Practical Deep Learning at Scale with MLflow: Bridge the gap between offline experimentation and online production

Practical Deep Learning at Scale with MLflow
Foreword
Contributors
About the author
About the reviewers
Preface
	Who this book is for
	What this book covers
	To get the most out of this book
	Download the example code files
	Download the color images
	Conventions used
	Get in touch
	Share Your Thoughts
Section 1 - Deep Learning Challenges and MLflow Prime
Chapter 1: Deep Learning Life Cycle and MLOps Challenges
	Technical requirements
	Understanding the DL life cycle and MLOps challenges
		Implementing a basic DL sentiment classifier
		Understanding DL\'s full life cycle development
		Understanding MLOps challenges
	Understanding DL data challenges
	Understanding DL model challenges
	Understanding DL code challenges
	Understanding DL explainability challenges
	Summary
	Further reading
Chapter 2: Getting Started with MLflow for Deep Learning
	Technical requirements
	Setting up MLflow
		Setting up MLflow locally using miniconda
		Setting up MLflow to interact with a remote MLflow server
	Implementing our first DL experiment with MLflow autologging
	Exploring MLflow\'s components and usage patterns
		Exploring experiments and runs in MLflow
		Exploring MLflow models and their usages
		Exploring MLflow code tracking and its usages
	Summary
	Further reading
Section 2 – Tracking a Deep Learning Pipeline at Scale
Chapter 3: Tracking Models, Parameters, and Metrics
	Technical requirements
	Setting up a full-fledged local MLflow tracking server
	Tracking model provenance
		Understanding the open provenance tracking framework
		Implementing MLflow model tracking
	Tracking model metrics
	Tracking model parameters
	Summary
	Further reading
Chapter 4: Tracking Code and Data Versioning
	Technical requirements
	Tracking notebook and pipeline versioning
		Pipeline tracking
	Tracking locally, privately built Python libraries
	Tracking data versioning in Delta Lake
		An example of tracking data using MLflow
	Summary
	Further reading
Section 3 – Running Deep Learning Pipelines at Scale
Chapter 5: Running DL Pipelines in Different Environments
	Technical requirements
	An overview of different execution scenarios and environments
	Running locally with local code
	Running remote code in GitHub locally
	Running local code remotely in the cloud
	Running remotely in the cloud with remote code in GitHub
	Summary
	Further reading
Chapter 6: Running Hyperparameter Tuning at Scale
	Technical requirements
	Understanding automatic HPO for DL pipelines
		Types of hyperparameters and their challenges
		How HPO works and which ones to choose
	Creating HPO-ready DL models with Ray Tune and MLflow
		Setting up Ray Tune and MLflow
		Creating the Ray Tune trainable for the DL model
		Creating the Ray Tune HPO run function
	Running the first Ray Tune HPO experiment with MLflow
	Running HPO with Ray Tune using Optuna and HyperBand
	Summary
	Further reading
Section 4 – Deploying a Deep Learning Pipeline at Scale
Chapter 7: Multi-Step Deep Learning Inference Pipeline
	Technical requirements
	Understanding patterns of DL inference pipelines
		Understanding the MLflow Model Python Function API
	Implementing a custom MLflow Python model
	Implementing preprocessing and postprocessing steps in a DL inference pipeline
		Implementing language detection preprocessing logic
		Implementing caching preprocessing and postprocessing logic
		Implementing response composition postprocessing logic
	Implementing an inference pipeline as a new entry point in the main MLproject
	Summary
	Further reading
Chapter 8: Deploying a DL Inference Pipeline at Scale
	Technical requirements
	Understanding different deployment tools and host environments
	Deploying locally for batch and web service inference
		Batch inference
		Model as a web service
	Deploying using Ray Serve and MLflow deployment plugins
	Deploying to AWS SageMaker – a complete end-to-end guide
		Step 1: Build a local SageMaker Docker image
		Step 2: Add additional model artifacts layers onto the SageMaker Docker image
		Step 3: Test local deployment with the newly built SageMaker Docker image
		Step 4: Push the SageMaker Docker image to AWS Elastic Container Registry
		Step 5: Deploy the inference pipeline model to create a SageMaker endpoint
		Step 6: Query the SageMaker endpoint for online inference
	Summary
	Further reading
Section 5 – Deep Learning Model Explainability at Scale
Chapter 9: Fundamentals of Deep Learning Explainability
	Technical requirements
	Understanding the categories and audience of explainability
		Audience: who needs to know
		Stage: when to provide an explanation in the DL life cycle
		Scope: which prediction needs explanation
		Input data format: what is the format of the input data
		Output data format: what is the format of the output explanation
		Problem type: what is the machine learning problem type
		Objectives type: what is the motivation or goal to explain
		Method type: what is the specific post-hoc explanation method used
	Exploring the SHAP Explainability toolbox
	Exploring the Transformers Interpret toolbox
	Summary
	Further reading
Chapter 10: Implementing DL Explainability with MLflow
	Technical requirements
	Understanding current MLflow explainability integration
	Implementing a SHAP explanation using the MLflow artifact logging API
	Implementing a SHAP explainer using the MLflow pyfunc API
		Creating and logging an MLflow pyfunc explainer
		Deploying an MLflow pyfunc explainer for an EaaS
		Using an MLflow pyfunc explainer for batch explanation
	Summary
	Further reading
	Why subscribe?
Other Books You May Enjoy
	Packt is searching for authors like you
	Share Your Thoughts