streaming inference / 8k -> 16k

[yonzhao]

This is a great job! i want to ask you the following two questions.
1、Does the model support streaming inference ?
2、It's possible to change the model to adapt the transfer form 8k to 16k ?

[ysharma3501]

@yonzhao

1. Unfortunately not really. It’s fast enough to just upsample 600ms audio in like 30ms but it doesn’t support true streaming.
2. Not right now, I am working on a far improved architecture for better quality and it will support audio any sampling rate from 8khz to 24khz. Will be still same speed if not a bit faster infact. Also will support music and general audio.

[shenbuguanni]

@ysharma3501 Could you please explain why streaming support is not available? Is it because it's not technically feasible? For example, I want to add a super-resolution model from 16k to 48k after the 16k noise reduction model. Is it impossible to achieve a streaming version for this?