Abstract Blue Angular gradient hero
DLPy model built using the Functional API

SAS DLPy Library
for developers

Getting started with SAS DLPy

The DLPy package provides a way to apply deep learning functionalities to solve the computer vision, NLP, forecasting and speech processing problems. DLPy allows for reading in and building deep learning models for image, text, audio and time series data. Keras and Caffe modelers will find the DLPy interface familiar.

High-level APIs for:

  • deep neural networks for tabular data
  • image classification and regression
  • object detection
  • RNN-based tasks – text classification, text generation and sequence labeling
  • RNN-based time series processing and modeling

Try DLPy for free!

  • DLPy is available on the VDMML trial image
  • nothing to download; a web browser is all you need
  • use your own data, or use our sample data and demo scripts
  • share the trial environment with four of your colleagues


>> DLPy Computer Vision Video Series

>> More DLPy Videos

Sample code

The sample code below establishs a connection and displays images from a data table. To see further sample code, visit the DLPy examples page on GitHub.

Display images from a data table


# Configure CAS and Import SWAT
In [1] : # You need to specify cashost and casport according to your own cas server.
           import swat
           sess = swat.CAS(cashost, casport)

# Import DLPy ImageTable Module
In [2] : from dlpy.images import ImageTable

# Specify Path to Images, Name and Load Image Table
In [3] : img_path='/disk/linux/dlpy/Giraffe_Dolphin'
         my_images = ImageTable.load_files(sess, path=img_path)

# Display Random Images from an Image Data Table
In [4] :  # randomly select 8 images from the table and arrange the display in 4 columns 
            my_images.show(nimages=8, ncol=4, randomize=True)
eight-animal-images

Back to Top