Running machine learning systems in production is hard work. If you’re a data scientist or ML engineer responsible for building models, you shouldn’t spend all your time configuring complex infrastructure to deploy them. Yet that’s what happens in data science teams struggling to put their models in production.
But it doesn’t have to be this way.
In this course, I will teach you how to use Amazon SageMaker, an end-to-end machine learning platform, to build, deploy, and monitor machine learning models in production. The techniques taught in this course will help you build scalable, efficient, and fault-tolerant machine learning systems.
SageMaker is built to solve the challenges of operating machine learning models at scale.
• No more countless hours spent figuring out how to distribute ML algorithms over clusters.
• No more configuring networking rules, manually spinning up EC2 instances, or writing APIs.
• No more writing custom code to track experiment results.
SageMaker makes it easy to run large scale experiments, record experimental metadata, deploy optimized models, and monitor your models so you can focus on what you do best: building powerful predictive models.
Learning objectives:
• Train machine learning models using top frameworks like scikit-learn, XGBoost, Tensorflow, and PyTorch.
• Run large scale experiments like hyperparameter sweeps over a distributed cluster without any knowledge about EC2.
• Easily collect, store, and analyze data from experiments.
• Perform batch inference on datasets.
• Deploy trained models as API endpoints that automatically scale with demand.
• Monitor deployed endpoints to detect concept drift.
This course includes:
• A condensed overview of the challenges of running production machine learning systems.
• In-depth explanations of how Amazon SageMaker solves production ML challenges.
• A detailed walkthrough describing how to set up your own SageMaker Studio development environment and connect to a GitHub repository.
• Jupyter Notebooks containing sample code for training, deploying and monitoring ML models.
• Screencasts providing step-by-step walkthroughs of all code samples as well as the AWS console.
• Recommendations for setting up common ML workflows.
Course benefits:
• Equip yourself with best practices and patterns for production machine learning systems.
• Advance your career as an ML professional by gaining highly in demand MLOps skills.
• Learn directly from a trusted industry expert and advanced practitioner.
This course is for:
• ML engineers and data scientists who want to deploy their models to production.
• Managers and Leaders of Data Science teams who wish to enable efficient practices within their organizations.
• Software developers seeking to move into machine learning.
• Site Reliability Engineers (SREs) and other DevOps responsible for deploying ML models.
• Note: If you aren’t familiar with Amazon Web Services, no problem! We’ll provide overviews of the relevant services used in the course.
Course Curriculum
- 3.1 Introduction to Model Training in SageMaker (4:56)
- 3.2 Training an XGBoost model using Built-in Algorithms (15:57)
- 3.3 Training a scikit-learn model using Pre-built Docker Images and Custom Code (12:39)
- 3.4 Installing Custom python Requirements (7:25)
- 3.5 Preprocessing Data with SageMaker Preprocessor (8:28)
Enroll now and you'll instantly receive:
- 30 high-quality video lessons, broken up into 6 sharply focused modules.
- 6 Jupyter Notebooks showing you exactly how to use the SageMaker Python SDK to productionize your machine learning.
- Unlimited access to the course. View it wherever and whenever you like.
- 100% money-back guarantee. If you don't love it, I'll refund you. It's as simple as that.
Frequently Asked Questions
Q1: When does the course start and finish?
The prerecorded course content will be available for viewing beginning on August 1st, 2020. The course offerings that include Study Group sessions will be available for purchase until August 8th, 2020. The Study Group sessions will be held weekly on Saturdays from 12 - 1:30pm EST. The first session will be held on August 8th and the last session will be held on September 12th. If you purchase course offerings that include a private consultation you’ll be able to schedule that with Luigi after your purchase.
Q2: How long do I have access to the course?
The content is yours forever as soon as you register. You can register for it in advance, but please note that the course content will be available starting on August 1st, 2020.
Q3: What’s your refund policy?
You have 14 days after you enroll to request a full refund. Please contact Luigi if you have any issues.
Q4: How do I know if this course is right for me?
This course is for:
- ML Engineers and Data Scientists who want to deploy their models to production.
- Managers and Leaders of Data Science teams who wish to enable efficient practices within their organizations.
- Software developers seeking to move into machine learning.
- Site Reliability Engineers (SREs) and other DevOps responsible for deploying ML models.
You do NOT need any knowledge of Amazon Web Services in order to take this course. While a basic understanding of services like EC2 or S3 would be helpful, we'll provide an overview in the course.
Q5: Can you provide more information about the six Study Group Sessions?
The first virtual Study Group will be held on August 8th, 2020. Each week students will meet with Luigi for 90 minutes to discuss the contents of that week’s lesson. He will field questions from students and tie the lessons to real world situations and examples. The Study Group sessions are a great way for students to ask Luigi specific questions about the course and about running production machine learning systems. He loves engaging with students in these settings and works hard to ensure every student feels comfortable asking questions, regardless of his or her level of knowledge.
Q6: When does the early bird discount end?
The early bird pricing ends on August 1st, 2020.
Q7: I came to this course from TWIML. Is this a TWIML course?
This course has been developed by MLinProduction.com, presented by Luigi Patruno and was not developed by TWIML. Luigi has partnered with TWIML to bring this course to as many students as possible!
Q8. What programming languages and frameworks are used in the course?
This course utilizes Python 3 as the main programming language. In order to interact with Amazon SageMaker, we rely on the SageMaker Python SDK and the SageMaker Experiments Python SDK. Additionally, we'll train models using the scikit-learn, XGBoost, Tensorflow, and PyTorch frameworks and associated Python clients.
“I have been following Luigi Patruno's writings on all things production ML. I find them clear, practical, and insightful. He has a nice mailer as well. Please checkout.”
- Venkata Pingali, Scribble Data
“Luigi, your newsletter is the best and most useful in the field of ML productionizing. Keep doing your great work!”
- MLinProduction.com Reader
“Just wanted to thank you for creating excellent content. I learn a lot your articles."
- MLinProduction.com Reader
MLinProduction content has been featured on prominent data science sites and newsletters, including O'Reilly Data, The Institute for Ethical AI & Machine Learning, Nathan Benaich's Your Guide to AI Newsletter, and KDnuggets.
Hi, I’m Luigi.
I'm the Founder of MLinProduction.com.
I lead machine learning teams. In the past I've held roles as a data scientist, ML engineer, and data engineer. I've worked for large public companies and for tiny startups, taught graduate courses in data analysis and big data engineering, and have consulted for Fortune 100 companies.
I've had one consistent goal throughout my career: To build real machine learning systems that deliver massive amounts of business value. Now I'm teaching other ML practitioners how to do the same. I'm excited to be learning with you!