NeurIPS 2019 Expo Demo
Dec. 8, 2019
How Performing Orchestration and Optimization Together Leads to Better Outcomes
Preparing data, engineering features and selecting the model are tasks that directly benefit from expertise. Orchestrating training infrastructure and optimizing these models are tasks that do not. Orchestration also introduces a process complication in which DevOps typically needs to be included in the model development process, thereby slowing the iterative model process and taxing internal resources more than otherwise necessary.
Automating these tasks with tools that support the model development process is a great way to bolster experts and encourage them to be more productive. And automating orchestrating in a way that is directly integrated in and tailored to the optimization process is critical for getting the best results from model training and tuning. Join this demo to learn how orchestration and optimization work together to accelerate model development.