Tokisaki_kurumi

Details

Download Files

About this version

I believe that this version is more stable and flexible than the alpha version, and it also fixes the eye color issue. I hope you like it and feel free to share your creation. For the best results, I recommend using either the Midnight Mixer or AbyssOrangeMix models with a LoRA weight of 0.4 to 1.

This version has been trained on 393 images sourced from Gelbooru, Yandere.re, and Konachan.

Model description

Example Prompt:

1girl, solo, (portrait:1.2), (face:1.3), (close-up:1.2), tokisaki_kurumi, 1girl, bare_shoulders, black_hair, blush, breasts, cleavage, (clock_eyes), dress, floating_hair, hairband, (heterochromia), lolita_fashion, lolita_hairband, long_hair, looking_at_viewer, medium_breasts, parted_lips, red_lips, lipstick, red_eyes, ribbon, solo, symbol-shaped_pupils, twintails, upper_body, wind, yellow_eyes

For best results, I recommend using the AbyssOrangeMix2 or Midnight Mixer V2 models with a LoRA weight of 0.5 to 1.

I don't know much about training, so I just followed Aitrepreneurs guide.

This model is an alpha version, trained with 84 hand-picked images from Gelbooru. In the future, I plan to learn more about training models and improve upon this

Images made by this model

Sort by

No Images Found.

Model Type	LORA
Base Model	SD 1.5
Published	2/18/2023