Live(?) transcription from the terminal

[JohnZolton]

This is a psuedo-live transcription program using Whisper.

https://github.com/JohnZolton/scribe

A microphone recorder function and a whisper function run in parallel using multiprocessing. The microphone continuously records 5 second audio files while the whisper function checks if a new audio file is ready for transcription and then displays the transcription on the terminal

[Agusteando]

Hey John, I am trying to run your implementation but I am getting

python livemic.py
Traceback (most recent call last):
File "C:\Users\desar\scribe\scribe\livemic.py", line 12, in
import whisper
File "C:\ProgramData\Anaconda3\lib\site-packages\whisper.py", line 69, in
libc = ctypes.CDLL(libc_name, mode=ctypes.RTLD_GLOBAL)
File "C:\ProgramData\Anaconda3\lib\ctypes_init_.py", line 364, in init
if '/' in name or '\' in name:
TypeError: argument of type 'NoneType' is not iterable

[JohnZolton]

I'm sorry I have no idea how to fix that

[jongwook]

@Agusteando I think you're using a different package named whisper. Please try reinstalling the package by:

pip uninstall whisper
pip install git+https://github.com/openai/whisper.git 

[guyos46]

I had the same problem. @jongwook solution worked for me.

[samueldonovan1701]

I got this error when I had the wrong version of Python. Switching from 3.11 to 3.10 fixed it.