DISTRIBUTED TRAINING AND TESTING OF AI SYSTEMS AND APPLICATIONS

Abstract

In various examples, supervised learning and reinforcement learning can each be performed at scale within appropriate clusters of compute nodes. The compute nodes in each cluster can use specialized hardware tailored to a specific task, and the tasks performed by the different clusters can be run together in a loop or circular workflow, using the output of one task as the input of the next. For example, parallel supervised learning jobs can continuously apply the latest skills or experience acquired during reinforcement learning to generate an updated model, and parallel reinforcement learning jobs can continuously use the latest version of the model to learn new skills or experience.The techniques presented here can be used to train and / or test humanoid robots, physical AI, or other AI systems and applications.