News
With AutoGluon, AWS Promises To Simplify Deep Learning for Non-Experts
Amazon Web Services (AWS) has launched several solutions in recent years that promise to "democratize" machine learning, from consulting services to courseware to products like SageMaker, DeepLens and Deep Learning Containers.
AutoGluon, announced Thursday, is the latest to join the growing roster of AWS products aimed at making machine learning -- specifically deep learning -- more accessible to developers with little experience in that area.
AutoGluon is an open source toolkit (available on GitHub here) that's designed for "developers building applications involving machine learning with image, text, or tabular data sets," AWS said in its announcement.
AutoGluon aims to help developers navigate the difficulties of creating deep learning models in particular. Software libraries like Keras and Theano have made it easier to harness deep learning with without laborious coding work, according to AWS, but developers still need help hurdling complex tasks such as "hyperparameter tuning, data pre-processing, neural-architecture search, and decisions related to leveraging transfer learning."
Enter AutoGluon, which AWS promises will enable developers to create a neural network model with very minimal coding -- as few as three lines, according to the announcement. It does this by automating the more complex processes involved in creating a model -- processes that had previously been the domain of very few and very skilled deep learning experts.
"There's no need for developers to manually experiment with the hundreds of individual choices that must be made while designing a deep learning model," AWS said. "Rather, they can simply specify when they would like to have their trained model ready. In response, AutoGluon leverages the available compute resources to find the strongest model within its allotted run-time."
More information is available on the AutoGluon Web site here, including tutorials and resources for beginner and advanced developers alike.