deeplearning-models
A collection of Jupyter notebooks showing many deep learning models in TensorFlow and PyTorch.
- Stars
- 17.6k
- Forks
- 4.1k
- Licence
- MIT
- Language
- Jupyter Notebook
- Last update
- 1 months ago
- Type
- Learning resource
What it is
This repository is a set of Jupyter notebooks, each implementing a specific deep learning model, from simple perceptrons and multilayer networks to convolutional architectures such as LeNet, AlexNet, VGG, DenseNet and MobileNet. The README says the notebooks are written for TensorFlow and PyTorch and are organized by topic, so people can open one and see how a given architecture is built and trained. It reads as a study and reference resource rather than a finished application.
- Open the LeNet-5 notebook on MNIST to see a small convolutional network trained end to end
- Compare the PyTorch and TensorFlow versions of a multilayer perceptron on MNIST
- Study the from-scratch backpropagation notebook to see how gradients are computed by hand
- Adapt the MobileNet-v3 CIFAR-10 notebook as a starting point for your own image classifier
Great for
- Students learning how classic neural network architectures are built
- Developers who want working PyTorch notebook examples to adapt
- Readers comparing the same model implemented in PyTorch and TensorFlow
Not ideal for
- People looking for a ready-made application or production library, since the README describes a collection of notebooks
- Readers who want detailed written explanations for each model, since most table descriptions in the README are marked TBD
How to get it
- 1
Clone and build from source
You'll need: Jupyter Notebook · Python
Clone the repository
git clone https://github.com/rasbt/deeplearning-models - 2
Run it
- Clone the repository to your machine
- Open the repository folder in Jupyter Notebook
- Pick a topic folder such as pytorch_ipynb/cnn and open a notebook
- Run the notebook cells in order
- 3
Get the most out of it
- Use the PyTorch Lightning versions of CNN notebooks to compare a higher-level training style with plain PyTorch
- Read the Traditional Machine Learning section first to build intuition before the CNN models
- Compare the same model across the PyTorch and TensorFlow notebook folders
Downloads
Strengths
- Covers many architectures across basic ML, MLPs, and CNNs in one place
- Provides both PyTorch and TensorFlow versions for several models
- Released under the MIT license, which the metadata lists
Trade-offs
- Most model descriptions in the README are placeholders marked TBD, so the notebooks must be read to learn their purpose
- The README excerpt provided is truncated, so the full scope is not visible here
- Install and run instructions are not present in the supplied README text
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.