BigDL is one of the leading deep learning libraries for Apache Spark. Since it was released to the community in December 2016, it has evolved into a vibrant open source project led by Intel. From a functional perspective, BigDL is feature rich, with 200+ layers and loss functions, model interoperability support with popular DL frameworks like Caffe, Torch, TensorFlow, and Keras, OpenCV on Spark, easily deployable Docker container support, and a model zoo of widely used neural network algorithms. On the performance side, BigDL takes advantage of the capabilities of the new Intel Xeon processor scalable family and Intel Math Kernel Library for training (node scaling) and inferencing (leveraging model quantization features). There’s been an explosion of use cases built with BigDL over the last 12 months, from recommendation engines to customer-merchant propensity models to large-scale image similarity search and inferencing.
Sergey Ermolin details the latest features, real-world use cases, and what’s in store for 2018 for BigDL on Intel Xeon processor-based data center and cloud deployments.
Sergey Ermolin is a software solutions architect for deep learning, Spark analytics, and big data technologies at Intel. A Silicon Valley veteran with a passion for machine learning and artificial intelligence, Sergey has been interested in neural networks since 1996, when he used them to predict aging behavior of quartz crystals and cesium atomic clocks made by Hewlett-Packard. Sergey holds an MSEE and a certificate in mining massive datasets from Stanford and BS degrees in both physics and mechanical engineering from California State University, Sacramento.
©2018, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • email@example.com