Flux2Real is a ComfyUI workflow designed to transform your illustrations, anime-style images, or sketches into highly realistic portraits using the power of Flux Kontext.
Description
FAQ
Comments (12)
Never mind- I am an IDIOT! Found my answer right in your workflow, thanks! And well done!
So how does a non-Chinese national get the F.1高清摄影系列_HDR_v1 Lora?
I'm having trouble finding the answer in the workflow. What was it or where should I look in the follow? Thanks for any help you can give! @microfino368
https://www.liblib.art/modelinfo/293b44c89503459eb4bb5adbf95a09b7?from=pic_detail&versionUuid=cd48683a2c7447248e3fa1e64dce9704
you can download it from this site, I downloaded it from here. I'm not Chinese by the way :))
@drmgmt If you tell me a little bit about your problem, I can help you.
@mertvakitsayan can you post the lora somewhere we we can download?
@mertvakitsayan thx^^ if you find something like this post here - https://civitai.com/models/138630/hardcore-asian-cosplay-porn?modelVersionId=1909392 we need better anime to realist lora XD
I don't know why, but it doesn't work for me. I selected your lora model, your workflow and kontext, but the image remains animated and not real. I left your promp "(transform to real person photo:1.4), reversal film photography, hdr,"
Have I done something wrong?
I dont Know. But If you can send me a screenshot showing the visible part of the nodes before the workflow runs and what's inside, maybe I can get something out of it.
mertvakitsayan Hello, it was due to the fact that I installed the Magcache node, and it must have corrupted the workflow. I re-downloaded the workflow from scratch and haven’t had the issue since. It’s working quite well now.
What is the purpose of Magcache? I don’t see much difference with or without it.
Light_x02 As for Magcache, its main purpose is to improve performance by caching intermediate tensors during execution. This can help reduce VRAM usage and speed up workflows when you're reusing the same inputs across nodes—especially useful in complex workflows or when using large models
mertvakitsayan ok I currently use Teacache which does this too ;) By the way I get interesting results using your lora strenght 1.00 with another lora (dtpr_photorealistic) strenght 0.65 ;) I put an image in your gallery.