Amazon SageMaker accelerates local ML code conversion to remote jobs
Amazon SageMaker Python SDK is an open source library for training and deploying machine-learning models on Amazon SageMaker. We are excited to announce SageMaker Python SDK now helps data scientists execute any local ML code authored in their preferred IDE and local notebooks along with the associated runtime dependencies as large-scale ML model training jobs with minimal code changes.
Amazon SageMaker Python SDK is an open source library for training and deploying machine-learning models on Amazon SageMaker. We are excited to announce SageMaker Python SDK now helps data scientists execute any local ML code authored in their preferred IDE and local notebooks along with the associated runtime dependencies as large-scale ML model training jobs with minimal code changes.
Data scientists need to add only a line of code (a Python decorator) to their local ML code and SageMaker Python SDK takes their code, datasets, and workspace environment setup and runs it as a SageMaker Training job. This decorator mode helps data scientists start their ML workflows on SageMaker more easily by reducing the need for custom code constructs and environment variable management. Further, this enhancement to local code to jobs experience reduces the time spent on container management via auto-capture and replication of local runtime so that data scientists can spend lesser time recreating their local environment in production-grade jobs.
This feature is now is available in all regions where Amazon SageMaker Python SDK is available. To get started with the new feature, read the Amazon SageMaker documentation, Amazon SageMaker Python SDK guide and the SageMaker Model Training page.
To checkout examples and learn more about how the local code to training jobs SDK interface can be used, visit the ML blog and checkout the sample notebooks in the SageMaker Python SDK repository.