Sign In

Muchen-SDXL-Portrait

101
676
27
Verified:
SafeTensor
Type
Checkpoint Trained
Stats
676
Reviews
Published
May 12, 2024
Base Model
SDXL 1.0
Usage Tips
Clip Skip: 2
Trigger Words
muchen
cngirl
Hash
AutoV2
AE9D7DB8F6

The official version is now available for download!

Simply repaired the hands and the whole body training set,

Not perfect, not perfect, not perfect, everyone. . .

Guide words: muchen, cngirl, if you look like a foreigner, please ➕ guide words

If you want to generate photo texture, a little reference is as follows:

imperfect skin body, (imperfect skin,

You can also add skin descriptions, in short XL can do a lot)

cinematic,(movie feel)

deep photo, (deep photo)

depth of field, (depth of field)

Superia 400, (film type, other optional)

shadows, (sense of light and shadow)

grainy

If you want to create a big-brand feel for your clothing, a few tips are as follows:

haute couture dress,

Brands or other clothing types and colors can be added.

Special thanks to the big guys for their strong technical and spiritual support during the model refining process!

names not listed in order:

Maiju, none, Miaoka, Lin He, Liao Wushi, Bai Ze, Tide, Yi Nian, Lao Xia Ciao, Yuny. .

........etc. and all the big guys in the group that I haven't mentioned yet. . (There are too many to finish)

Sampling method: DPM++ 2M Karras (other sampling is welcome to try)

Number of iteration steps: 30 or more (depends on your mood)

Recommended size: 768*1024 (welcome to try other sizes)

Prompt word guidance coefficient (CFG): 5-7

HD resolution fix: 8x_NMKD-Superscale_150000_G

If there is a picture that meets the requirements, use high-definition enlargement, and the redrawing range should not be too high.

<Notes>

-There is a chance that foreigners will appear in the V1 version, and subsequent versions will continue to fix it.

-There is a chance that the hand will collapse in the V1 version, and subsequent versions will continue to fix it.

-The V1 version may not be friendly to the whole body support (maybe it’s a problem with the training set), or maybe I haven’t tested it too much. Feedback is welcome!

- When the character's whole body or proportions are small, the face will appear broken and blurred. Please open ADetailer and check to enable it, and select face_yolov8n to repair it!

- Some prompt words may not have obvious effects, and subsequent versions will expand their versatility;

————————————————————————————————

Model purpose statement:

1. You may not host this model and its derivatives (such as the Fusion Model version) on websites/applications that plan to earn revenue or donate. If you need to use this model and its derivatives for commercial purposes (such as generative services, selling pictures, using pictures to publish articles or publications on short video platforms, etc.), please contact me through the site;

2. You may not directly sell images generated by this model unless you have made sufficient manual modifications to the image so that it can be fully determined as your personal work in a legal sense. If you violate this article, you will be personally responsible for all legal consequences, and I will not be held responsible;

3. You cannot use this model to intentionally create or share illegal or harmful content and output. Please abide by public order and morality and use this model for positive purposes.