santis_yang
santis_yang
@CharlesShang @jasonhsiao97 Is it right that all head architecture(Mask branch, Fast RCNN) should be shared on every scale of FPN layer? @CharlesShang I have few questions about your code: 1....
Looking forward to a opencv-like `remap` function in DALI. It is a very common augmentation operation nowadays. ❤️
Same issue here.
Yes. Notable findings @ChawDoe . Also for distributed training, please make sure that you are loading different data at every spawned process, namely paying attention to your random seed setting...
@lucidrains @adelacvg Is this because in the pointed DDPM implementation, the UNnet has default attention layers every level while in this repo, you have to point to which layer should...
@lucidrains Have you evalute how these two repos perform on unconditional generation tasks? I appreciate your contributions very much and hope get more hints on this.
@lucidrains Any plan to add the DPM solver in the near future? This project is awesome but it seems that it lacks some quick sampler method.
Same here. Eager to see cross attention implementation in DiT.
> Thanks. Copying the compiled files to site-packages solved my issue.
> > I found when using EMA for cosine code book, the l2-norm of the input to the vq module would grow gradually, from 22 -> 20000, leading to growing...