Ceyuan Yang

Hi @jongh222, Could you please try to load any checkpoint file on your platform? We didn't debug it on windows before. For instance, you can load the weight of pre-trained...

Hi @ryanwebster90 , The dataset actually samples the subset with a given [random seed](https://github.com/genforce/insgen/blob/52bda7cfe59094fbb2f533a0355fff1392b0d380/training/dataset.py#L42). We follow the default seed 0 in all experiments.

Thanks for your testing. Could you please tell me which boundary you used? We will follow up then.

Actually, HiGAN applies the manipulation in w space instead of z. And yes, feeding the edited w code into the pretrained generator results in the corresponding image.