akanimax/BMSG-GAN
[MSG-GAN] Any body can GAN! Highly stable and robust architecture. Requires little to no hyperparameter tuning. Pytorch Implementation
Implements multi-scale gradient connections between generator and discriminator intermediate layers to enable synchronized training across resolution levels, eliminating the need for progressive growing schemes. Supports multiple loss functions including relativistic hinge and Wasserstein with gradient penalty, plus optional equalized learning rates and exponential moving averages for improved convergence. Integrates with PyTorch and AWS SageMaker for distributed training across multiple GPUs.
627 stars. No commits in the last 6 months.
Stars
627
Forks
103
Language
Python
License
MIT
Category
Last pushed
Jun 17, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/akanimax/BMSG-GAN"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
NVIDIA/pix2pixHD
Synthesizing and manipulating 2048x1024 images with conditional GANs
GaParmar/clean-fid
PyTorch - FID calculation with proper image resizing and quantization steps [CVPR 2022]
albertpumarola/GANimation
GANimation: Anatomically-aware Facial Animation from a Single Image (ECCV'18 Oral) [PyTorch]
yuanming-hu/exposure
Learning infinite-resolution image processing with GAN and RL from unpaired image datasets,...
yiranran/APDrawingGAN
Code for APDrawingGAN: Generating Artistic Portrait Drawings from Face Photos with Hierarchical...