October 2020

23 October 2020

Features

New out of the box models are available

OS Packages > Image Analysis > ObjectDetection: This is a generic, retrainable deep learning model to perform Object Detection based on YOLO. This ML Package is pretrained on COCO Dataset so you can directly create an ML Skill which can be used for identifying 80 classes of COCO Dataset. Full documentation is there.

OS Packages > Tabular Data > TPOTAutoMLRegression and TPOTXGBoostRegression these two models were already available for classification problems they are now available for Regression as well. They will help you predict numerical values based on numerical tabular data. Full documentation available here.

Improvements

• Queued state of training pipeline is now split in two different states Packaging and Waiting for resources for better understanding of what is happening. Packaging state means that the docker image that will execute the training pipeline is being build. If this is the first time you are training this specific version of the ML Package this can takes up to 20 minutes. Waiting for resources means that there is no license available for the pipeline to execute. Pipeline checks every 5 minutes if new license is available (this will happen if you remove running ML Skills or if another Pipeline complet) and will start as soon as this is the case.
• Improving accessibility of our App: Accessibility is very important for us and a lot of progress has been made on making the entire app accessible this month. We are not perfect yet but we are getting better! Do not hesitate to let us know if you see anything missing.

Bug fix

Timeout set on activity side is now properly enforced.