Amazon SageMaker

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
Amazon SageMaker
Developer(s)Amazon, Amazon Web Services
Initial release29 November 2017; 19 months ago (2017-11-29)
TypeSoftware as a service
Websiteaws.amazon.com/sagemaker

Amazon SageMaker is a cloud machine learning platform that was launched in November 2017.[1] SageMaker enables developers to create, train, and deploy machine learning (ML) models in the cloud.[2] SageMaker also enables developers to deploy ML models on embedded systems and edge-devices.[3][4]

Capabilities[edit]

SageMaker enables developers to operate at a number of levels of abstraction when training and deploying machine learning models. At its highest level of abstraction, SageMaker provides pre-trained ML models that can be deployed as-is.[5] In addition, SageMaker provides a number of built-in ML algorithms that developers can train on their own data.[6][7] Further, SageMaker provides managed instances of TensorFlow and Apache MXNet, where developers can create their own ML algorithms from scratch.[8] Regardless of which level of abstraction is used, a developer can connect their SageMaker-enabled ML models to other AWS services, such as the Amazon DynamoDB database for structured data storage,[9] AWS Batch for offline batch processing,[9][10] or Amazon Kinesis for real-time processing.[11]

Development interfaces[edit]

A number of interfaces are available for developers to interact with SageMaker. First, there is a web API that remotely controls a SageMaker server instance.[12] While the web API is agnostic to the programming language used by the developer, Amazon provides SageMaker API bindings for a number of languages, including Python, JavaScript, Ruby, and Java.[13][14] In addition, SageMaker provides managed Jupyter Notebook instances for interactively programming SageMaker and other applications.[15][16]

History and features[edit]

  • 2017-11-29: SageMaker is launched at the AWS re:Invent conference.[1][6][2]
  • 2018-02-27: Managed TensorFlow and MXNet deep neural network training and inference are now supported within SageMaker.[17][8]
  • 2018-02-28: SageMaker automatically scales model inference to multiple server instances.[18][19]
  • 2018-07-13: SageMaker adds support for recurrent neural network training, word2vec training, multi-class linear learner training, and distributed deep neural network training in Chainer with Layer-wise Adaptive Rate Scaling (LARS).[20][7]
  • 2018-07-17: AWS Batch Transform enables high-throughput non-realtime machine learning inference in SageMaker.[21][22]
  • 2018-11-08: Support for training and inference of Object2Vec word embeddings.[23][24]
  • 2018-11-27: SageMaker Ground Truth "makes it much easier for developers to label their data using human annotators through Mechanical Turk, third party vendors, or their own employees."[25][3]
  • 2018-11-28: SageMaker Reinforcement Learning (RL) "enables developers and data scientists to quickly and easily develop reinforcement learning models at scale."[26][3]
  • 2018-11-28: SageMaker Neo enables deep neural network models to be deployed from SageMaker to edge-devices such as smartphones and smart cameras.[27][3]
  • 2018-11-29: The AWS Marketplace is launched. The AWS Marketplace enables 3rd-party developers to buy and sell machine learning models that can be trained and deployed in SageMaker.[28]
  • 2019-01-27: SageMaker Neo is released as open-source software.[29]

Uses[edit]

  • NASCAR is using SageMaker to train deep neural networks on 70 years of video data.[30]
  • Carsales.com uses SageMaker to train and deploy machine learning models to analyze and approve automotive classified ad listings.[31]
  • Avis Budget Group and Slalom Consulting are using SageMaker to develop "a practical on-site solution that could address the over- and under-utilization of cars in real-time using an optimization engine built in Amazon SageMaker."[32]
  • Volkswagen Group uses SageMaker to develop and deploy machine learning in its manufacturing plants.[33]
  • Peak and Footasylum use SageMaker in a recommendation engine for footwear.[34]
  • Intel, H2O, and Persistent Systems provide SageMaker-enabled algorithms to other developers via the AWS Marketplace.[28]

Awards[edit]

In 2019, CIOL named SageMaker one of the "5 Best Machine Learning Platforms For Developers," alongside IBM Watson, Microsoft Azure Machine Learning, Apache PredictionIO, and ai-one.[35]

People[edit]

  • Alex Smola, Director of AI at AWS.[36][37]
  • Anima Anandkumar, former Principal Scientist in Amazon SageMaker group.[38]
  • Swami Sivasubramanian, Vice President of Amazon AI.[4]

See also[edit]

References[edit]

