Thanks for the idea :-) One of our product uses this repository, many changes are made to integrate this repository to our environment, I finally got some time to back port some of them (non-application specific part) into this repository, I hope this helps.
Notable changes are:
Change the description in the README.md "Note that we split the speech signals into chunks of 200ms with 10ms overlap" to "... with 190ms overlap", since this (190ms) is what implemented in the code.
Fix the offset-by-one error when making batches
Add support for DataParallel
Reformat the code according to PEP8 (except 2 spaces is used as indentation)
Thanks for the idea :-) One of our product uses this repository, many changes are made to integrate this repository to our environment, I finally got some time to back port some of them (non-application specific part) into this repository, I hope this helps.
Notable changes are: