Systém pro spouštění a monitorování výpočtu neuronových sítí

Abstract

This diploma thesis is focused on the development of a system that would allow monitoring of neural network learning and starting their calculations in the environment of a remote computational server and supercomputer. As part of this work is description of neural network learning monitoring, task management in remote environments and methods of metric persistence. This thesis also describes implementation of three main parts, which include management server, client library and user interface.