============================= Cookiecutter Pytorch Template

A Cookiecutter <https://github.com/audreyr/cookiecutter>_ template for PyTorch projects.

.. contents:: Table of Contents :depth: 2

Requirements

• Python >= 3.8
• PyTorch >= 1.1
• Tensorboard >= 1.4

Features

• Clear folder structure which is suitable for many deep learning projects.
• Runs are configured via .yml files allowing for easy experimentation.
• Checkpoint saving and resuming.
• Tensorboard logging

Usage

.. code::

$ pip install cookiecutter
$ cookiecutter https://github.com/khornlund/cookiecutter-pytorch
$ cd path/to/repo

A template project has now been created! You can run the MNIST example using:

.. code::

$ conda env create --file environment.yml
$ conda activate <your-project-name>
$ <your-project-name> train

Example Projects

Here are some projects which use this template:

1. Severstal Steel Defect Detection (Kaggle) <https://github.com/khornlund/severstal-steel-defect-detection>_
2. Aptos Blindness Detection (Kaggle) <https://github.com/khornlund/aptos2019-blindness-detection>_
3. Understanding Cloud Organization (Kaggle) <https://github.com/khornlund/understanding-cloud-organization>_

Custom Defaults

If you fork this repo, you can modify cookiecutter.json to provide personalised defaults eg. name, email, username, etc.

Acknowledgements

This template was based on PyTorch Template <https://github.com/victoresque/pytorch-template>_.