ImaginMixReal

Version description: A new fusion method is used based on mix80

; mixed 90 is generated 23 times. I hope to generate more stylish character images for everyone. I personally recommend sampling: elua

, step count 30, HD restoration 0.2-0.4 We are a small circle that loves realistic style. The upload model is a partner who hopes to have like-minded people who like “realism”

Let's “teach” us. Together, we want to “learn”

model data thanks to the 23 link data overlays used in all models in fusion superposition

(wheat orange, rabbit, old white, chocolate, DynamicWang, etc.)

All of the images in the example are randomly generated. In theory, seeds can be locked and sampled consistently; Circle WeChat Group QR code (posted irregularly every weekend) WeChat: Fashion

-i

Join the circle: return to us. Your return image can be a high quality return image and is not limited to using our model

, Your high-quality return image We invite you to play with us

(realistic

, 2.5D, magical, mechanical, etc.) Negative hint recommendations:

(worst quality), (low quality), normal quality, lowres, bad anatomy, bad hands, (monochrome), ( Grayscale) watermark

, moles