Loading
Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
LoHA trained on Iki hiyori, Yato, Yukine from Noragami
Model source
Source description
LoHA trained on Iki hiyori, Yato, Yukine from Noragami
gif idea inspired by @Elluna Rem and Ram
models:
anime: MeinaAlter
2,5D: MeinaUnreal
tags:
Iki hiyori 1: iki hiyori, 1girl, pink scarf, long hair, brown hair, coat
Iki hiyori 2: iki hiyori, 1girl, long hair, school uniform, serafuku
u could also add a tail but its not that consistent
Sources
1 sourceModel artifacts
1 artifactSource excerpts
1 excerptSource context: 555 downloads · Type LoCon · Base model SD 1.5
Yato 1: yato, 1boy, track suit, bandana
to get the bandana consistently increase the weight of the LoHA to 1 or just be lucky :D
Yato 2: yato, 1boy, male focus, japanese clothes, short ponytail
Yukine 1: yukine, 1boy, fur-trimmed jacket, blonde hair
Yukine 2: yukine, 1boy, hoodie, beanie, blonde hair
negative embeddings: embedding:bad-hands-5 , embedding:ng_deepnegative_v1_75t , embedding:bad-artist , embedding:EasyNegative
Recommended settings:
Make sure you are aware on the usage instructions of LORA
VAE: orangemix.vae (for 2,5D pictures), vae-ft-mse-840000-ema-pruned (for realistic models), kl-f8-anime2 , blessed2.vae (for anime models)
Sampler: DPM++ SDE Karras or DPM++ 2M Karras
Steps: 20 to 35
Clipskip: 1 or 2
Upscaler : img2img in comfyui using 4x-UltraMix_Smooth
CFG Scale : 5 to 9 (mostly 7 and 8)
LORA weight for txt2img: 0.5-1 (for the examples i mostly used 0.8)
Denoise strength for img2img: 0.4 to 0.7
If you want the exact same settings, just drag and drop the picture into ComfyUI
for the pictures i used comfyui
