STYLEGAN Procedure

PROC STYLEGAN Statement

PROC STYLEGAN IMAGESIZE=number <options>;

The PROC STYLEGAN statement invokes the procedure. Table 1 summarizes the options in this statement.

Table 1: PROC STYLEGAN Statement Options

Option	Description
DATA=	Specifies the name of the input table
DETERMINISTIC	Specifies deterministic cuDNN implementations
IMAGESIZE=	Specifies the image size in the training data. This option is required.
NUMSAMPLES=	Specifies the number of observations to generate
SCORESEED=	Specifies the random number seed to use for GAN scoring
SEED=	Specifies the random number seed to use for GAN training
USEGPU	Specifies that a GPU device is to be used
USEGPU(DEVICE=)	Specifies the ID of the GPU device that is to be used

You must specify the following option:

IMAGESIZE=number: specifies the image size (width of square images in pixels) in the training data. The number must be one of the values 4, 8, 16, 32, 64, 128, 256, 512, 1024.

You can also specify the following options:

DATA=libref.data-table

names the input data table for PROC STYLEGAN to use. The default is the most recently created data table. libref.data-table is a two-level name, where

libref: refers to a collection of information that is defined in the LIBNAME statement and includes the library, which includes a path to the data. For more information about libref, see the section Using SAS Viya Workbench.
data-table: specifies the name of the input data table.

DETERMINISTIC

tells the NVIDIA CUDA Deep Neural Network library (cuDNN) that you want only the deterministic implementations. Note: Specifying this option can significantly slow down the training process.

NUMSAMPLES=number

specifies the number of sample images to generate.

By default, NUMSAMPLES=1.

SCORESEED=number

specifies the seed to use for the random number generator for scoring. If you do not specify a seed or if you specify a value less than or equal to 0, the seed is generated by reading the time of day from the computer’s clock.

SEED=number

specifies the seed to use for the random number generator for training. If you do not specify a seed or if you specify a value less than or equal to 0, the seed is generated by reading the time of day from the computer’s clock.

USEGPU(suboptions)

specifies that a GPU device is to be used. You can specify the following suboption:

DEVICE=id-number: specifies the ID number of the GPU device that is to be used.

By default, DEVICE=0.

Last updated: September 23, 2026