Now that we've done some validation on our models, it's time to talk about monitoring your model during training. You saw some of this before in the section Measuring classification performance in CNTK and the previous Chapter 2, Building Neural Networks with CNTK, through the use of the ProgressWriter class, but there are more ways to monitor your model. For example: you can use TensorBoardProgressWriter. Let's take a closer look at how monitoring in CNTK works and how you can use it to detect problems in your model.
Monitoring your model
Using callbacks during training and validation
CNTK allows you to specify callbacks in several spots in the API. For example: when you call train on a loss function, you can...
