Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/blinorot/pytorch_project_template
PyTorch Template for DL projects
https://github.com/blinorot/pytorch_project_template
configuration deep-learning deep-learning-template experiment-tracking machine-learning machine-learning-template project project-template python python-template pytorch pytorch-template
Last synced: 9 days ago
JSON representation
PyTorch Template for DL projects
- Host: GitHub
- URL: https://github.com/blinorot/pytorch_project_template
- Owner: Blinorot
- License: mit
- Created: 2024-03-07T17:10:19.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2024-10-12T12:21:02.000Z (4 months ago)
- Last Synced: 2025-01-17T03:06:58.199Z (16 days ago)
- Topics: configuration, deep-learning, deep-learning-template, experiment-tracking, machine-learning, machine-learning-template, project, project-template, python, python-template, pytorch, pytorch-template
- Language: Python
- Homepage:
- Size: 461 KB
- Stars: 69
- Watchers: 2
- Forks: 9
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# PyTorch Template for DL projects
About •
Tutorials •
Examples •
Installation •
How To Use •
Useful Links •
Credits •
License
## About
This repository contains a template for [PyTorch](https://pytorch.org/)-based Deep Learning projects.
The template utilizes different python-dev techniques to improve code readability. Configuration methods enhance reproducibility and experiments control.
The repository is released as a part of the [HSE DLA course](https://github.com/markovka17/dla), however, can easily be adopted for any DL-task.
This template is the official recommended template for the [EPFL CS-433 ML Course](https://www.epfl.ch/labs/mlo/machine-learning-cs-433/).
## Tutorials
This template utilizes experiment tracking techniques, such as [WandB](https://docs.wandb.ai/) and [Comet ML](https://www.comet.com/docs/v2/), and [Hydra](https://hydra.cc/docs/intro/) for the configuration. It also automatically reformats code and conducts several checks via [pre-commit](https://pre-commit.com/). If you are not familiar with these tools, we advise you to look at the tutorials below:
- [Python Dev Tips](https://github.com/ebezzam/python-dev-tips): information about [Git](https://git-scm.com/doc), [pre-commit](https://pre-commit.com/), [Hydra](https://hydra.cc/docs/intro/), and other stuff for better Python code development. The YouTube recording of the workshop is available [here](https://youtu.be/okxaTuBdDuY).
- [Seminar on R&D Coding](https://youtu.be/sEA-Js5ZHxU): Seminar from the [LauzHack Deep Learning Bootcamp](https://github.com/LauzHack/deep-learning-bootcamp/) with template discussion and reasoning. It also explains how to work with [WandB](https://docs.wandb.ai/). The seminar materials can be found [here](https://github.com/LauzHack/deep-learning-bootcamp/blob/main/day03/Seminar_WandB_and_Coding.ipynb).
- [HSE DLA Course Introduction Week](https://github.com/markovka17/dla/tree/2024/week01): combines the two seminars above into one with some updates, including an extra example for [Comet ML](https://www.comet.com/docs/v2/).
- [PyTorch Basics](https://github.com/markovka17/dla/tree/2024/week01/intro_to_pytorch): several notebooks with [PyTorch](https://pytorch.org/docs/stable/index.html) basics and corresponding seminar recordings from the [LauzHack Deep Learning Bootcamp](https://github.com/LauzHack/deep-learning-bootcamp/).
To start working with a template, just click on the `use this template` button.
You can choose any of the branches as a starting point. [Set your choice as the default branch](https://docs.github.com/en/repositories/configuring-branches-and-merges-in-your-repository/managing-branches-in-your-repository/changing-the-default-branch) in the repository settings. You can also [delete unnecessary branches](https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/proposing-changes-to-your-work-with-pull-requests/creating-and-deleting-branches-within-your-repository).
## Examples
> [!IMPORTANT]
> The main branch leaves some of the code parts empty or fills them with dummy examples, showing just the base structure. The final users can add code required for their own tasks.
You can find examples of this template completed for different tasks in other branches:
- [Image classification](https://github.com/Blinorot/pytorch_project_template/tree/example/image-classification): simple classification problem on [MNIST](https://yann.lecun.com/exdb/mnist/) and [CIFAR-10](https://www.cs.toronto.edu/~kriz/cifar.html) datasets.
- [ASR](https://github.com/Blinorot/pytorch_project_template/tree/example/asr): template for the automatic speech recognition (ASR) task. Some of the parts (for example, `collate_fn` and beam search for `text_encoder`) are missing for studying purposes of [HSE DLA course](https://github.com/markovka17/dla).
## Installation
Installation may depend on your task. The general steps are the following:
0. (Optional) Create and activate new environment using [`conda`](https://conda.io/projects/conda/en/latest/user-guide/getting-started.html) or `venv` ([`+pyenv`](https://github.com/pyenv/pyenv)).
a. `conda` version:
```bash
# create env
conda create -n project_env python=PYTHON_VERSION
# activate env
conda activate project_env
```
b. `venv` (`+pyenv`) version:
```bash
# create env
~/.pyenv/versions/PYTHON_VERSION/bin/python3 -m venv project_env
# alternatively, using default python version
python3 -m venv project_env
# activate env
source project_env
```
1. Install all required packages
```bash
pip install -r requirements.txt
```
2. Install `pre-commit`:
```bash
pre-commit install
```
## How To Use
To train a model, run the following command:
```bash
python3 train.py -cn=CONFIG_NAME HYDRA_CONFIG_ARGUMENTS
```
Where `CONFIG_NAME` is a config from `src/configs` and `HYDRA_CONFIG_ARGUMENTS` are optional arguments.
To run inference (evaluate the model or save predictions):
```bash
python3 inference.py HYDRA_CONFIG_ARGUMENTS
```
## Useful Links:
You may find the following links useful:
- [Report branch](https://github.com/Blinorot/pytorch_project_template/tree/report): Guidelines for writing a scientific report/paper (with an emphasis on DL projects).
- [CLAIRE Template](https://github.com/CLAIRE-Labo/python-ml-research-template): additional template by [EPFL CLAIRE Laboratory](https://www.epfl.ch/labs/claire/) that can be combined with ours to enhance experiments reproducibility via [Docker](https://www.docker.com/).
- [Mamba](https://github.com/mamba-org/mamba) and [Poetry](https://python-poetry.org/): alternatives to [Conda](https://conda.io/projects/conda/en/latest/user-guide/getting-started.html) and [pip](https://pip.pypa.io/en/stable/installation/) package managers given above.
- [Awesome README](https://github.com/matiassingers/awesome-readme): a list of awesome README files for inspiration. Check the basics [here](https://github.com/PurpleBooth/a-good-readme-template).
## Credits
This repository is based on a heavily modified fork of [pytorch-template](https://github.com/victoresque/pytorch-template) and [asr_project_template](https://github.com/WrathOfGrapes/asr_project_template) repositories.
## License
[](/LICENSE)