From the course: Amazon SageMaker for Generative AI Applications
Join today to access over 25,500 courses taught by industry experts.
Challenge: Implement MLOps in SageMaker - Amazon SageMaker Tutorial
From the course: Amazon SageMaker for Generative AI Applications
Challenge: Implement MLOps in SageMaker
(upbeat music) - [Instructor] You've explored the pieces, model monitoring, versioning, automation, and Pipelines. Now it's time to bring it all together. In this hands-on challenge, you'll build a simplified MLOps workflow using SageMaker Pipelines. Your goal is to automate a few core concepts of the machine learning lifecycle, data loading, model fine-tuning, and model registration. First, you'll start by uploading the dataset you pre-processed in the 3-4 video, "Customize a Pre-Trained Model Challenge." Your file name was the formatted AmazonTraindata.txt file. Use that file. Next, you'll fine-tune a model using the uploaded dataset, and you'll register the train model into the SageMaker model registry with an approved status. Each step in the pipeline will be reusable, traceable, and automated, core principles of MLOps. You don't need to handle model deployment in this challenge. Instead, focus on building a pipeline that runs cleanly end to end and gives your team a repeatable…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.