using speaker diarization pipeline offline, however program tries to find embedding online

[justinvictor]

I wish to use speaker diarization pipeline offline. however although i managed to point to the offline segmentation model successfully, the program keeps trying to get speechbrain/spkrec-ecapa-voxceleb embedding online.
i downloaded the folder and tried to point the embedding value to the offline folder, however this does not work too.
Any advice is much appreciated, thanks!

pipeline:
name: pyannote.audio.pipelines.SpeakerDiarization
params:
clustering: AgglomerativeClustering
embedding: speechbrain/spkrec-ecapa-voxceleb
embedding_batch_size: 32
embedding_exclude_overlap: true
segmentation: "C:/VTT/segmentation/pytorch_model.bin"
segmentation_batch_size: 32

[justinvictor]

Found the answer. clone speechbrain/spkrec-ecapa-voxceleb and put it in a folder called speechbrain (the term speechbrain has to appear somewhere in the folder path name). Then the speaker diarization config.yaml file embedding value will now accept a filepath such as embedding: "C:/VTT/speechbrain/spkrec-ecapa-voxceleb"