0% Precision when using French Model

[ghost]

Hello, i am trying to use the French Model (vosk-model-fr-0.22) for Voice Recognition in NodeJS using VoskJS Source: https://github.com/solyarisoftware/voskJs/

Recognition Works flawlesslyh using an English Model, as soon as i used the french model on french sentences, its going hog wild

Here is the voskJs output: "hum mon dieu en un non hum hum bon euh bon bon le punch" Expected output: "Le chien regarde, le bateau glisse, la fille dors, la dame brosse les cheveux de la petite fille"

Obviously, there is something wrong, here are two examples files i tried: https://www.lightbulblanguages.co.uk/resources/audio/tempsfrance.mp3 https://www.lightbulblanguages.co.uk/resources/audio/scene-de-plage.mp3

let me know what could be wrong