When trying to download Tubo in speech-to-text settings, it says checking available models and won’t change
Sometimes Whisper gets kind of stuck with this. The trick is to run it from the command line once, and then things tend to work.
Lat’s assume you are on Linux and you have Whisper installed and ready to go (if not, use your package manager to install it now).
Get a short clips, maybe just a few seconds with some speaking and run whisperwith the model you want to install:
whisper -model turbo yourvideoclip.mp4
Whisper will then download the model, in this case turbo and do its thing.
You can discard whatever Whisper produces, if you want. The next time you go to use Kdenlive, you should find Turbo has been installed and, in my experience, you will be able to install more models directly from within Kdenlive.
If you are using the FlatPak version of Kdenlive, check the article here.
1 Like