anonymous721

It's causing me a lot of annoyance too. 15-20 minutes testing a couple different models, and I'm already over 20GB of system RAM used.

If the ROCm fork does get merged in, would the Int8 Matmul compatibility improvements also work for AMD GPUs?

"Apply color correction to img2img results to match original colors" was on, and turning it off solves the problem. I'm not entirely sure what it does beyond what the name...

The old hash method already had a split when the safetensors format came about to replace ckpt, since a ckpt and a safetensors version of the same model have different...

It is impossible for Stable Diffusion to generate images that "exploit" cause "harm" to children. The "children" it generates are not real, therefore it is impossible to harm or exploit...

"Things that aren't real are not real" isn't circular, it's tautological. Important distinction. And I have no idea what you're talking about since his victim was not not real, so...

>Old hires. fix made it very easy to commit errors Old hires. fix was incredibly easy to use without trouble. You set the resolution you wanted, ticked the box, and...

>Recreating old seeds should be simple - just take the first pass values and plug those directly into height & width How do we determine what the old firstpass values...

It looks like it may be impossible to recreate some old results. On a 512x768 image whose metadata reported 0x0 as the first pass resolution, feeding it into the PNG...

Running that version, it didn't stick on the scanning step at all. Which is good for usability but I'm not sure how useful the log will be. Here it is...