Bootoshi/retroanime
Bootoshi
Texto a imagen
LoRA de difusión para generación de imágenes en estilo anime retro, diseñado como adaptador para FLUX.1-dev. El modelo se activa con el término "retro anime" y está publicado en Hugging Face con licencia Apache 2.0.
Como usar
Instalación y uso con Diffusers:
pip install -U diffusers transformers accelerate
import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained(
"black-forest-labs/FLUX.1-dev",
dtype=torch.bfloat16,
device_map="cuda"
)
pipe.load_lora_weights("Bootoshi/retroanime")
prompt = "A retro anime style drawing of a female woman that has short, black hair, bangs, and hair tendrils. She has cat-like ears with dark fur on the outside and white fur on the inside. Her outfit is a tight, glossy black latex bodysuit with a revealing design, emphasizing her chest and legs. She is wearing long, thigh-high boots that match the bodysuit. She has a white and black, fluffy tail, resembling that of an wolf. Her skin is smooth and fair, and she has a slender, curvy body type with a petite, hourglass figure. She is hot with big green eyes a long face with a sharp jawline. She is driving an open topped car"
image = pipe(prompt).images[0]
Funcionalidades
- Tarea: texto a imagen.
- Tipo: adaptador LoRA para Diffusers.
- Modelo base: black-forest-labs/FLUX.1-dev.
- Formato de pesos disponible: Safetensors.
- Palabra de activación recomendada: retro anime.
- Compatible con Diffusers e inferencia mediante proveedores como fal.
Casos de uso
- Generar ilustraciones de personajes con estética de anime retro.
- Aplicar un estilo visual retroanime a imágenes generadas con FLUX.1-dev.
- Probar un LoRA de estilo en flujos locales o notebooks con Diffusers.
- Usar el adaptador en aplicaciones compatibles como Draw Things o DiffusionBee.