book

Machine Learning Engineering with Python - Second Edition

by Andrew P. McMahon

August 2023

Intermediate to advanced

462 pages

11h 20m

English

Packt Publishing

Read now

Unlock full access

Who this book is forWhat this book coversTo get the most out of this bookGet in touch
Technical requirementsDefining a taxonomy of data disciplinesData scientistML engineerML operations engineerData engineerWorking as an effective teamML engineering in the real worldWhat does an ML solution look like?Why Python?High-level ML system designExample 1: Batch anomaly detection serviceExample 2: Forecasting APIExample 3: Classification pipelineSummary
Technical requirementsSetting up our toolsSetting up an AWS accountConcept to solution in four stepsComparing this to CRISP-DMDiscoverUsing user storiesPlayDevelopSelecting a software development methodologyPackage management (conda and pip)PoetryCode version controlGit strategiesModel version controlDeployKnowing your deployment optionsUnderstanding DevOps and MLOpsBuilding our first CI/CD example with GitHub ActionsContinuous model performance testingContinuous model trainingSummary
Technical requirementsDefining the model factoryLearning about learningDefining the targetCutting your lossesPreparing the dataEngineering features for machine learningEngineering categorical featuresEngineering numerical featuresDesigning your training systemTraining system design optionsTrain-runTrain-persistRetraining requiredDetecting data driftDetecting concept driftSetting the limitsDiagnosing the driftRemediating the driftOther tools for monitoringAutomating trainingHierarchies of automationOptimizing hyperparametersHyperoptOptunaAutoMLauto-sklearnAutoKerasPersisting your modelsBuilding the model factory with pipelinesScikit-learn pipelinesSpark ML pipelinesSummary
Technical requirementsWriting good PythonRecapping the basicsTips and tricksAdhering to standardsWriting good PySparkChoosing a styleObject-oriented programmingFunctional programmingPackaging your codeWhy package?Selecting use cases for packagingDesigning your packageBuilding your packageManaging your environment with MakefilesGetting all poetic with PoetryTesting, logging, securing, and error handlingTestingSecuring your solutionsAnalyzing your own code for security issuesAnalyzing dependencies for security issuesLoggingError handlingNot reinventing the wheelSummary
Technical requirementsArchitecting systemsBuilding with principlesExploring some standard ML patternsSwimming in data lakesMicroservicesEvent-based designsBatchingContainerizingHosting your own microservice on AWSPushing to ECRHosting on ECSBuilding general pipelines with AirflowAirflowAirflow on AWSRevisiting CI/CD for AirflowBuilding advanced ML pipelinesFinding your ZenMLGoing with the KubeflowSelecting your deployment strategySummary
Technical requirementsScaling with SparkSpark tips and tricksSpark on the cloudAWS EMR exampleSpinning up serverless infrastructureContainerizing at scale with KubernetesScaling with RayGetting started with Ray for MLScaling your compute for RayScaling your serving layer with RayDesigning systems at scaleSummary
Going deep with deep learningGetting started with PyTorchScaling and taking deep learning into productionFine-tuning and transfer learningLiving it large with LLMsUnderstanding LLMsConsuming LLMs via APICoding with LLMsBuilding the future with LLMOpsValidating LLMsPromptOpsSummary
Technical requirementsUnderstanding the forecasting problemDesigning our forecasting serviceSelecting the toolsTraining at scaleServing the models with FastAPIResponse and request schemasManaging models in your microservicePulling it all togetherContainerizing and deploying to KubernetesContainerizing the applicationScaling up with KubernetesDeployment strategiesSummary
Technical requirementsUnderstanding the batch processing problemDesigning an ETML solutionSelecting the toolsInterfaces and storageScaling of modelsScheduling of ETML pipelinesExecuting the buildBuilding an ETML pipeline with advanced Airflow featuresSummary

Content preview from Machine Learning Engineering with Python - Second Edition

8 Building an Example ML Microservice

This chapter will be all about bringing together some of what we have learned in the book so far with a realistic example. This will be based on one of the scenarios introduced in Chapter 1, Introduction to ML Engineering, where we were required to build a forecasting service for store item sales. We will discuss the scenario in a bit of detail and outline the key decisions that have to be made to make a solution a reality, before showing how we can employ the processes, tools, and techniques we have learned through out this book to solve key parts of the problem from an ML engineering perspective. By the end of this chapter, you should come away with a clear view of how to build your own ML microservices ...