Hello Shree !
Thank you for posting on Microsoft Learn.
I checked similar issues shared in the community common causes :
- service-side regression (bug introduced in recent updates)
- storage permission change (editor and dataset modules rely on internal storage, if RBAC or storage account access changed, export might fail)
- corrupted or malformed transcripts (even in previously working datasets some validation rules might have changed)
- region-specific issue (specific Azure region Speech Studio backend having issues)
Try using REST API directly to upload data into a training dataset instead of relying on the Studio UI https://learn.microsoft.com/en-us/azure/cognitive-services/speech-service/rest-speech-to-text#create-dataset
or if possible, create a test project in a different region to check if it’s region-specific.
If this persists (and especially since it worked before), this likely requires support intervention.