User LoRA

SD

Hyperdetailed Quality

Hyperdetailed Quality - AI Model cover image
Entrenado por
Avatar image of たつ

たつ

Entrena tu propio LoRA
Subido en
7 sept 2023, 16:07
Utiliza
382.47k
Ganado
200,000
Reseñas
Excelente  (24)
Palabras clave

(HDR), (8k), (32k), (masterpiece:1.1), (Best quality), (Ultra detailed eyes)

Permisos
  • Allow image generation & sharing
  • Permita que los usuarios descarguen su modelo
  • Usos comerciales

Descripción

This is a LOra made by しTatsusuki

"Have you imagined the world of anime being filled with wonderful illustrations only using one LoRA?" well, Hyperdetailed Quality is there for you! Even if trained in Pixai.Art with only 20 images, it can generate beautiful artworks with the only limitation being your creativity!

Recommendations:

  • CGF SCALE 1.0-2.0
  • Sampling Steps 40.0-50.0
  • 0.5 to 0.9 LOra weight

Observation: The LoRA won't change your style, but only make your artwork more attractive! (doesn't work too well with realistic images, as it'll create an artificial look to it)

Feel free to mix other LoRAs with this one, or else it'll feel lonely :(

Comentarios5