[WinError 2] The system cannot find the file specified

[Atoli]

Greetings.

I have an issue.
I installed whisper, however, when i try to use with via command line with an audio file, it says it cannot find said file (even if it's there).

Installed it as the main page says: pip install git+https://github.com/openai/whisper.git .
Attached image is the errors that pop up when run whisper audiofilename.wav.

What could the issue be?

OS: Windows 10.
Python ver: 3.10.

I have read this could happen if python is not installed but i already checked pip, it says ffmpeg-python version 0.2.0 so that could not be the problem unless this ffmpeg is not recognized by whisper.

[phineas-pta]

re-check with ffmpeg -version

[Atoli]

re-check with ffmpeg -version

Greetings, thank you for answer.

Weird, i tried ffmpeg -version in command line but it does not recognize ffmpeg as a command.

I get 'ffmpeg' is not recognized as an internal or external command, operable program or batch file.

If i try pip show ffmpeg-python i DO get:

Name: ffmpeg-python
Version: 0.2.0
Summary: Python bindings for FFmpeg - with complex filtering support
Home-page: https://github.com/kkroening/ffmpeg-python
Author: Karl Kroening
Author-email: [email protected]
License: UNKNOWN
Location: c:\python310\lib\site-packages
Requires: future

It says i have ffmpeg only if check with pip, checking by itself does not work.

[phineas-pta]

ffmpeg is not ffmpeg-python

whisper needs the former not the latter

[SunJR24]

hello I have the same question.
I tried unstall ffmpeg-python and install ffmpeg ，but it still print error

code

import whisper

model = whisper.load_model("medium")
result = model.transcribe("G:\Work-base\processed_3.wav")
print(result["text"])

error

Traceback (most recent call last):
File "G:\Anaconda3\envs\T3\lib\site-packages\whisper\audio.py", line 46, in load_audio
ffmpeg.input(file, threads=0)
AttributeError: module 'ffmpeg' has no attribute 'input'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "G:\Work-base\Colab-WorkBase\VITS-fast-fine-tuning\scripts\t5.py", line 4, in
result = model.transcribe("G:\Work-base\Colab-WorkBase\VITS-fast-fine-tuning\custom_character_voice\VO_nahida\processed_3.wav")
File "G:\Anaconda3\envs\T3\lib\site-packages\whisper\transcribe.py", line 121, in transcribe
mel = log_mel_spectrogram(audio, padding=N_SAMPLES)
File "G:\Anaconda3\envs\T3\lib\site-packages\whisper\audio.py", line 130, in log_mel_spectrogram
audio = load_audio(audio)
File "G:\Anaconda3\envs\T3\lib\site-packages\whisper\audio.py", line 50, in load_audio
except ffmpeg.Error as e:
AttributeError: module 'ffmpeg' has no attribute 'Error'

进程已结束,退出代码1

[Atoli]

ffmpeg is not ffmpeg-python

whisper needs the former not the latter

I see.

Excuse me for the noob question, but how do i install ffmpeg in my system? I know how to use ffmpeg but when i downloaded it was just an exe file in a folder, it's not "installed" in the system, but something i just run.

Repositories like whisper and other utilities can not use it for this reason, i guess.

[phineas-pta]

u can add the ffmpeg folder to environment variable PATH

something like this https://linuxhint.com/add-directory-to-path-environment-variables-windows/

[jpthiru]

The instructions are here : https://www.geeksforgeeks.org/how-to-install-ffmpeg-on-windows/

May be useful for someone.

[SunJR24]

I solved it

you need to use anaconda

conda install -c conda-forge ffmpeg