How to get some meaningful debug when training a Neural Network.
Is there anyway a network can be trained one iteration at a time, printing out some debug (e.g. iteration count, current error) compare current error with max error ,...? Currently I can only blind train a network, not knowing anything happening until the training is done, which is terrible because my current neural network is huge and training should take days or weeks.
I guess you need to open up the source code and add debug lines in the iteration steps. It will require you to adapt the source code and rebuilt opencv yourself! Nevertheless it is the only way.
@crazyfffan: That's the main issue with the OpenCV implementation. There's no function to monitor the learning phase. It would be helpful to plot the curves related to bias/variance problems.