lstm-transducer model for Chinese data

[lucasjinreal]

Does there a lstm-transducer eg for Any Chinese dataset?

[csukuangfj]

Yes, please see

https://k2-fsa.github.io/sherpa/python/streaming_asr/lstm/index.html#chinese

[csukuangfj]

And here is the demo using the above Chinese model.

https://k2-fsa.github.io/sherpa/python/streaming_asr/endpointing.html#endpointing-demo-chinese

[csukuangfj]

Does there a lstm-transducer eg for Any Chinese dataset?

We will upload the code soon.

[lucasjinreal]

OK. hoping for it. BTW, does wenet speech dataset is currently most accurate model for daily life voice recognition?

[csukuangfj]

I think Wenetspeech is the largest open-source Chinese dataset so far, to the best of my knowledge.