Azure AI Speech
An Azure service that integrates speech processing into apps and services.
1,876 questions
Acoustic data import failed: Zero transcriptions could be parsed from the given input. Error: invalid input line format
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(112.00000000000001, 63%, 20%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EAA%3C/text%3E%3C/svg%3E)
abdou abdou
1
Reputation point
Trying to upload the zip attached at this link --> https://we.tl/t-jXZchy74pE to train my model, I got this error
Acoustic data import failed: Zero transcriptions could be parsed from the given input.
Error: invalid input line format
Can help me to fix it , please?
{count} votes
-
romungi-MSFT 48,431 Reputation points Microsoft Employee2022-05-16T14:33:13.443+00:00 @abdou abdou I believe you are trying to create a custom speech dataset from the Azure speech studio with a transcript and audio files? If Yes, then the format of the transcript(text file) should contain the audio file name along with the text where the audio file corresponding to this text should be separated for each line. In your case the transcript file contains the text but no audio file name and there is only one audio file which seems to have pronunciations for all the text. Could you try adding a single line in your transcript file along with one audio file of the text in this line to test your scenario?
Format for transcript:
filename.wav articulationsThe file
filename.wavshould contain the audio for the textarticulations. The next line in the transcript file should contain another audio file name and the corresponding text. I hope this helps to build your dataset. Thanks!! -
romungi-MSFT 48,431 Reputation points Microsoft Employee
2022-05-16T14:46:59.357+00:00 @abdou abdou I have used your data and tested the upload, The transcript format will work with the format mentioned in my earlier comment but your audio file is using multi channel, this will throw the following error after parsing your dataset.
Number of success: 0 Number of failure: 1 Error message: [ { filename: inputText: articulations... message: Error: audio file has wrong format: Audio files with multiple channels are not supported. status: Failed } ]Please use the audio format in the correct format as mentioned in the documentation i.e mono channel with the correct sample rate and format. Thanks!!
Sign in to comment
Sign in to answer