PyTorch implementation of GAN-INT-CLS and AttnGAN
In addition, you may need other packages...
data_dir
|- COCO
|- filenames
|- train2014
|- val2014
|- text
|- train2014
|- val2014
|- image
|- train2014
|- val2014
Train DAMSM models:
python DAMSM_main.py
Train AttnGAN models:
python main.py
python main.py --help
)I prepared notebook for evaluation (AttnGAN/eval.ipynb).
You can evaluate generated images by
You can also generate images from your own captions.
AttnGAN
|- results
|- DAMSM/COCO/2019_05_04_00_32/model
|- image_encoder600.pth
|- text_encoder600.pth
|- AttnGAN/COCO/2019_05_14_17_08
|- model
|- G_epoch50.pth
|- config.txt