Main-Branch:
Develop-Branch:
DeepSaki is an add-on to TensorFlow. It provides a variaty of custom classes ranging from activation functions to entire models, helper functions to facilitate connectiong to your, compute HW and many more!
The project started as fun project to learn and to collect the code snippets I was using in my projects. Now it has been transformed into a modern SW package featuring CI/CD and a documentation.
🎖️ Some highlights:
- Layers to transform data into the frequency domain using FFTs, like FFT2D and FFT3D.
- Layers to perform calculations in the frequency domain supporting complex values like FourierPooling2D.
- Wrapper to make initializer and activation functions complex-valued.
- Utilities to auto-detect your compute hardware.
- autoencoder like models like the UNet and discriminator models like a Layout-Content-Discriminator.
- Augmentations like Cut-Mix and Cut-Out
- Custom constraints for your layers like NonNegative.
- And many more...
⌚ Coming soon:
- A CycleGAN framework as used in VoloGAN
- A diffusion model framework as used in RGB-D-Fusion
- Further support for complex valued deep learning
git clone https://github.com/sascha-kirch/DeepSaki.git
cd DeepSaki
pip install .
pip install DeepSaki
I highly encourage you to contribute to DeepSaki. Checkout our contribution guide to get started.