NVlabs/stylegan3

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

6943 stars · 1238 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Official PyTorch implementation of StyleGAN3 (Alias-Free GANs), a state-of-the-art generative adversarial network for high-fidelity image synthesis with translation- and rotation-equivariant generators. It includes training, visualization, video generation, and equivariance evaluation tools.

Use cases

• generate photorealistic synthetic images with a GAN
• train a StyleGAN3 model on my own image dataset
• create interpolation videos from a trained GAN model
• analyze aliasing and spectral properties of a generator
• evaluate GAN equivariance to translation and rotation
• load and run pretrained StyleGAN2 pickles

When to choose

• you need high-quality image synthesis with equivariant, alias-free generators
• you want to train or fine-tune StyleGAN2/3 models on custom datasets
• you need official, research-grade GAN tooling with visualization and metrics

When to avoid

• you need a lightweight or CPU-only image generator
• you want diffusion models instead of GANs
• you lack a modern NVIDIA GPU, since training requires substantial CUDA resources

Facets

library · maturity maintenance

machine-learning deep-learning image-processing gpu-computing deep-learning computer-vision image-processing artificial-intelligence python cli gan stylegan generative-models pytorch image-synthesis nvidia-research neurips-2021 gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="NVlabs/stylegan3")