Featured presentation: Optimizing, Profiling, and Deploying High Performance Spark ML, Scikit-Learn, and TensorFlow AI Models in Production with GPUs
by Chris Fregly (http://linkedin.com/in/cfregly), Research Engineer @ PipelineIO (http://pipeline.io/)
Using the latest advancements from TensorFlow including the Accelerated Linear Algebra (XLA) Framework, JIT/AOT Compiler, and Graph Transform Tool, I’ll demonstrate how to optimize, profile, and deploy TensorFlow Models in GPU-based production environment.
This talk is contains many Spark ML and TensorFlow AI demos using PipelineIO's 100% Open Source Community Edition. All code and Docker images are available to reproduce on your own CPU or GPU-based cluster.
Chris Fregly (https://linkedin.com/in/cfregly) is Founder and Research Engineer at PipelineIO (http://pipeline.io/), a Streaming Machine Learning and Artificial Intelligence Startup based in San Francisco. He is also an Apache Spark Contributor, a Netflix Open Source Committer, founder of the Global Advanced Spark and TensorFlow Meetup, author of the O’Reilly Training and Video Series titled, "High Performance TensorFlow in Production."
Previously, Chris was a Distributed Systems Engineer at Netflix, a Data Solutions Engineer at Databricks, and a Founding Member and Principal Engineer at the IBM Spark Technology Center in San Francisco.
Accelerated TensorFlow development with packages
by Garrett Smith, project lead for Guild AI (https://guild.ai)
Software package managers such as Homebrew, apt, and rpm letdevelopers quickly install and use software. They provide an essentialrole in accelerating software development through component reuse. Isit possible to apply similar patterns and tooling to TensorFlowdevelopment? In this talk, Garrett Smith, project lead of Guild AI,will present a specification for code reuse in TensorFlow thatincludes pre-trained models, datasets, and project templates. He'lldemonstrate Guild's packaging command set, which can be used to build,deploy, and install TensorFlow packages, saving developers timegetting started with their projects.
Garrett Smith is project lead for Guild AI, an open source toolkit that provides essential productivity features for TensorFlow development. Garrett has over twenty years of software development experience and has expertise in building reliable, distributed back-end systems and in operations. Prior to founding Guild AI, Garrett led CloudBees PaaS division, which hosted hundreds of thousands of Java applications at scale. Garrett is a frequent instructor and speaker at software conferences and an active contributor to several prominent open source projects.
R bindings for Keras
by Rajiv Shah
This talk introduces the new Keras interface for R. This is significant, because it opens up all the great innovation using Keras with a Tensorflow backend. The talk will walk through a simple image classification project using this new interface.
With the Keras interface for R, it is possible to build a convolutional neural network, augment the data, use pre-trained networks, and save the model weights. These are all very important functions and very simple to do with the Keras interface. As a result, it will now be possible to build production quality models from R.
Rajiv Shah (https://www.linkedin.com/in/rcshah?authType=NAME_SEARCH&authToken=TVD5&locale=en_US&srchid=854662381470146485173&srchindex=1&srchtotal=1&trk=vsrp_people_res_name&trkInfo=VSRPsearchId%3A854662381470146485173%2CVSRPtargetId%3A15452467%2CVSRPcmpt%3Aprimary%2CVSRPnm%3Atrue%2CauthType%3ANAME_SEARCH) is a senior data scientist at Caterpillar and an Adjunct Assistant Professor at the University of Illinois at Chicago. He is an active member of the data science community in Chicago with an interest into public policy issues, such as surveillance in Chicago. He has a PhD from the University of Illinois at Urbana Champaign.
Thanks to DePaul University
Our warm thanks to DePaul University for hosting this month's meeting!