Skip to content
Pytorch implementation of paper: AttnGAN Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


Pytorch implementation for reproducing AttnGAN results in the paper AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks.


python 2.7


In addition, please add the project folder to PYTHONPATH and pip install the following packages:

  • python-dateutil, easydict, pandas, torchfile, nltk, scikit-image


  • Download metadata (text and filename):

    • Download preprocessed metadata for coco and save them to data/
    • Extract caption files in '''' and '''' to data/coco/text/
    • [Optional] If you want to use the per-trained models, please download the dictionary, captions.pickle, otherwise it will be generated by
  • Download images:

    • Download coco dataset and extract both train2014 and val2014 images to data/coco/images/


  • Pre-train DAMSM models:

    • For coco dataset: python --cfg cfg/DAMSM/coco.yml --gpu 0
  • Train AttnGAN models:

    • For coco dataset: python --cfg cfg/coco_attn2.yml --gpu 0
  • *.yml files are example configuration files for training/evaluation our models.

Pretrained Model


  • Run python --cfg cfg/eval_coco.yml --gpu 1 to generate examples from captions in files listed in "./data/coco/example_filenames.txt". Results are saved to DAMSMencoders/.
  • Change the eval_*.yml files to generate images from other pre-trained models.
  • Input your own sentence in "./data/coco/example_captions.txt" if you wannt to generate images from customized sentences.


  • To generate images for all captions in the validation dataset, change B_VALIDATION to True in the eval_*.yml. and then run python --cfg cfg/eval_coco.yml --gpu 1
  • We compute inception score for models trained on coco using improved-gan/inception_score.

Examples generated by AttnGAN [Blog]

bird example coco example

Citing AttnGAN

If you find AttnGAN useful in your research, please consider citing:

  author    = {Tao Xu, Pengchuan Zhang, Qiuyuan Huang, Han Zhang, Zhe Gan, Xiaolei Huang, Xiaodong He},
  title     = {AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks},
  Year = {2018},
  booktitle = {{CVPR}}
You can’t perform that action at this time.