Hi,
When running my network in inference mode in my Android app, if I use the normalizer I stored with the network, when executing
normalizer.revertLabels(reverted2);
where reverted2 is the saved network
if normalizer is the one stored with the saved network, when I print
Log.i(TAG, " - neuralNetworkloadAndRun - reverted2.toStringFull() - "+reverted2.toStringFull());
The output consists of the lat/lon pairs produced by the network.
If, instead, normalizer is
NormalizerStandardize normalizer = new NormalizerStandardize();
the print statement after normalizer.revertLabels reports the normalized label values, i.e. the ones that would be displayed printing the network before applying
normalizer.revertLabels
This becomes a problem when using samediff because I have not been able to find a way to attaching a normalizer to a network when saving a samediff network.
Is there a way to do that? Am I doing something wrong?