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
When I follow the installation instructions in the readme (pip install timething), I'm unable to run the align-long example. My investigation has led me to believe it's because it's installing an older version of the code that doesn't yet support that command.
How to reproduce
Steps to reproduce the behavior:
In a new project/python virtual environment, pip install timething
Copy keanu.mp3 and keanu.cleaned.txt to the appropriate subdirectories
Usage: timething [OPTIONS] COMMAND [ARGS]...
Try 'timething --help' for help.
Error: No such command 'align-long'.
If I try timething -- help, it indeed only offers the command options align, download, and recut.
Expected
I expected (or hoped, anyway) that it would try to align the audio with the transcript and give me an output file with that info.
Additional context
I installed timething in a python virtual environment to get it working, since the default python 3 on my machine was 3.10. By digging around the virtual environment files, it looks like the local version of init.py has version number "0.0.27", in case that's helpful. I'd be happy to try to run timething from source in the meantime just to see how well it'll work for my usecase, but I'm not sure how to do that. Do I need to run any sort of setup command first?
The text was updated successfully, but these errors were encountered:
What
When I follow the installation instructions in the readme (
pip install timething
), I'm unable to run thealign-long
example. My investigation has led me to believe it's because it's installing an older version of the code that doesn't yet support that command.How to reproduce
Steps to reproduce the behavior:
pip install timething
timething align-long --audio-file fixtures/audio/keanu.mp3 --transcript-file fixtures/keanu.cleaned.txt --alignments-dir aligned --batch-size 10 --n-workers 5
If I try
timething -- help
, it indeed only offers the command optionsalign
,download
, andrecut
.Expected
I expected (or hoped, anyway) that it would try to align the audio with the transcript and give me an output file with that info.
Additional context
I installed timething in a python virtual environment to get it working, since the default python 3 on my machine was 3.10. By digging around the virtual environment files, it looks like the local version of
init.py
has version number "0.0.27", in case that's helpful. I'd be happy to try to run timething from source in the meantime just to see how well it'll work for my usecase, but I'm not sure how to do that. Do I need to run any sort of setup command first?The text was updated successfully, but these errors were encountered: