This model has a new advanced fine-tune format with hyper-parameters or hyper leaning rate. This new training technique is a discovery, which reduces the cost of training models in SD 1.5. It has not been tested in other SDXL and Flux Dev models. This technique forces accelerated training so that the model has no losses, having high stability and consistency in image generation.
Introducing the new version of the Perfect Woman SD model, trained with 9308 photos at 640x960 resolution, featuring a greater variety of faces, skin tones, hair colors, many lingerie looks, and beautiful landscapes of rivers, lakes, and waterfalls with gorgeous women.
I recommend using the Adetailer extension.
Use this extension to fix hand errors:
https://github.com/licyk/advanced_euler_sampler_extension
It is strictly prohibited to distribute, inside and outside civit ai, this template or post it anywhere or on hosting services, on third party image generation services, you can use it for personal use, if you know who distributed it without my consent , report it immediately.
Resolution settings:
Standard resolutions 512 x 640, 640x 768, 640x960
Resolution: 540x540 select the Hires.fix checkbox, result 1080 x 1080 pixels
Note: I am testing the sampling methods and I will add the others to the settings.
Use these recommended settings for generation:
Sampling method: Restart
Sampling steps: 30-50
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: Kohaku_LoNyu_Yog
Sampling steps: 30-50
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: Euler_Max
Sampling steps: 30-50
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: Euler_Smea_Dy
Sampling steps: 18-50
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: Euler a
Sampling steps: 18-50
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: LCM
Sampling steps: 18-30
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: DDPM Karras
Sampling steps: 18-30
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: DPM++ 2M
Sampling steps: 18-30
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: DPM++ SDE Karras
Sampling steps: 18-30
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Sampling method: DPM++ 2M SDE Karras
Sampling steps: 18-30
CFG Scale: 2.0 - 7.0
Skip clip: 1-2
Use the recommended negative prompt:
(CyberRealistic_Negative-neg), cartoon, painting, illustration, (grayscale:1.4), (worst quality:2), (low quality:2), (normal quality:2), ugly, fat, 3D rendering, bad hands, malformed hand, amorphous hand, multiplied fingers
Description
This model only generates images in portrait format because it was trained at that resolution of 640x960. I do not recommend generating images in landscape format; it produces realistic and photographic images.
FAQ
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.