import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("black-forest-labs/FLUX.1-Fill-dev", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("Ling277/FaceComposer")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]FaceComposer
Model description
Official research code for FaceComposer: Learning Coherent and Realistic Face Composition from References.
FaceComposer performs reference-based face composition: given a source portrait, a reference portrait, and selected facial components such as eyebrows, eyes, nose, or mouth, the model edits the source ROI toward the reference while preserving source pose, expression, illumination, and context.
Download model
Download them in the Files & versions tab.
- Downloads last month
- 14
Model tree for Ling277/FaceComposer
Base model
black-forest-labs/FLUX.1-Fill-dev