Amazon Creates New Tool To Engineer AI Models With Just A Few Lines Of Code

Navin Bondade 0 Comments Artificial Intelligence, Data Science, Deep Learning, Machine Learning

Building a machine learning or deep learning model is a complex and time-consuming task. It requires feature engineering, data must be transformed and preprocessed to create the features that contribute AI algorithms performance after that testing the model, customizing and optimizing it to generate accurate results.

Amazon Web Services recently launched an open-source library called AutoGluon that can cut down most of the developer’s workload, with AutoGluon one can write AI-imbued apps with only a few lines of code.

AutoGluon lets developers harness machine learning models with image, text, or tabular data sets, without. the need to manually experiment. Developers can achieve strong, predictive performance in their model.

AutoGluon eliminates most of the cumbersome processes and delivers developers a truly hands-off-the-wheel experience AutoGluon helps by creating a model with as few as three lines of code by automatically tuning choices within default ranges that are known to perform well for a given task.

AutoGluon is designed to be an easy-to-use and easy-to-extend AutoML toolkit, suitable for both beginners and experts. It enables prototyping deep learning models with a few lines, automatic hyperparameter tuning, model selection, and data processing; and automatic utilization of SOTA deep learning models.

After importing the AutoGluon package, developers can simply specify a task of interest, load the appropriate dataset, and finally have AutoGluon quickly and automatically train many models under thousands of different hyperparameter configurations and then return the best model.

Developers simply specify when they’d like to have their trained model ready, and in response, AutoGluon leverages available compute resources to find the strongest model within the allotted runtime.

“We developed AutoGluon to truly democratize machine learning, and make the power of deep learning available to all developers,” AWS scientist Jonas Mueller said. “AutoGluon solves this problem as all choices are automatically tuned within default ranges that are known to perform well for the particular task and model.”

AutoGluon endeavors to make both the creation of the neural net architecture and the selection of appropriate hyperparameters easier. The AutoGluon framework help researchers to customize and improve their existing bespoke models and data pipelines.

One of the applications of AutoGluon is object detection in images, achieved through target identification and localization within a bounding box. The author of AutoGluon Tabular and his colleagues used AutoGluon to train an object detector on a small toy dataset generated using the motorbike category of the VOC dataset.

The task was to localize motorbikes in the given pictures. With a single call to fit() and using predict() to test the model, AutoGluon was able to generate a reasonably accurate visualization image (shown below).

AutoGluon is currently capable of creating models for image classification, text classification, object detection, and tabular prediction. AutoGluon’s API is also intended to allow more experienced developers to be able to customize the auto-generated model and improve performance.

At the moment, AutoGluon is only available for Linux and it requires Python version 3.6 or 3.7 but Amazon says that Mac OSX and Windows versions will be available soon.

In explaining the reasoning behind AutoGluon, Amazon said the deployment of deep learning models with state-of-the-art inferencing accuracy typically has required extensive expertise. Developers have had to invest a considerable amount of time and effort into training deep learning models.

Despite advancements such as the Keras library, for more easily specifying parameters and layers in deep learning models, developers have had to grapple with complex issues such as hyperparameter tuning. AutoGluon is intended to democratize machine learning and make deep learning available to all developers.

AutoGluon is based on work initially begun by Microsoft and Amazon in 2017. The original Gluon was a machine learning interface designed to let developers mix and matched optimized components to create their own models, but AutoGluon just creates a model end-to-end, based on the desires of the user.

The AutoGluon project is on GitHub.