Use the run_mnist.py script for training, and validation and test distribution networks for MNIST dataset.
python run_mnist.py <options>
python run_mnist.py -h
usage: run_mnist.py [-h] [--seed N] [-b N] [--path PATH] [--savepath
[--start-epoch N] [--gpu N] [--action ACTION]
Training the distribution network for MNIST
optional arguments:
-h, --help show this help message and exit
--seed N the random seed
-b N, --batch-size N mini-batch size (default: 128)
--path PATH data path
--savepath SAVEPATH path to save models
-d N, --dim N the dimention of latent space (default:10)
-s SIGMASHAPE, --sigmashape SIGMASHAPE
the form of covariance matrix
-e N, --epochs N number of total epochs to run
--start-epoch N manual epoch number (useful on restarts)
--gpu N the GPU number (default auto schedule)
--action ACTION train or test (default train-test)
Use the run_cifar.py script for training, and validation and test distribution networks for CIFAR10 dataset.
python run_cifar.py <options>
python run_cifar.py -h
usage: run_cifar.py [-h] [--seed N] [-b N] [--path PATH] [--savepath SAVEPATH] [-d N] [-s SIGMASHAPE] [-e N]
[--start-epoch N] [--gpu N] [--action ACTION]
Training the distribution network for CIFAR10
optional arguments:
-h, --help show this help message and exit
--seed N the random seed
-b N, --batch-size N mini-batch size (default: 128)
--path PATH data path
--savepath SAVEPATH path to save models
-d N, --dim N the dimention of latent space (default:10)
-s SIGMASHAPE, --sigmashape SIGMASHAPE
the form of covariance matrix
-e N, --epochs N number of total epochs to run
--start-epoch N manual epoch number (useful on restarts)
--gpu N the GPU number (default auto schedule)
--action ACTION train or test (default train-test)
Use the run_ohsu.py script for training, and validation and test distribution networks for Ohsumed dataset.
python run_ohsu.py <options>
python run_ohsu.py -h
usage: run_ohsu.py [-h] [--seed N] [-b N] [--path PATH] [--wordvecfile WORDVECFILE] [--savepath SAVEPATH] [-d N]
[-s SIGMASHAPE] [-e N] [--start-epoch N] [--gpu N] [--action ACTION]
Training the distribution network for Ohsumed
optional arguments:
-h, --help show this help message and exit
--seed N the random seed
-b N, --batch-size N mini-batch size (default: 128)
--path PATH data path
--wordvecfile WORDVECFILE
wordvector files
--savepath SAVEPATH path to save models
-d N, --dim N the dimention of latent space (default:10)
-s SIGMASHAPE, --sigmashape SIGMASHAPE
the form of covariance matrix
-e N, --epochs N number of total epochs to run
--start-epoch N manual epoch number (useful on restarts)
--gpu N the GPU number (default auto schedule)
--action ACTION train or test (default train-test)