Difference in portal and the SDK results (Pronunciation Assessment)

Jimmy 21

Hi:
I am running the model through the Python SDK for pronunciation assessment and the JSON that comes back is different than the one that is returned in the online portal (Speech Studio) with the same wave file and the reference text.
How can I get the same results as returned by the Speech studio using the Python SDK?

Regards,
Jimmy

YutongTie-MSFT 46,991 Reputation points

2022-06-10T09:19:49.083+00:00

Hello @Jimmy

Thanks for reaching out to us, could you share the sample you are trying to and the results? We are looking into this.

Regards,
Yutong
YutongTie-MSFT 46,991 Reputation points

2022-06-15T17:44:13.257+00:00

Hello @Jimmy

I think you have deleted your post, have you solved the issue? Thanks.
Sagar Kulkarni 0 Reputation points

2023-04-21T10:36:18.34+00:00

@YutongTie-MSFT I'm also facing this similar issue