Loss Inspector

You can use the Loss Inspector to monitor how validation loss changes when training High Detail mode tools. By monitoring the validation loss, you can check the following characteristics:

The fluctuation and the saturation of the validation loss to get the best loss for the best model.
How good is the model at generalization
How much the model is overfitted as this loss is from the validation set.

For more information about validation loss, see Validation Loss.

Note: The Loss Inspector does not provide the graph of validation loss of a High Detail tool trained in the earlier version of VisionPro Deep Learning 1.1. For example, if you imported a Green Classify High Detail Mode tool trained in VisionPro Deep Learning 1.0, the Loss Inspector does not provide its loss graph. But, if you are-train this tool in 1.1 or 2.0, the loss inspector shows its loss from this new training session.


Status Variable	Description
Elapsed Time	The elapsed time for the current training.
Elapsed Epoch	The elapsed epoch for the current training.
Loss	The validation loss of the trained model at a certain time. Loss is calculated in every 0.125 epoch.
Best Epoch	The epoch that recorded the lowest loss among all the elapsed epochs.
Best Loss	The lowest loss among all the elapsed epochs.

When the Loss remains the same over several Elapsed Epochs, the loss inspector uses an internal metric to select the Best Epoch from the epochs with the same Loss value. The metric selects an epoch to provide the best results. In the example below, the internal metric selects 5 as the Best Epoch among 3, 4, and 5.

Elapsed Epoch	Loss	Best Epoch
3	0.2	3
4	0.2	3
5	0.2	5
6	0.2	5
7	0.2	7

For more information about using the Loss Inspector, see Using the Loss Inspector.