Issue in offline use of speaker-diarization 3.1 model

[ashu5644]

@hbredin , I am trying to use speaker-diarization 3.1 model in offline mode , but I am getting following error:

sess = C.InferenceSession(session_options, self._model_path, True, self._read_config_from_model)
onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from /home/ubuntu/workspace/pyannote_models/pyannote-wespeaker-voxceleb-resnet34-LM-embedding.bin failed:Protobuf parsing failed.

Code:

pipeline = Pipeline.from_pretrained('config.yaml')
pipeline.to('cuda')
audio_, _ = librosa.load('audio.wav', sr=16000)
audio = {"waveform": torch.tensor(np.reshape(audio_,(1,-1))), "sample_rate": 16000}
output = pipeline(audio)

config.yaml:

version: 3.1.0

pipeline:
  name: pyannote.audio.pipelines.SpeakerDiarization
  params:
    clustering: AgglomerativeClustering
    embedding: '/home/ubuntu/workspace/pyannote-wespeaker-voxceleb-resnet34-LM-embedding.bin'
    embedding_batch_size: 32
    embedding_exclude_overlap: true
    segmentation: '/home/ubuntu/workspace/pyannote_models/pyannote-segmentation-3.0-pytorch-model.bin'
    segmentation_batch_size: 32

params:
  clustering:
    method: centroid
    min_cluster_size: 12
    threshold: 0.7045654963945799
  segmentation:
    min_duration_off: 0.0

Similar code is working for speaker-diarization 3.0 in offline mode by replacing embedding argument in corresponding config with .onnx model path. But for speaker-diarization 3.1 no .onnx model exists here. Can you please suggest some workaround for this issue ?

Thank you.

[github-actions[bot]]

Thank you for your issue. We found the following entry in the FAQ which you may find helpful:

• Does pyannote support streaming speaker diarization?

Feel free to close this issue if you found an answer in the FAQ.

If your issue is a feature request, please read this first and update your request accordingly, if needed.

If your issue is a bug report, please provide a minimum reproducible example as a link to a self-contained Google Colab notebook containing everthing needed to reproduce the bug:

• installation
• data preparation
• model download
• etc.

Providing an MRE will increase your chance of getting an answer from the community (either maintainers or other power users).

Companies relying on pyannote.audio in production may contact me via email regarding:

• paid scientific consulting around speaker diarization and speech processing in general;
• custom models and tailored features (via the local tech transfer office).

This is an automated reply, generated by FAQtory

[ashu5644]

Solved by upgrading to pyannote.audio 3.1

[seanzhang-zhichen]

Solved by upgrading to pyannote.audio 3.1

unuseful