ntc-ai/SDXL-LoRA-slider.oil-painting

Instalación y uso básico con Diffusers:

pip install -U diffusers transformers accelerate

import torch
from diffusers import DiffusionPipeline

# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("ntc-ai/SDXL-LoRA-slider.oil-painting")

prompt = "oil painting"
image = pipe(prompt).images[0]

Ejemplo con carga explícita del archivo Safetensors y peso del adaptador:

from diffusers import StableDiffusionXLPipeline
from diffusers import EulerAncestralDiscreteScheduler
import torch

pipe = StableDiffusionXLPipeline.from_single_file("https://huggingface.co/martyn/sdxl-turbo-mario-merge-top-rated/blob/main/topRatedTurboxlLCM_v10.safetensors")
pipe.to("cuda")
pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config)

# Load the LoRA
pipe.load_lora_weights('ntc-ai/SDXL-LoRA-slider.oil-painting', weight_name='oil painting.safetensors', adapter_name="oil painting")

# Activate the LoRA
pipe.set_adapters(["oil painting"], adapter_weights=[2.0])

prompt = "medieval rich kingpin sitting in a tavern, oil painting"
negative_prompt = "nsfw"
width = 512
height = 512
num_inference_steps = 10
guidance_scale = 2

image = pipe(prompt, negative_prompt=negative_prompt, width=width, height=height, guidance_scale=guidance_scale, num_inference_steps=num_inference_steps).images[0]
image.save('result.png')