Simple Comfy UI workflow for a simple 4K image generation with multi-step upscaling and multi-face fix. All nodes should be able to be automatically downloaded through comfy UI manager.
I will upload the SD1.5 upscaler model soon
Will also make a video on how to use this soon, but for most people who are comfortable with comfy UI this should be pretty self explanatory.
Simple Video here.
For the SD1.5 portion of the workflow, you can get the embeddings here. https://huggingface.co/Loarintang/SD1.5_Embeddings_pack
Description
FAQ
Comments (10)
Do I need Tensor to use this
No, you don't. I sometimes use sage attention and I need to include the node disabling it or else the SD1.5 part doesn't work. So you can either bypass or delete it if you want.
Thanks! Waiting for your RealMixEuropean1. BTW I see you use various LoRAs, can you add links to them? I found [Hand 5.5](https://civitai.com/models/200255?modelVersionId=997426), [Detailed Eyes v3](https://civitai.com/models/120723/detailedeyesxl), [Fractalism](https://civitai.com/models/240059?modelVersionId=270828). But I have doubts about [Add Detail](https://huggingface.co/PvDeep/Add-Detail-XL/tree/main) and [Cute Person](https://huggingface.co/wsj1995/LORA/resolve/main/542409/603077/cuteperson.safetensors?download=true)
nice, thank you
Looks like you found them!
Thanks for the hard work on this workflow, that works perfectly. Quick comments though :
- you lack a "quick test" button that will compute the image in small, to test seed/prompt more quickly (I dont want to wait 5 minutes each time).
- I do not understand the double process with CR Lora stack and all, cant see any image improvement.
- there is a bug on CR SDXL aspect ratio: if you pick "portrait" it generates a landscape and vice-versa
this workflow is genius, thank you