prithivMLmods/Qwen-Image-Studio-Realism

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("Qwen/Qwen-Image", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("prithivMLmods/Qwen-Image-Studio-Realism")

prompt = "Studio Realism, a close-up shot of a young womans face features a neutral expression. The womans eyes are a deep brown, her hair is a dark brown, and her eyebrows are a lighter shade of brown. She is wearing a black and yellow t-shirt, with a cream collar around her neck. Her ear is adorned with a gold earring, adding a touch of shine to her face. The backdrop is a stark white, and the womans hair is pulled back in a ponytail."
image = pipe(prompt).images[0]

Configuración local indicada en la tarjeta del modelo:

import torch
from diffusers import DiffusionPipeline

base_model = "Qwen/Qwen-Image"
pipe = DiffusionPipeline.from_pretrained(base_model, torch_dtype=torch.bfloat16)

lora_repo = "prithivMLmods/Qwen-Image-Studio-Realism"
trigger_word = "Studio Realism"
pipe.load_lora_weights(lora_repo)

device = torch.device("cuda")
pipe.to(device)

Incluye siempre "Studio Realism" al inicio o dentro del prompt para activar el estilo del LoRA.