upload fail

[MRTHtlk1]

Hi! I've been having trouble with upload, and it's already 1 month or even more I can't get an answer via e-mail about this issue. Moreover I have toped up my account, and I can't get any answer or my money back. I have checked the previous opened issues here in Github, and I see that now there is the same problem as was 1 year ago. After upload was almost done, the message pops up saying 'Upload fail'.
I've tried different types of files (mp4, m4a, mp3) in different browsers (Safari, Chrome, even Mozilla) + I tried to upload the file, which was previously successfully transcribed, so the problem isn't in corrupted file. I would much appreciate if you could look into this issue and fix it. Thanks

[gbonnema]

You could have a look at faster-whisper on repository SYSTRAN in github, which provides a python implementation of the whisper model and appears to be faster. You run this locally on your own machine. I personally use faster-whisper-xxl which includes all sublibraries (especially handy for windows users) but I forgot which github library that is.

So, all I do is open up a (linux) terminal (I work from pop-os) and call the faster-whisper-xxl with parameters and the filename. Because I like full sentences, I specify "--sentence". It runs on my computer on only CPU due to the videocard I use.

Link is https://github.com/SYSTRAN/faster-whisper.

I think openai on github offers the same capability, but according to faster-whisper it is slower.

I hope this helps.