-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Our notebooks contain a few tools for debugging, that are present in most notebooks and relate to typcial issues that might occur, for instance, in case of system incompatibility or improper installation.
If you would like to use the GPU, and it fulfils the requirements, it should be displayed alongside the CPU in the first cell of every notebook, where the libraries are imported.
Tensorboard is a convenience tool that displays the progress of the training. Our notebooks are already configured in a manner that all the logs can be read by Tensorboard. However, Tensorboard is known for having issues with ports, so based on the installation approach that you use, there are two ways to properly launch it. The cell, that follows the cell with loading the Tensorboard, displays the ports where the instances of Tensorboard can be found and connected in case it is not displayed in Jupyter.