We have an internal dependency on our STT and for now there are no plans for Chinese unless we sign some sort of contract / deal (we do not speak Chinese alas, we would need at least one person with language / culture)

You are encouraged nevertheless to try the vad as is - there is a high probability it will just work out of the box (please report if it does)

If you will be trying it, please make sure you use a visualization tool to tune the params

We will be making the post processing simpler soon though

We have an internal dependency on our STT and for now there are no plans for Chinese unless we sign some sort of contract / deal (we do not speak Chinese alas, we would need at least one person with language / culture)

You are encouraged nevertheless to try the vad as is - there is a high probability it will just work out of the box (please report if it does)

If you will be trying it, please make sure you use a visualization tool to tune the params

We will be making the post processing simpler soon though

So this repo may work on Chinese now?if yes,i will try

So this repo may work on Chinese now?if yes,i will try

The vad may

Language classifier / number detector obviously won't

So this repo may work on Chinese now?if yes,i will try

The vad may

Language classifier / number detector obviously won't

Got the result below:
[{'end': 290000, 'start': 36000}]
what's the meaning of 290000 and 36000?
Another question:
I do the vad using the function below:
speech_timestamps = get_speech_ts(wav, model, visualize_probs=True,num_steps=4)
but no chart appears when the scripts finished.Any suggestions?thanks!

what's the meaning of 290000 and 36000?

#25

speech_timestamps = get_speech_ts(wav, model, visualize_probs=True,num_steps=4)
but no chart appears when the scripts finished.Any suggestions?thanks!

this has to be run in a notebook