These files are Custom Workflows for ComfyUI
ComfyUI is a super powerful node-based, modular, interface for Stable Diffusion. I have a brief overview of what it is and does here. And full tutorial on my Patreon, updated frequently.
Please consider joining my Patreon! Advanced SD tutorials, settings explanations, adult-art, from a female content creator (me!) patreon.com/theally
Description
Advanced LoRA flow which also includes Hires fix upscaling.
FAQ
Comments (13)
Probably a dumb question but how do you download the upscale models? My folder is void of any. I installed comfy with the standalone zip.
Hey! You can get the upscale models from here - https://upscale.wiki/wiki/Model_Database I highly recommend 4x-foolhardy-remacri, second in the table.
no need to download anything if you have automatic1111 installed on your computer, just use the extra_model_path.yaml.example and edit the paths to automatic1111 folders, no duplicity needed
Thank you Theally, for the hint to ComfyUI. That is probably really the future. I already get good results with the default configuration without HiresFix, with fewer problems with the hands and faces. At the moment, the only thing I miss is the Checkpoint Merger.
I want to try it but it's so hard to step away from Auto's WebUI... I've seen this node style workflow before with a upscaler idr the name of, they're scary.
@undelirium Are you referring to chaiNNer ?
Create an image like you usually do in A1111, then load that image into ComfyUI (can just drag&drop into the web interface of ComfyUI). The nodes should be created for you, and you can learn the rope from there. At least this is how I started with ComfyUI. (Note that if you use img2img from txt2img then this won't work, as the PNGInfo only show what was used in img2img)
The issue with ComfyUI is that it does not have much extensions (if any) for easy incremental workflow - drawing masks for inpainting for example.
:) why not both....there is a sd web ui extension for comfyui that makes it a tab inside auto1111
Complete garbage. it is not compatible with ComfyUI, because this package does not contain the init.py file
It was 100% compatible with the version I developed it on. I haven't kept it up to date with ComfyUI updates, though. Thanks for letting me know!
Dude. Maybe be a little bit nicer to the folks that make this stuff for the rest of us?
it really work, thanks for the great work sharing. It was very useful to me.,appreciated. i hope if you got time, could addon mutiple lora loader,at least for 3 loras,textual inversion for negative embeding.
You can do that by "chaining" loRa Loaders together, as many as you want. The only issue is you will add more time to the overall generation time, for each one you use/add. You could loop out LoRa's so you don't use so many when you don't need them.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.