Azure Speech-to-text. Missing a base model with support for custom model training with audio

Alex 6 Reputation points
2022-05-03T12:19:05.837+00:00

Speech-to-text
In the docs en-US is listed as a language that supports custom model training with audio data.
198484-s1.png

But in the Speech Studio there is no base models with audio data training support. Where did they go?!
198492-s2.png

Azure AI Speech
Azure AI Speech
An Azure service that integrates speech processing into apps and services.
1,413 questions
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. YutongTie-MSFT 46,991 Reputation points
    2022-05-04T00:08:59.287+00:00

    @Alex Sorry for the misunderstanding and unclear document, this table you shared is not about the baseline model, it's about the lanauage supports for each feature. I know the description is confused, I will contact the content author to modify it.

    The table you shared is related to input data and data type as below screenshots, this means, for languages in this table, the audio data input is supported:
    198576-image.png

    In the studio, it reflect below:

    198594-image.png

    Sorry again for the misunderstanding.

    I hope this helps.

    Regards,
    Yutong

    -Please kindly accept the answer if you feel helpful, thanks a lot.