Efficient Multi-Instance Generation with Janus-Pro-Dirven Prompt Parsing
Paper • 2503.21069 • Published
import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("Jonas179/MIGLoRA", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]Model Description
This is the official weight of "Efficient Multi-Instance Generation with Janus-Pro-Dirven Prompt Parsing". We propose an efficient Text-to-Image model that can generate high-quality images with reasonable layouts according to the requirements.
You can read our paper on arXiv to dive deeper into the theoretical foundations and experiments.
Todo List:
- The complete codebase will be hosted on GitHub
- The full dataset will be made available through Hugging Face Datasets
- Downloads last month
- -
Model tree for Jonas179/MIGLoRA
Base model
stabilityai/stable-diffusion-3.5-medium