Author: Zen and the Art of Computer Programming
1 Introduction
On January 27, Amazon SageMaker Studio finally released the official version, which can turn the SageMakerStudio experience into a new development tool. This article is based on the development entry series of the latest version of SageMaker Studio, and brings you a guide for beginners. I hope that reading this article can help readers better understand SageMakerStudio and quickly get started with machine learning-related development practices.
On July 10, 2019, Amazon SageMaker announced the launch of Studio, which provides an interactive ML workspace, enabling data scientists, AI engineers, analysts, decision makers, and business users to build, train, and deploy. The tool includes:
- SageMaker Notebook: Data scientists can write code, run notebooks, perform model training and development, and more in the SageMaker Notebook environment.
- SageMaker Experiments: Provides the ability to manage the ML lifecycle and track experimental results. Data scientists, AI engineers, analysts, and decision makers can use this function to record and share experiments.
- SageMaker Model Registry: Provides data scientists with model registration, version control, search and discovery capabilities, allowing them to easily find, compare and use models trained in the past.
- SageMaker Pipelines: Automate machine learning workflows through pipelines, support data-based machine learning lifecycles, and allow data scientists, AI engineers, analysts, and decision makers to efficiently and reliably handle complex tasks.
- SageMaker Data Wrangler: Connect to S3 buckets or DynamoDB tables in the browser and handle data transformations without writing code.
This article mainly introduces the basic knowledge of SageMaker Studio Notebook components, including how to open Studio Notebook, write code, install dependent libraries, configure environment variables, and use Git version control