# znxlwm/UGATIT-pytorch

Official PyTorch implementation of U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation

Repository: https://github.com/znxlwm/UGATIT-pytorch
Canonical: https://ross.abutalabs.com/products/ugatit-pytorch
Language: Python
License: MIT
License Family: permissive
Last push: 2023-03-16T02:38:05+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2592, "days_push": 1267, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2536, forks 461 (observed 2026-08-28T04:06:59.145189+00:00)

## What it is
Official PyTorch implementation of U-GAT-IT, an unsupervised image-to-image translation model using attention modules and adaptive layer-instance normalization (AdaLIN). It trains GAN-based models to translate images between domains, such as selfies to anime portraits.

## Use cases
- convert selfie photos to anime style
- train an unsupervised image-to-image translation model
- reproduce U-GAT-IT paper results in PyTorch
- translate images between two unpaired image domains
- experiment with attention-based GAN architectures

## When to choose
- you need unsupervised domain-to-domain image translation without paired data
- you want the official reference implementation of the U-GAT-IT paper
- your translation task involves large shape and texture changes, like selfie-to-anime

## When to avoid
- you need a maintained production-ready image translation pipeline
- you want paired pix2pix-style translation with aligned training data
- you lack a GPU, since training is memory-intensive

## Facets
- artifact type: library
- maturity: maintenance
- function: image-processing, machine-learning, deep-learning
- domain: deep-learning, computer-vision, image-processing, artificial-intelligence
- platform: python
- tags: gan, image-to-image-translation, pytorch, attention, selfie2anime, research-code, gpu, linux

## Member repositories
- znxlwm/UGATIT-pytorch (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:59.145189+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 2026-08-30T02:25:14.102882+00:00, confidence not recorded.
  - readme: https://github.com/znxlwm/UGATIT-pytorch (fetched 2026-08-28T04:06:59.145189+00:00, sha 09a169d2d9e5)
- Data as of 2026-08-30T08:39:29.467469+00:00.