Note: Yes, I'm aware of the left hand looking "smaller" on the "flagship" image, my first choice image got marked as 13+, so I went with this image instead and didn't pay enough attention to the details. Don't let it discourage you from trying it yourself, it works fine.
Trained on Anything V5, 15 images. You might experience issues with the hat generation, or at least the "cowboy hat" generation.
Works great with ThorneMix_v1, Dark Sushi 2.5D 大颗寿司2.5D, Coffee NSFW.
darkSushi 2.5D_V2.0, Mistoon_Anime v2.0 (kl-f8-anime2 VAE) and GrahamAingPikas-ClearVAE 2.3 should also produce good results.
I use Euler A
23 sampling steps ( I use this little because of testing, feel free to use more)
7 CFG
Thumbnail Model: ThorneMix_v1