Alegant: a elegant training framework
Project description
Alegant
Alegant is an elegant training framework for PyTorch models.
Install Alegant
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 fairseq and develop locally:
python setup.py develop
Usage
To use alegant, follow the steps below:
- Define your Model.
- Define your DataModule.
- Define your Trainer.
- Run the training script using the following command:
python --config_file run.py
Make sure to replace config_file with the path to your configuration file.
Configuration
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.
Project Structure
alegant
├── tensorboard
├── data
├── alegant
│ ├── data_module.py
│ ├── trainer.py
│ └── utils.py
├── src
│ ├── dataset.py
│ ├── loss.py
│ ├── model
│ │ ├── modeling.py
│ │ ├── poolers.py
│ ├── trainer.py
│ └── utils.py
├── config.yaml
├── run.py
└── setup.py
Contact
If you have any questions or inquiries, please contact us at zhuhh17@qq.com
Thank you for using Alegant! Happy training!
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
