Unlike the “Ahegao” expression, the “Ohogao (オホ顔)” expression is characterized by a vertically elongated mouth without a protruding tongue.
While most standard image generation models can generate Ahegao expressions, generating Ohogao expressions is difficult.
This LoRA supports the generation of Ohogao expressions.
Description
FAQ
Comments (4)
Thanks a lot for this LoRA. It is indeed very difficult to render the "ohogao" danbooru tag for as far as this is a tag with a limited dataset/score (about 300).
Tho i noticed a HUGE style change from your image comparison when using your LoRA at score of 1 vs a score of 0.01. Is this really a style change resulting from your LoRA?
Generally, LoRAs do affect image quality.
However, in this case, it seems the “ohogao” tag in the base model contains specific character traits. Please try the following parameters and model.
If you include the “ohogao” tag, I think it will generate a low-quality young woman with black hair and thick eyebrows.
====
ohogao,1girl,solo,
Negative prompt: nsfw,
Steps: 16, Sampler: Euler a, Schedule type: Automatic, CFG scale: 7, Seed: 1017738115, Size: 768x1344, Model hash: f116b0c78f, Model: waiIllustriousSDXL_v170, VAE hash: 235745af8d, VAE: sdxl_vae.safetensors, Clip skip: 2, Version: v1.10.1
====
MAGICAL 😭😭😭
Any chance for anima?
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.