
MAGIC DETAILZ - by pAInCREAT0R
Works on SDXL and PONY
New image enhancer that I have been working with as a side project. The goal is an enhancer that works with various models to alter/improve details and colors. I wanted something that worked on a variety of subjects and styles. Similar to adding "enhance style" prompts to the base prompt.
(Use V2.0 and not 1.0. There was an issue in some of the weights in 1.0 that I fixed in 2.0. Did not uncover them until I ran it in COMFYUI. Sorry for any inconvenience.)
I have not done extensive testing yet, but early results seemed good enough to share. It works with PONY Realism models fairly well - but the base LoRA was trained in SDXL. (I have only recently begun using PONY models to any extent, so that is the least tested aspect of it.)
I am posting some WITH/WITHOUT side-by-sides for the main pictures. I will post all of the individual pictures as well with prompts, etc. All but one of these were quick, simple prompts just designed to show some of the difference the LoRA impacts.
Works at a strength .5 to 1.0. No trigger word needed. Enjoy.
描述:
Corrected the block weight console errors in COMFYUI and reduced the file size.
训练词语:
名称: Creat0r_MagicDetailzV2_FIX.safetensors
大小 (KB): 450655
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success