3 ms·
we used existing recipies for training (tuda german), slightly adapted and with additional training data. i am sure with more knowledge we would have made some
by ftreml 7y ago
we used existing recipies for training (tuda german), slightly adapted and with additional training data. i am sure with more knowledge we would have made some extra percent ... performance is not really good, but good enough for our purposes, and with a satisfying cost vs quality trade off
- Nimitz14 7y agoYou should look at the latest recipe's for tedlium, librispeech, swbd etc you'll get significantly better results with those models.