Upgrade to V2 version, utilizing more image training and aligning with displayed images.
升级到V2版本，采用更多图片训练，跟展示图片对齐
Description
updated train on additional image to align with the demos
采用更多类型图片训练跟展示图片对齐
FAQ
Comments (2)
Did you upload the wrong file? The Lora used in the picture is not related to the downloaded file
I seem to got dang near identical to one of their images with this LORA. I checked properties of one of their images and gathered the detailed gen data. You can usually download images from Civitai and open the images properties then go to details and check the comments. Here it is for theirs;
Steps: 24, Sampler: Euler, CFG scale: 14, Seed: 140915885, Size: 512x768, Model hash: e4a30e4607, Model: majicmixRealistic_v6, Denoising strength: 0.2, Clip skip: 2, Hires upscale: 2, Hires steps: 12, Hires upscaler: 4x-UltraSharp, Dynamic thresholding enabled: True, Mimic scale: 7, Threshold percentile: 100, Mimic mode: Half Cosine Up, Mimic scale minimum: 6, CFG mode: Half Cosine Up, CFG scale minimum: 6