Description:

Added nearly 2000 high-quality photographs and sci-fi images on top of the SDXL base model for training and fine-tuning, resulting in significant improvements in composition aesthetics and image details.

Parameters:

Base parameters:

CLIP end layers: 2

vae: none

Sampling: DPM++ 2M Karras, Euler a

Steps: 20-40

CFG: 7

High-resolution restoration parameters:

Scaling algorithm: R-ESRGAN 4x+ Anime6B

Redraw magnitude: 0.4-0.6

Points:

This is the first large model I have trained. It took a lot of time to collect materials and train to achieve the version I am most satisfied with now. So, if you like it, please give it a thumbs up and share more images! Feel free to leave me comments if you have any questions, and I will speed up the update! Thank you for your support!

Attention:

This model is prohibited for any form of commercial use, only for communication and learning purposes. You are not allowed to directly sell images generated by this model. Without my permission, you are not allowed to repost the model on other websites or sell it. Do not arbitrarily integrate this model. Respect others, do not use this model to generate images that are harmful to public order and good customs or insult others. You will be personally responsible for any legal consequences, and I will not be held responsible!!!