  1. ^ a b Miller, Ron (2017-11-29). "AWS releases SageMaker to make it easier to build and deploy machine learning models". TechCrunch. Retrieved 2019-06-09.
  2. ^ a b Woodie, Alex (2017-11-29). "AWS Takes the 'Muck' Out of ML with SageMaker". datanami. Retrieved 2019-06-09.
  3. ^ a b c d Rodriguez, Jesus (2018-11-30). "With These New Additions, AWS SageMaker is Starting to Look More Real for Data Scientists". Towards Data Science. Retrieved 2019-06-09.
  4. ^ a b Terdiman, Daniel (2018-10-05). "How AI is helping Amazon become a trillion-dollar company". Fast Company. Retrieved 2019-06-09.
  5. ^ Ponnapalli, Priya (2019-01-30). "Deploy trained Keras or TensorFlow models using Amazon SageMaker". AWS. Retrieved 2019-06-09.
  6. ^ a b "Introducing Amazon SageMaker". AWS. 2017-11-29. Retrieved 2019-06-09.
  7. ^ a b Nagel, Becky (2018-07-16). "Amazon Updates SageMaker ML Platform Algorithms, Frameworks". Pure AI. Retrieved 2019-06-09.
  8. ^ a b Roumeliotis, Rachel (2018-03-07). "How to jump start your deep learning skills using Apache MXNet". O'Reilly. Retrieved 2019-06-09.
  9. ^ a b Marquez, Ernesto. "Evaluate when to use added AWS Step Functions actions". TechTarget. Retrieved 2019-06-09.
  10. ^ "AWS Step Functions Adds Eight More Service Integrations". AWS. 2018-11-29. Retrieved 2019-06-09.
  11. ^ "Deploy Amazon SageMaker and a Data Lake on AWS for Predictive Data Science with New Quick Start". AWS. 2018-08-15. Retrieved 2019-06-09.
  12. ^ Olsen, Rumi (2018-07-19). "Call an Amazon SageMaker model endpoint using Amazon API Gateway and AWS Lambda". AWS. Retrieved 2019-06-09.
  13. ^ "Amazon SageMaker developer resources". AWS. Retrieved 2019-06-09.
  14. ^ Wiggers, Kyle (2018-11-21). "Amazon updates SageMaker with new built-in algorithms and Git integration". Retrieved 2019-06-09.
  15. ^ "Use Notebook Instances". AWS. Retrieved 2019-06-09.
  16. ^ Gift, Noah (2018-08-17). "Here Come The Notebooks". Forbes. Retrieved 2019-06-09.
  17. ^ "Amazon SageMaker now supports TensorFlow 1.5, Apache MXNet 1.0, and CUDA 9 for P3 Instance Optimization". AWS. 2018-02-27. Retrieved 2019-06-09.
  18. ^ "Auto Scaling in Amazon SageMaker is now Available". AWS. 2018-02-28. Retrieved 2019-06-09.
  19. ^ "Amazon Sagemaker Now Uses Auto-scaling". Polar Seven. 2018-03-24. Retrieved 2019-06-09.
  20. ^ "Amazon SageMaker Announces Several Enhancements to Built-in Algorithms and Frameworks". AWS. 2018-07-13. Retrieved 2019-06-09.
  21. ^ "Amazon SageMaker Now Supports High Throughput Batch Transform Jobs for Non-Real Time Inferencing". AWS. 2018-07-17. Retrieved 2019-06-09.
  22. ^ Simon, Julien (2019-01-24). "Making the most of your Machine Learning budget on Amazon SageMaker". Medium. Retrieved 2019-06-09.
  23. ^ "Introduction to Amazon SageMaker Object2Vec". AWS. 2018-11-08. Retrieved 2019-06-09.
  24. ^ "Amazon SageMaker Now Supports Object2Vec and IP Insights Built-in Algorithms". AWS. 2018-11-19. Retrieved 2019-06-09.
  25. ^ "Introducing Amazon SageMaker Ground Truth - Build Highly Accurate Training Datasets Using Machine Learning". AWS. 2018-11-28. Retrieved 2019-06-09.
  26. ^ "Introducing Reinforcement Learning Support with Amazon SageMaker RL". AWS. 2018-11-28. Retrieved 2019-06-09.
  27. ^ "Introducing Amazon SageMaker Neo - Train Once, Run Anywhere with up to 2x in Performance Improvement". AWS. 2018-11-28. Retrieved 2019-06-09.
  28. ^ a b Robuck, Mike (2018-11-29). "AWS goes deep and wide with machine learning services and capabilities". FierceTelecom. Retrieved 2019-06-09.
  29. ^ Janakiram, MSV (2019-01-27). "Amazon Open Sources SageMaker Neo To Run Machine Learning Models At The Edge". Forbes. Retrieved 2019-06-09.
  30. ^ Digman, Larry (2019-06-04). "NASCAR to migrate 18 petabytes of video archives to AWS". ZDNet. Retrieved 2019-06-09.
  31. ^ Crozier, Ry (2019-05-02). "Carsales builds Tessa AI to check vehicle ads". IT News. Retrieved 2019-06-09.
  32. ^ "Avis Budget Group and Slalom Further Digitize the Car Rental Process with Machine Learning on AWS". AWS. 2019-05-31. Retrieved 2019-06-09.
  33. ^ "Volkswagen and AWS Join Forces to Transform Automotive Manufacturing". Metrology News. 2019-05-24. Retrieved 2019-06-09.
  34. ^ Mari, Angelica (2019-05-14). "Footasylum steps up artificial intelligence to drive customer centricity". Computer Weekly. Retrieved 2019-06-09.
  35. ^ Pandey, Ashok (2019-02-21). "5 Best Machine Learning Platforms For Developers". CIOL. Retrieved 2019-06-09.
  36. ^ "Alex Smola". CrunchBase. Retrieved 2019-06-09.
  37. ^ Duberstein, Billy (2018-02-09). "The 3 Most Disruptive Accomplishments Inside Amazon's Earnings Report". The Motley Fool. Retrieved 2019-06-09.
  38. ^ Lahoti, Sugandha (2018-09-01). "Anima Anandkumar, the machine learning guru behind AWS bids adieu to AWS". Packt. Retrieved 2019-06-09.