ba-dls-deepspeech icon indicating copy to clipboard operation
ba-dls-deepspeech copied to clipboard

Published 20 hours ago • verified publisher icon baidu-research

Diverge with augmented data

Open a00achild1 opened this issue 8 years ago • 0 comments

I'm trying to train a model with augmented LibriSpeech train-other-500 dataset. The speech was augmented by adding noise and changing the tempo. But the loss couldn't converge. I have tried letting the learning rate small(1e-5), adding more GRU units, or adding more nodes, but none of these works. Is anyone has a solution or suggestions? plot_7rnn1200_libri500_aug

a00achild1 avatar Feb 11 '17 13:02 a00achild1