Every version can be accurately generated, close ups and full body.
NSFW is OK.
I hope you enjoy creating it.
（Sales are prohibited. Please use for personal enjoyment）
Uploading to Tensor Art is prohibited.
If you like, please try "anime-ver." as well.
https://civarchive.com/models/1170176/diving-illustrious-anime
Adjusting the lighting can make a big difference in how face looks.
Recommended Settings
Sampling method：DPM++ SDE
Schedule type：Karras
Size：896 × 1152 (You can set it to your liking here)
Step：15〜
CFG : 4 (3〜6)
ADetailer：mediapipe_face_mesh
masterpiece, ultra-HD, photorealistic, high detail, best quality, 8k, best quality, sharp focus, ray-tracing,
realistic, depth of field, shallow depth of field, raw photo.V1.0
v1.0 pursued realism.
V2.0
v2.0 is now easier to use. It's realistic even without LoRA.
Add suggested LoRA to your liking.
V3.0
v3.0 is easier to use.
Description
FAQ
Comments (12)
can this be used in SD or is this for Comfy? cause am trying in SD but its giving only black images
Works for me on SD Forge just like all other Illustrious checkpoints do
what prompt can i use to avoid extremely elongated torso?
Thank you! A very good model. However, sometimes when generating male characters, the facial features of the male characters tend to be more feminine. Could this be improved in the future?
Please leave some notes in the "about" section for each version. Cause there is just no way to know what the intended improvements are without a bunch of trial & error.
Did you train this on any Japanese, Chinese, or Korean characters? I'm curious if the kanji and hiragana I added to my prompts makes a difference.
Because this is a illustrious XL model,and illustrious XL model has strong character knowledge.
You can try other characters,for example, aerith_gainsborough,and will found It also knows.
You can try it with any oher illustrious XL based models and found it still works.
This is what illustrious XL model works,not this model.
@YSCysc Thank you! When I typed characters I meant kanji, hiragana, hangul, and jamo for example.
v3 is best,other is not good.
Your Asian models are incredible! Is it possible to create a Z-Image model? That would be amazing if you could!
I made a Z-Image model.
https://civitai.com/models/2276359/diving-zimageturbo-real-asian
amazing model, please comeback to use directly in civit generator
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.