GAN (generative adversial networks) and some variants implementation.
- InfoGan: infomative GAN model.
- AeGan: auto-encoder GAN model.
- WGan-GP: Wasserstein GAN model with gradient peanlty.
- LsGan: least-square GAN model.
- Gan-cls: LSGAN model with features and labels together in input of discriminator.
To train/infer/generate GAN models, you may need to run codes by following the steps in each model folder:
the below steps are based on InfoGAN model, others have similar steps,
> python Train_Mnist.py
> python Inference_Mnist.py
> python Generate_Mnist.py
> python Plot.py
