search

feldberlin / timething

Timething is a library for aligning text transcripts with their audio recordings.
MIT License
92 stars 7 forks source link

cannot load audiofile #23

Open osama84-hup opened 1 year ago

osama84-hup commented 1 year ago

when i run the code " timething align --language arabic --metadata text.csv --alignments-dir aligned --batch-size 4 --n-workers 4 --use-gpu False" in the terminal i got the following:

setting up aligner... starting aligment... 0%| | 0/1 [00:00<?, ?it/s]formats: can't open input file `audio/chapter01.wav

and it ends without any results!!!

so if there is any error in using the function or any thing else please help thanks

gameoverhack commented 1 year ago

I had the same problem, eventually I tracked it down to ffmpeg not being installed for python - unintuitively it turns out ffmpeg and ffmpeg-python are not the same thing!

This worked for me:

conda install -c conda-forge ffmpeg-python

then open a python command line and run:

import torchaudio
torchaudio._extension._init_ffmpeg()
print(torchaudio._extension._FFMPEG_INITIALIZED)

Should return True if successful.