About model

It's megre of Flat-2D Animerge V3 and AbyssOrangeMix2 - Hardcore. VAE is enabled. Model created for ???? cartoon generation. Author Telegram (my generations are here).

Important

Use 512x768 defenition for basic results (thick lines), use 640x832 for golden mean results, use 768x1152 for good results (fine and beautiful lines), use 1024x1536 for experimental pictures.

Fix FACES

Use img2img OR ADetailer extension! For ADetailer I recomend to use my settings:

Inpaint mask blur: 5
Inpaint denoising strength: 0.55 - 0.6
ADetailer model: face_yolov8n.pt
Negative: |like in the picture|

Usage

Author of original Flat-2D recommends using 10.0 CFG scale, but I recomend using 7.5. Also, I recommend to use Euler (without a) sampler with 22-24 steps. My favorite negative for generation is:

mutation, ugly hands, badly drawn hands, fingers, BadDream, badhandv4,  ng_deepnegative_v1_75t, <lora:easynegative:1> easynegative, (watermark, logo:1.2) 

I used 1 LoRA and few Embeddings: lora:easynegative, BadDream.pt, badhandv4.pt, ng_deepnegative_v1_75t.pt.