: GPU recommended for most deep learning tasks.
• Links: Module 0

Highlighted Details

• Covers a wide range of deep learning topics, from basic tensor operations to advanced models like GANs, Transformers, and Diffusion Models.
• Includes implementations from scratch for core concepts like backpropagation and MLPs.
• Features practical applications such as recommender systems, word embeddings, and image segmentation.
• Offers solutions for many practical exercises and homework assignments.

Maintenance & Community

The repository appears to be associated with academic courses, with content reflecting a 2023 schedule. Specific contributor or community activity details are not prominent in the README.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README.

Limitations & Caveats

The README does not specify a license, which may impact commercial use or redistribution. Some notebooks are marked as "empty" or "empty.ipynb", indicating they are templates requiring completion. The content is tied to a specific course structure and schedule.