Darius-H

I met the same problem. If I use utf-8 to save the code, I can open it again, but it will output chinese characters wrongly. If I use ANSI to...

> Hi @Darius-H ! > > To make sure I understand, you're saying the mask should also be resized to be the `shape` passed to `gr.Image`? I mean, you can...

bug not fixed yet

``` from mmengine.dist import get_dist_info,init_dist ```

I met the same problem on Windows10 python3.9.7. Why this problem has not been tackled yet?

You can download it here: https://openmmlab-share.oss-cn-hangzhou.aliyuncs.com/mmhuman3d/models/expose/expose/expose-d9d5dbf7_20220708.pth?versionId=CAEQRBiBgMC8vbbujhgiIDg0NWUyM2ZiZGY3MzQ0YmI5YjFjYTA0Y2Q5NDE3MDEw Please read this doc first: https://github.com/open-mmlab/mmhuman3d/tree/main/configs/expose

I found the download link https://a3s.fi/swift/v1/AUTH_a235c0f452d648828f745589cde1219a/specvqgan_public/melception-21-05-10T09-28-40.pt in https://github.com/v-iashin/SpecVQGAN, is that the one you used in your project ?

The reconstruction performance is a little worse than 2D VAE but it works better when using with diffusion. Training 1D VAE with 2D discriminators is prone to instability problems, leading...

> i got the bug about nccl when i try to train ,so can i run it on window without nccl I haven't tried running it on windows. You say...