Accidentally deleted this whole model post, redoing it from memory.
Colorizer Manga Doujinshi ANIMA edit: https://civarchive.com/models/2685547/manga-doujinshi-colorizer-anima-edit-lora
Manhwa Uncensor Lora: https://civarchive.com/models/2685875/manhwa-uncensor-lora-flux-2-klein-9b-blank-censor
V1 Overview:
Provide a Blank Censored squared image (512x512) and additionally a reference image of the genital and will uncensor using that reference (the reference image is required). In resources of this post I attached some reference images from the dataset for you to use.
CODEX made two node for getting the squared crop from any image easily and stitch it back: https://github.com/Norian11/comfyui_ManhwaCropping.git
Workflow with the custom node: https://pastebin.com/TSpuPgxT
Workflow without the custom node: https://pastebin.com/tRjqrXC5
Recommendations:
Model: Distilled Klein 9B
Steps: 4
Prompt: (NONE)
Input image at 512x512 (my node does this automatically)
Description
FAQ
Comments (2)
I tested it, but it doesn't appear to be working. Could you provide your test images? Like a .zip file with the .json file containing the input images. I tested all your other workflows, but this is the only one that doesn't seem to work.
Its ass