# torchgan/torchgan

Research Framework for easy and efficient training of GANs based on Pytorch

Repository: https://github.com/torchgan/torchgan
Canonical: https://ross.abutalabs.com/products/torchgan
Homepage: https://torchgan.readthedocs.io/en/latest/
Language: Python
License: MIT
License Family: permissive
Topics: machine-learning, computer-vision, generative-adversarial-networks, python3, pytorch, neural-networks, deep-learning, generative-model, python, gans
Last push: 2022-10-23T03:55:31+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2904, "days_push": 1410, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1425, forks 167 (observed 2026-08-28T04:04:41.502893+00:00)

## What it is
TorchGAN is a PyTorch-based research framework for designing and training Generative Adversarial Networks. It provides modular building blocks for popular GAN models while allowing customization of loss functions and architectures for cutting-edge research.

## Use cases
- train GAN models on my dataset
- implement a custom GAN loss function
- experiment with generative adversarial network architectures
- compare popular GAN variants like DCGAN or WGAN
- visualize GAN training with logging backends
- prototype new generative model research ideas

## When to choose
- you want a modular PyTorch framework specifically for GAN research
- you need to mix and match GAN losses, trainers, and architectures
- you want built-in logging and visualization for GAN training

## When to avoid
- you need actively maintained software with recent updates
- you need general-purpose generative modeling beyond GANs (e.g., diffusion models)
- you prefer a higher-level framework like PyTorch Lightning for general training

## Facets
- artifact type: framework
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: machine-learning, deep-learning, computer-vision
- platform: python
- tags: gan, generative-adversarial-networks, pytorch, research-framework, generative-models, linux, macos, gpu

## Member repositories
- torchgan/torchgan (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:41.502893+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:37:32.491247+00:00, confidence not recorded.
  - readme: https://github.com/torchgan/torchgan (fetched 2026-08-28T04:04:41.502893+00:00, sha 2d3872a4d095)
  - registry_pypi: https://pypi.org/pypi/torchgan/json (fetched 2026-08-29T11:49:57.973526+00:00, sha f4892f3e047c)
- Data as of 2026-08-30T08:39:29.467469+00:00.