Need help - transitioning from ChatGPT image Gen to SD
https://redd.it/1s6r6rh
@rStableDiffusion

LTX 2.3 Reasoning VBVR Lora comparison on facial expressions

https://redd.it/1s6uthp
@rStableDiffusion

Hunger of "Workflow!?"
https://redd.it/1s6uh4t
@rStableDiffusion

SFW Prompt Pack v3.0 — 670 styles · 29 categories

https://redd.it/1s6tlch
@rStableDiffusion

[Training-Free] Bring Famous Paintings to Life! Every Painting Awakened (I2V)

https://redd.it/1s6x7gq
@rStableDiffusion

ZImageTurbo nodes
https://redd.it/1s6o0ey
@rStableDiffusion

What model did they use here?

https://redd.it/1s72igk
@rStableDiffusion

GTA becomes Anime

https://redd.it/1s73c7m
@rStableDiffusion

Whats the verdict on Sage Attention 3 now? or stick with Sage 2.2?

I use Image Z Turbo, Wan 2.2 and LTX 2.3

I noticed that Sage Attention 3 altered the dress in a video of a dancing woman to a trousers when using LTX 2.3, I switched to Sage 2.2 and also tried disabling it and the issue was fixed

I actually thought it was the GGUF text encoder that causes the dress to turn into a pants but to my surprise it was Sage 3 that was causing it.

I went back to 2.2 only lost a few seconds speed by the quality was like if it' was disabled very good.

https://redd.it/1s73r4e
@rStableDiffusion