DeepAlignmentNetwork icon indicating copy to clipboard operation
DeepAlignmentNetwork copied to clipboard
verified publisher icon MarekKowalski

why did you split your network into two stages to train? what is the different between all to train and split train?

Open liulei2140 opened this issue 7 years ago • 1 comments

liulei2140 avatar Jul 19 '18 07:07 liulei2140

The stages were separated to allow the first stage to learn fully before the second stage is added. Thanks to this approach the second stage doesn't start lerning before the first stage is able to make sensible predictions.

Marek

MarekKowalski avatar Jul 25 '18 00:07 MarekKowalski