Replies: 1 comment
-
In this line pipeline_builder.map(map_file, selector="audio") in s2tt file, "audio" is the name of relative address of the text files. You should have a header in the data.tsv file and put the name of audio file address "audio".
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello,
I am trying to do an evaluation with streaming_evaluate. The format of my input data tsv file is
filenamespeakerdurationtarget_language_text
xxxx.wavAAA11.02target text transcription
Also I give the root directory of *.wav files in the source language, as following:
streaming_evaluate --task s2tt --data-file data_file.tsv --audio-root-dir wav-files-directory --output output-directory --ref-field target_language_text --tgt-lang fra
I face this error:
ValueError: The input data does not have an element at path 'audio'.
fairseq2.data.data_pipeline.DataPipelineError: The map operation has failed. See nested exception for details.
Is there any specification that I missed my input data file?
Thanks for your help.
Beta Was this translation helpful? Give feedback.
All reactions