SAS deep learning with Python made easy using DLPy


I look forward to Pi Day every year at SAS because it's a day of celebration including yummy pies and challenging games that challenge you to recall the digits after the decimal point in Pi.  Plus you get to wear Pi t-shirts (I have about seven). Today, though we are going to talk about DLPy which sounds like Pi and Pie but really stands for Deep Learning with Python.  

DLPy is an open source package that data scientists can download to apply SAS deep learning algorithms to image, text and audio data. And you don’t need to write SAS  code to reap the benefits of deep learning. DLPy is a toolset in a Python-style shell to SAS scripting language and the SAS deep learning actions from SAS® Visual Data Mining and Machine Learning.

DLPy is available in SAS Viya 3.4 and accessed via Jupyter Notebook. DLPy is designed to provide an efficient way to apply deep learning functionalities to solve computer vision, natural language processing, forecasting, and speech processing problems. DLPy APIs are created following the Keras APIs with a touch of PyTorch flavor. This library is available on GitHub.

Deep learning with DLPy

DLPy features include the following:

  • Read in and build deep learning models for image, text, audio and time series data.
  • 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.
  • Processing audio files and training deep learning algorithms to create a language model for speech recognition applications.
  • Predefined network architectures such as LeNet, VGG, ResNet, DenseNet, Darknet, Inception and YoloV2 and Tiny_Yolo. And many are provided with pre-trained weights!
  • Enhanced data visualization such as heat maps and feature maps to aid in the interpretation of deep learning computer vision models.
  • Import and export deep learning models in ONNX format.

We’ve created a series of six videos to introduce DLPy. We suggest you start with the introduction below and then move to your area of interest. I'll include the introduction here and link to all the videos below.

  1. Introduction to the series
  2. Image Classification with Convolutional Neural Networks
  3. Object Detection With Tiny YOLOv2 Model
  4. Import and Export Deep Learning Models With ONNX
  5. Text Classification and Text Generation With Recurrent Neural Networks
  6. Time Series Forecasting With Recurrent Neural Networks

With DLPy, you can have your Py!

DLPy makes it easy to take advantage of deep learning with its preconfigured models. Simply choose your model, modify it and begin deep learning. And if you want to contribute to the DLPy library, create a pull request as SAS gladly accepts pull requests on GitHub.

Just a note, to use the package for model development, a SAS Visual Data Mining and Machine Learning license is required.

Learn how to do deep learning with SAS


About Author

Susan Kahler

Global Product Marketing Manager for AI

Susan is a Global Product Marketing Manager for AI at SAS. She has her Ph.D. in Human Factors and Ergonomics, having used analytics to quantify and compare mental models of how humans learn complex operations. Throughout her well-rounded career, she has held roles in user centered design, product management, customer insights, consulting and operational risk. Susan recently completed her Master of Science in Analytics, focusing on healthcare analytics. She also holds a patent for a software navigation system to guide users through dynamically changing systems.

Related Posts


  1. Pingback: Understanding object detection in deep learning - The SAS Data Science Blog

  2. Pingback: Building a YOLO object detection model using SAS - The SAS Data Science Blog

Leave A Reply

Back to Top