# taoxugit/AttnGAN

Repository: https://github.com/taoxugit/AttnGAN
Canonical: https://ross.abutalabs.com/products/attngan
Language: Python
License: MIT
License Family: permissive
Last push: 2024-07-25T04:54:47+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3117, "days_push": 769, "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 1363, forks 425 (observed 2026-08-28T04:04:30.410349+00:00)

## What it is
A PyTorch implementation of AttnGAN, a fine-grained text-to-image generative adversarial network with multi-stage attention mechanisms from a CVPR 2018 paper. It includes training scripts, pretrained models for the CUB birds and COCO datasets, and sampling code to generate images from text captions.

## Use cases
- generate images from text descriptions
- reproduce AttnGAN paper results
- train a text-to-image GAN on birds or COCO dataset
- experiment with attentional GAN architectures
- generate bird images from custom captions
- research fine-grained text-to-image synthesis

## When to choose
- you need to reproduce or build on the AttnGAN CVPR 2018 paper
- you want pretrained text-to-image GAN models for birds or COCO
- you are studying attention mechanisms in GANs for research

## When to avoid
- you need modern high-quality text-to-image generation (use diffusion models instead)
- you need a production-ready or maintained product
- you cannot use Python 2.7-era dependencies or older PyTorch versions

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: machine-learning, deep-learning, computer-vision
- platform: python
- tags: text-to-image, gan, pytorch, attention, research-code, cvpr-2018, natural-language-processing, linux, gpu

## Member repositories
- taoxugit/AttnGAN (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:30.410349+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 2026-08-30T04:41:27.011666+00:00, confidence not recorded.
  - readme: https://github.com/taoxugit/AttnGAN (fetched 2026-08-28T04:04:30.410349+00:00, sha 3b1a24c8cc67)
- Data as of 2026-08-30T08:39:29.467469+00:00.