You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Describe the bug
I wanted to put a minimum_timestamp_duration to visualize the result given by nemo forced aligner with Gecko as the documentation recommends.
When I change the value of minimum_timestamp_duration to 0.4, it's all good.
If I try to change the minimum_timestamp_duration for a higher value like 0.5, I get a ValueError : Steps/Code to reproduce bug
Notice the option ctm_file_config.minimum_timestamp_duration=0.5 to change the value of minimum_timestamp_duration. To be sure to have the error, you can put an extremely high value like 50.
python <path to NeMo>/tools/nemo_forced_aligner/align.py pretrained_name="stt_en_fastconformer_hybrid_large_pc" manifest_filepath=nemo/manifest.json output_dir=nemo/output ctm_file_config.minimum_timestamp_duration=0.5
The value that is too high and causes the error depends on the respective duration of your tokens/words/segments that are calculated in this loop. The loop iterates over each token/word/segment of the transcription and finally produces the ctm file).
Expected behavior
No error should be raised.
Environment overview
Method of NeMo install: source
Additional context This is a minor error that I've already fixed locally, so I'll make a pull request in a few minutes
The text was updated successfully, but these errors were encountered:
Describe the bug
I wanted to put a
minimum_timestamp_duration
to visualize the result given by nemo forced aligner with Gecko as the documentation recommends.When I change the value of
minimum_timestamp_duration
to 0.4, it's all good.If I try to change the
minimum_timestamp_duration
for a higher value like 0.5, I get aValueError
:Steps/Code to reproduce bug
minimum_timestamp_duration
. To be sure to have the error, you can put an extremely high value like 50.The value that is too high and causes the error depends on the respective duration of your tokens/words/segments that are calculated in this loop. The loop iterates over each token/word/segment of the transcription and finally produces the
ctm
file).Expected behavior
Environment overview
Additional context
This is a minor error that I've already fixed locally, so I'll make a pull request in a few minutes
The text was updated successfully, but these errors were encountered: