pyctcdecode usage in Speechbrain .scorer/.binary extension

[VitoPepe]

Hello, I use coqui-stt which uses Kenlm to generate a language model. I would like to use pyctcdecode to add this language model to the CTC of my Speechbrain engine.

But the model produced by Coqui-stt is in .scorer or .binary format. I tried to give these two models as argument in the kenlm_model_path argument of the build_ctcdecoder function but it does not work.

Do you have any idea which method I should use to make it work?

Thanks