Chaofan Luo
Chaofan Luo
Yes, I think you're right. I'm using GB/T-7714-2015-numeric. Do you have any ideas of what might be causing this issue?
Here is the CSL: [csl.zip](https://github.com/Future-Scholars/paperlib/files/14572357/csl.zip)
> hi! thanks for the issue note that `sample` is scaled with `1/ ((sigma**2 + 1) ** 0.5)` for euler scheduler > > `self.scheduler.scale_model_input()` is always called inside pipeline at...
> Have you managed to solve it? Because I haven't.
To my knowledge, you should probably alter `hint_channels` under `control_stage_config` instead of `channels`
same here, anyone solved yet?
Literally solving it could be difficult because it's equivalent to the halting problem. The possible solution could be engineering workarounds such as putting a hard time limit etc.
Have you fixed this? Because I've encountered a similar ploblem.
> vae_transparent_encoder Hello, have you figured out yet?
> 3\. LatentTransparencyOffsetEncoder(alpha, RGB) + sdvae.encode( masked_rgb ) -> decode, it seems like not adding the offset performs better. Is this expected? > Thanks again. Hello, have you figured out...