LCM_SoteMix / README.md
Disty0's picture
Update README.md
9ffc3b5
metadata
license: creativeml-openrail-m
language:
  - en
library_name: diffusers
pipeline_tag: text-to-image
tags:
  - art
  - anime
  - stable diffusion
  - lcm
  - latent consistency model
  - openvino
  - onnx

A Stable Diffusion 1.5 anime model with LCM that allows image generation with only 4 steps.
More details on LCM: https://latent-consistency-models.github.io/

Original SoteMix is available here:
https://huggingface.co/Disty0/SoteMix

Coverted to LCM with this script:
https://github.com/vladmandic/automatic/blob/master/cli/lcm-convert.py

Positive Prompts:

(masterpiece, best quality, highres), 1girl,

Negative Prompts:

(worst quality, low quality, lowres), zombie, interlocked fingers,

Sampler: LCM
Steps: 4-8
Clip Skip: 1
CFG: 0 or 1
Base Resolution: 512x / 768x / 1024x / 768x1280

Second Pass / Hires:
Sampler: LCM
Steps: 2-6
Upscaler: RealESRGAN 4x+ Anime6B / ESRGAN 4x-AnimeSharp with 0.3-0.4 denoise strength.
CFG: 0 or 1
Resolution: 2x of the base resolution.

Mege:

0.7(0.75(0.75(0.75(0.75(0.75(0.5(SomethingV2_2) + 0.5(meinapastel_v5Anime)) + 0.25(meinaalter_v2)) + 0.25(meinamix_meinaV9)) + 0.25(nyanMix_230303Intense)) + 0.25(mixProV4_v4)) + 0.25(breakdomain_M2150)) + 0.3(0.4(SomethingV2_2) + 0.6(meinapastel_v5Anime))

And baked in Blessed2 VAE

image/png