Alegant is an elegant training framework for PyTorch models.
Before installing Alegant, please make sure you have the following requirements:
- Python >= 3.7
- torch >= 1.9
Simple installation from PyPI
pip install alegant
To install Alegant and develop locally:
python setup.py develop
For examples on how to use elegant, please refer to the examples directory in this repository. It contains sample configuration files and code snippets to help you get started.
To use Alegant, follow the steps below:
- Define your Model.
- Define your DataModule.
- Define your Trainer.
- Set your configuration.
- Run the training script using the following command:
cd alegant/example
python example_main.py # for simply use the DataModule and Trainer
python example_runner.py # for using the runner from Alegant
To customize the training process, you need to provide a configuration file. This file specifies various parameters such as dataset paths, model architecture, hyperparameters, etc. Make sure to create a valid configuration file before running the framework.
alegant
├── alegant
│ ├── __init__.py
│ ├── runner.py
│ ├── trainer.py
│ ├── data_module.py
│ ├── utils.py
│ └── example
│ ├── data
│ ├── config.yaml
│ ├── example_main.py
│ ├── example_runner.py
│ ├── logs
│ ├── README.md
│ ├── src
│ │ ├── dataset.py
│ │ ├── loss.py
│ │ ├── model
│ │ │ ├── modeling.py
│ │ │ ├── poolers.py
│ │ ├── trainer.py
│ │ └── utils.py
│ └── tensorboard
└── setup.py
If you have any questions or inquiries, please contact us at zhuhh17@qq.com
Thank you for using Alegant! Happy training!