Saik0s / Whisperboard

The open-source iOS app that's making quality voice transcription more accessible on mobile devices.
GNU General Public License v3.0
675 stars 70 forks source link

Crash on Transcribe/Start Over (was: Crash on Open) #27

Open leotulipan opened 8 months ago

leotulipan commented 8 months ago

Igor @Saik0s , what is the best way to help you debug and/or collect debugging info?

iOS 17.2.1 iPhone 14 Pro (not max) Whisperboard 1.11.6

As I have a couple of finished voice recording transscripts in the app which I have not yet saved elsewhere I do not want to de-install as I will loose all the data

What would be helpful here would be a "Files" Folder in the iOs Files App where the recordings and transcripts reside. That way I could access them without the app and also remove buggy files if needed....

Saik0s commented 8 months ago

Hi @leotulipan , thanks for reporting it, I will take a look into it ASAP and let you know!

Saik0s commented 8 months ago

New version(1.11.7) with a fix should be already available on App Store. Let me know if there are any issues.

It had two main issues: a memory problem and the automatic resumption of transcription on launch. I improved memory handling to prevent crashes. I also turned off the auto-resume feature. Now, you have to press "Resume" to continue transcription.

leotulipan commented 8 months ago

New version just downloaded.

  1. The crash on start is gone. I see the "Resume" - "Start over" buttons.
  2. I had to download the large model again (see new separate issue) and after a couple of trials it worked
  3. I still have one audio file I imported from Rode Reporter on August 4th that shows "Resume" and "Start Over" but the app always crashed even after changing model or selecting the Language in the settings instead of using "Auto".
Saik0s commented 8 months ago

How long is the audio? Also, try new 1.11.8 version please.

The resume function always uses the model that was initially selected, while the start over function uses the currently selected model.

leotulipan commented 8 months ago

WOW. Thanks for the quick release cycles you are running and thanks for working on this!

The audio is just 02:23 and in WAV. I did send you a dm via email with a link. It is a voice recording from me in German.

I tried tiny, large-v3 and the quantized version with language set to auto or German with and without the new GPU setting

on iPhone 14 Pro with latest iOs 17.2.1 and App version 1.11.8 it always crashes

as I was able to save to files and then re-import I tried that with a new name "Buggy.wav" and this still crahes. that is the file I sent you.

If you need more debugging from me let me know