Hello Luca,
thanks for your message!
The open ai whisper module is using torch internally, so you will have to install CUDA enabled torch:
https://github.com/openai/whisper/discussions/47
Have a nice weekend,
Peter

I really like your programm, I can confirm that using .mp3-files works just fine.
However you mentioned in the video, that the use of the GPU instead the CPU could possible accelerate the program. Is this configuration easy to change (with relatively little programming knowledge) and if so how would this work?
Thanks again for your work!

Greetings,
Luca

Hello [Your Name] 😉
Thanks for your message.
To transcribe pre recorded audio files, you can simply skip the first step of recording and manually input the path to your pre recorded audio file into the „Audio Datei:“ field.
It will definitely work for .wav files – I’m not sure with other file types though. You can simply test it out, but might have to convert them first.
Best regards,
Peter

Dear Widerporst Team,

I hope this message finds you well. Firstly, I would like to commend you on your exceptional work in creating such a brilliant script. The Widerporst program has been remarkably helpful and straightforward to use, and I genuinely appreciate the effort you’ve put into making it user-friendly.

I have been successfully using Widerporst to record audio and convert it into text, following the easy steps laid out in your instructional video. However, I have some pre-recorded audio files on my device that I would like to convert into text using your program.

I am writing to inquire whether there is a possibility to import these pre-recorded files into Widerporst for conversion. If this feature is available, could you kindly guide me through the process? If not, would you consider this as a possible feature to add in a future update?

Thank you once again for your excellent work on Widerporst. I look forward to your response.

Best Regards,

[Your Name]