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
I believe that it has become impossible to use transcripition with openai-whisper on macOS (and presumably Windows) in the latest version of manim-voiceover.
Let me explain: Recently the transcription module was updated to be compatible with openai-whisper>20230306. This whisper version adds a dependency on openai's triton, which unfortunately has only manylinux wheels, as you can see here. Hence, on non-Linux machines manually building triton (in version >=2.0.0) is the only option to get the transcription part of manim-voiceover to work.
To me personally this means, that I will stick to the Azure service or even use two different python environments (I wanted to use GTTS for early versions of my next video because its free).
Installation logs
Terminal output
poetry error when setting up the environment:
Unable to find installation candidates for triton (2.0.0.post1)
System specifications
System Details
OS: macoS 13.3.1
RAM: 64 GB
Python version (python/py/python3 --version): 3.11
LaTeX details
LaTeX distribution (e.g. TeX Live 2020):
Installed LaTeX packages:
FFMPEG
Output of ffmpeg -version:
PASTE HERE
Additional comments
The text was updated successfully, but these errors were encountered:
Preliminaries
installation instructions.
Description of error
I believe that it has become impossible to use transcripition with
openai-whisper
on macOS (and presumably Windows) in the latest version ofmanim-voiceover
.Let me explain: Recently the transcription module was updated to be compatible with
openai-whisper>20230306
. This whisper version adds a dependency on openai'striton
, which unfortunately has onlymanylinux
wheels, as you can see here. Hence, on non-Linux machines manually building triton (in version>=2.0.0
) is the only option to get the transcription part ofmanim-voiceover
to work.To me personally this means, that I will stick to the Azure service or even use two different python environments (I wanted to use GTTS for early versions of my next video because its free).
Installation logs
Terminal output
poetry error when setting up the environment:
System specifications
System Details
python/py/python3 --version
): 3.11LaTeX details
FFMPEG
Output of
ffmpeg -version
:Additional comments
The text was updated successfully, but these errors were encountered: