Speech recognition model stuck at download

[mato25579]

Hi there folks, new user here. I’m on Windows 10 and Kdenlive Version 24.12.2.

When I create a new project and either put in the video or not and I try to setup speech recognition. I’m getting into downloading a model which recommends me to the turbo. Either if I select manual or recommended selection it doesn’t work, upon downloading it gets stuck at 25% and won’t move any further even upon long waiting session. Should i try using another speech model or is there any fix around it? Speech to text as “numba, numpy, openai-whisper, srt and torch” seemed to be working and were downloaded correctly. Thanks.

[Tonned]

Hi and welcome to the community.

Not sure about Windows but a similar issue has been adressed for Linux and MacOS I believe. Take a look here:

Where does Kdenlive 24.12.2 store speech models in Flatpak?

That is good for you. I use Mint and have used it for over ten years. Nevertheless, thank you for at least making the effort to try and help me. By the way, Whisper is correctly installed from within the Flatpak Kdenlive. So only the speech/language models won’t download/install. It is my belief that the latest Flatpak should have the entire proces ‘automated’ when required. At least that is how I understand. It uses a Python venv to accomplish things.

You may have to do a bit of digging in Windows but the gist might be clear to you.

[antimidia]

I’m using Manjaro and I also have this issue, but on 24.12.3.

[Eugen_Mohr]

The model installation could be a bit “clunky”. I would abort the download and try with another model. Then try if Whisper Speech to Text is working.

[Tonned]

I see this is a while ago so you may have found the solution.
If not, did you check if the folder called “whisper” (without the quotes and lower case only)
is where it should be? If not, simply create it and you’ll be fine.