hustvl / DiG

DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Could you also share the inference code?

goldhuang opened this issue · comments

I think it's more straightforward to test if I could train DiG to reproduce the results shown in the paper, using your inference code that has already been verified, because inference pipeline includes preprocessing , VAE encoder, DiG, decoder and post processing.
Thanks!