LAPGAN

this project implements algorithm LAPGAN for high resolution image generation introduced in paper Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks.

install prerequisite packages

install with command

pip3 install -r requirements.txt

how to train

train with the following command

python3 train.py --batch_size=<batch size> --num_workers=<workers> --download=(True|False) [--gpus=<gpu number>]

if you train with GPU, --gpus must be over zero.

if you want to train with pretrained checkpoint, run with command

python3 train.py --batch_size=<batch size> --num_workers=<workers> --download=(True|False) --checkpoint=<path/to/ckpt/file> [--gpus=<gpu number>]

how to save model from checkpoint

save model with the following command

python3 save_model.py --checkpoint=<path/to/ckpt/file>

the saved models will be placed under directory models.