PyTorch framework for Deep Learning research and development.
It was developed with a focus on reproducibility,
fast experimentation and code/ideas reusing.
Being able to research/develop something new,
rather than write another regular train loop.
Break the cycle - use the Catalyst!
Project manifest. Part of PyTorch Ecosystem. Part of Catalyst Ecosystem:
- Alchemy - Experiments logging & visualization
- Catalyst - Accelerated Deep Learning Research and Development
- Reaction - Convenient Deep Learning models serving
Catalyst.Gan [WIP]
You will learn how to train your GAN using the Catalyst framework. The main advantage is to customize your experiments in the yaml config instead of the code.
Installation
pip install -r requirements.txt
Run examples
MNIST
# (Goodfellow et. al., 2014: https://arxiv.org/pdf/1406.2661.pdf) catalyst-dl run -C examples/mnist/configs/vanilla_gan.yml # (Arjovsky et. al., 2017: https://arxiv.org/abs/1701.07875) catalyst-dl run -C examples/mnist/configs/wasserstein_gan.yml # (Gulrahani et. al., 2017: https://arxiv.org/abs/1704.00028) catalyst-dl run -C examples/mnist/configs/wasserstein_gan_gp.yml # (Mirza and Osindero, 2014: https://arxiv.org/abs/1411.1784) catalyst-dl run -C examples/mnist/configs/conditional_gan.yml
Advanced [under construction]
If you want to try right now run from console
(you should download FFHQ dataset before that and specify path in examples/advanced/tconfigs/data/FFHQ.yml)
./examples/advanced/experiments_setup/run.sh