LibreSpeech is 1000h, WSJ is 73h. Google is training on datasets that are as big as 30kh and MS seems to work on a 10k h dataset. At the moment, I am working

LibreSpeech is 1000h, WSJ is 73h. Google is training on datasets that are as big as 30kh and MS seems to work on a 10k h dataset. At the moment, I am working on a similar e2e system but 80h big dataset makes it a really challenging task to generalize well.