New Method/Model for 4-Step image generation with Flux and QWen Image - Code+Models posted yesterday
https://github.com/Lakonik/ComfyUI-piFlow

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

I am currently training a realism LoRA for Qwen Image and really like the results - Would appreciate people's opinions

https://redd.it/1osp42y
@rStableDiffusion

RTX 3090 24 GB VS RTX 5080 16GB

Hey, guys, I currently own an average computer with 32GB RAM and an RTX 3060, and I am looking to either buy a new PC or replace my old card with an RTX 3090 24GB. The new computer that I have in mind has an RTX 5080 16GB, and 64GB RAM.

I am just tired of struggling to use image models beyond XL (Flux, Qwen, Chroma), being unable to generate videos with Wan 2.2, and needing several hours to locally train a simple Lora for 1.5; training XL is out of the question. So what do you guys recommend to me?

How important is CPU RAM when using AI models? It is worth discarding the 3090 24GB for a new computer with twice my current RAM, but with a 5080 16GB?

https://redd.it/1oso8md
@rStableDiffusion