Here resides my practice for generative models. These codes can be writen by myself or adapted from other sources with some modification. In case of adoption, I include credit in the corresponding notebooks.
| Model | Pytorch | Tensorflow |
|---|---|---|
| CycleGAN | cycleGAN_h2z: transform horse to zebra and vice versa, experiment with different generators' architectures and coefficients for loss components (identity, GAN, and cycle)  |