From Experimentation to Products: The Production Machine Learning Journey • Robert Crowe • GOTO 2021

This presentation was recorded at GOTOpia February 2021. #GOTOcon #GOTOpia http://gotopia.eu Robert Crowe - TensorFlow Developer Advocate at Google @robertcrowe3724 ABSTRACT A machine learning (ML) journey typically starts with trying to understand the world, and looking for data that describes it. This leads to an experimentation phase, where we try to use that data to model the parts of the world that we’re interested in, often because they directly affect our users or our business. Once we have one or more models that deliver good results, it’s time to move those models into production. Deploying advanced machine learning technology to serve customers and/or business needs requires a rigorous approach and production-ready systems. This is especially true for maintaining and improving model performance over the lifetime of a production application. Unfortunately, the issues involved and approaches available are often poorly understood. A ML application in production must address all of the issues of modern software development methodology, as well as issues unique to ML and data science. Often ML applications are developed using tools and systems which suffer from inherent limitations in testability, scalability across clusters, training/serving skew, and the modularity and reusability of components. In addition, ML application measurement often emphasizes top level metrics, leading to issues in model fairness as well as predictive performance across user segments. In this talk, Robert will discuss the use of ML pipeline architectures for implementing production ML applications, and in particular we review Google’s experience with TensorFlow Extended (TFX), as well as the advantages of containerizing pipeline architectures using platforms such as Kubeflow. Google uses TFX for large scale ML applications, and offers an open-source version to the community. TFX scales to very large training sets and very high request volumes, and enables strong software methodology [...] TIMECODES 00:00 Intro 02:15 Production ML 05:41 We need MLOps 06:21 Continuous integration, deployment and testing 07:29 MLOps level 0: Manual Process 09:02 Experiment 12:11 Tales from the trenches 13:02 TensorFlow Extended (TFX) 14:28 TFX production components 16:43 What is a TFX component? 18:20 TFX orchestration 19:16 Difference between TFX & Kubeflow pipelines 23:00 Distributed pipeline processing: Apache Beam 25:28 TFX standard components 25:53 Components: ExampleGen, StatisticsGen & SchemaGen 28:17 Components: ExampleValidator, Transform & Trainer 31:45 Components: Tuner, Evaluator & InfraValidator 32:51 Components: Pusher & BulkInferrer 33:37 TFX pipeline nodes 34:43 TRFX custom components 36:09 Very high level architecture 37:03 Outro Download slides and read the full abstract here: https://gotopia.eu/february-2021/sessions/1689/from-experimentation-to-products-the-production-ml-journey RECOMMENDED BOOKS Holden Karau, Trevor Grant, Boris Lublinsky, Richard Liu & Ilan Filonenko • Kubeflow for Machine Learning • https://amzn.to/3JVngcx Phil Winder • Reinforcement Learning • https://amzn.to/3t1S1VZ Kelleher & Tierney • Data Science (The MIT Press Essential Knowledge series) • https://amzn.to/3AQmIRg Lakshmanan, Robinson & Munn • Machine Learning Design Patterns • https://amzn.to/2ZD7t0x Lakshmanan, Görner & Gillard • Practical Machine Learning for Computer Vision • https://amzn.to/3m9HNjP Aurélien Géron • Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow • https://amzn.to/2XZaQy8 https://twitter.com/GOTOcon https://www.linkedin.com/company/goto- https://www.facebook.com/GOTOConferences #MachineLearning #ML #TensorFlow #TF #TFX #TensorFlowExtended #Kubeflow #AI #ArtificialIntelligence #DataScience #MLOps #CI #ContinuousIntegration #Testing #Orchestration #ApacheBeam #ExampleGen #StatisticsGen #SchemaGen Looking for a unique learning experience? Attend the next GOTO conference near you! Get your ticket at https://gotopia.tech SUBSCRIBE TO OUR CHANNEL - new videos posted almost daily. https://www.youtube.com/user/GotoConferences/?sub_confirmation=1