Awesome

⚡ TmPL: Template for Pytorch Lightning

Lightning Docs  •  Installation  •  Usage  •  Reference  •  Contributing  •  License

A template for rapid & flexible DL experimentation development, built upon Lightning & Hydra with best practice.

What's New

v2.0 was released on Sep 5 '23.

Installation

pip install -r requirements.txt

It is recommended to manually install PyTorch and Torchvision before running the installation command, referring to the official PyTorch website for instructions.

Usage

0. Setup

   export PYTHONPATH=`pwd`:$PYTHONPATH
   

1. Training

   python tools/train.py [--config-name config[.yaml]] [trainer.devices=4] [data.loader.batch_size=16]
   

   • Override the default config file with --config-name.
   • You can also override any value in the loaded config from the command line, refer to the following for more infomation.
     • https://hydra.cc/docs/tutorials/basic/your_first_app/config_file/
     • https://hydra.cc/docs/advanced/hydra-command-line-flags/
     • https://hydra.cc/docs/advanced/override_grammar/basic/

2. Tips for Further Development

   The code is designed to be flexible and customizable to meet your specific needs.
   Useful comments can be found in the source code.

Reference

• PyTorch Lightning Docs ↗
• Hydra Docs ↗

Contributing

Contributions are welcome and appreciated!
Feel free to open an issue or PR! 🎉

License

Released under the MIT License.