Released in 2019, Amazon SageMaker Studio provides one place for all end-to-end machine learning (ML) workflows, from data preparation, creation, and experimentation, to training, hosting, and monitoring. As we continue to innovate to increase data science productivity, we are excited to announce the enhanced SageMaker Studio experience, allowing users to select the managed integrated development environment (IDE) of their choice, while having access to resources and SageMaker Studio tools. through IDEs. This updated user experience (UX) gives data scientists, data engineers, and ML engineers more options for where to build and train their ML models within SageMaker Studio. As a web application, SageMaker Studio has improved load times, faster IDE and kernel startup times, and automatic updates.
In addition to JupyterLab and RStudio managed on Amazon SageMaker, we also launched management Visual Studio Open Source Code (Code-OSS) with SageMaker Studio. Once a user selects Code Editor and launches the Code Editor space backed by the compute and storage of their choice, they can take advantage of SageMaker tools and the Amazon Toolkit, as well as integration with Amazon EMR , Amazon CodeWhisperer, GitHub and the capacity. to personalize the environment with custom images. As they can do today with JupyterLab and RStudio in SageMaker, users can change the Code Editor calculation on the fly based on their needs.
Finally, to streamline the data science process and prevent users from having to jump from the console to Amazon SageMaker Studio, we added the ability to view training jobs and endpoint details in the user interface (UI ) of SageMaker Studio and enabled the ability to see all running instances in launched applications. Additionally, we enhanced our Jumpstart base model (FM) experience so users can quickly discover, import, register, tune, and deploy an FM.
Solution Overview
Launch IDE
With the new release of Amazon SageMaker Studio, the JupyterLab server is updated to provide faster startup times and a more reliable experience. SageMaker Studio is now a multi-tenant web application from where users can not only launch JupyterLab, but also have the option to launch Visual Studio Code open source (Code-OSS), RStudio and Canvas as managed applications. The SageMaker Studio user interface allows you to access and discover SageMaker resources and machine learning tools, such as jobs, endpoints, and pipelines, consistently, regardless of which IDE you choose.
SageMaker Studio contains a default private space that only you can access and run in JupyterLab or Code Editor.
You also have the option to create a new space in SageMaker Studio Classic, which will be shared with all users in your domain.
Improved machine learning workflow
With the new interactive experience, there are significant improvements and simplification of parts of the existing Amazon SageMaker machine learning workflow. Specifically, within Training and Hosting there is a much more intuitive UI-driven experience for creating new jobs and endpoints, while also providing metric tracking and monitoring interfaces.
Training
For training models in Amazon SageMaker, users can perform training of different types, either through a Studio Notebook through a Notebook job, a dedicated training job, or a tuning job through SageMaker JumpStart. With the improved UI experience, you can track current and past training jobs using the Studio Training dashboard.
You can also switch between specific training jobs to understand performance, the location of model artifacts, and also settings such as the hardware and hyperparameters behind a training job. The user interface also provides the flexibility of being able to start and stop training jobs through the console.
Accommodation
There are also a variety of different hosting options within Amazon SageMaker that you can use for model deployment within the user interface. To create a SageMaker endpoint, you can go to the Models section where you can use existing templates or create a new one.
Here you can use a single model to deploy an Amazon SageMaker real-time endpoint or multiple models to work with SageMaker's advanced hosting options.
Optionally, for FMs, you can also use the Amazon SageMaker JumpStart panel to toggle through the list of available FMs and fine-tune or deploy them through the user interface.
Setting
The updated Amazon SageMaker Studio experience launches alongside the Amazon SageMaker Studio Classic experience. You can try the new user interface and choose to subscribe to make the updated experience the default option for new and existing domains. The documentation lists the steps to migrate from SageMaker Studio Classic.
Conclusion
In this post, we show you the features available in the new and improved Amazon SageMaker Studio. With the updated SageMaker Studio experience, users now have the ability to select their preferred IDE backed by the compute of their choice and boot the kernel in seconds, with access to SageMaker tools and resources through the SageMaker Studio web app. . The addition of training details and endpoints within SageMaker Studio, as well as the enhanced Amazon SageMaker Jumpstart UX, provides seamless integration of ML steps within SageMaker Studio UX. Get started with SageMaker Studio here.
About the authors
Mair Hasco is an ai/ML specialist for Amazon SageMaker Studio. Helps customers optimize their machine learning workloads using Amazon SageMaker.
Ram Vegiraj is a machine learning architect on the SageMaker service team. He focuses on helping clients build and optimize their ai/ML solutions on Amazon SageMaker. In his free time he loves to travel and write.
Lauren Mullennex is a Senior Solutions Architect specializing in ai/ML at AWS. He has a decade of experience in DevOps, infrastructure and ML. She is also the author of a book on computer vision. In her free time she likes to travel and hike.
Khushboo Srivastava is a Senior Product Manager at Amazon SageMaker. She enjoys creating products that simplify machine learning workflows for customers and loves playing with her 1-year-old daughter.
