thewh1teagle / vibe

Transcribe on your own!
https://thewh1teagle.github.io/vibe/
MIT License
1.09k stars 68 forks source link

Bug: Crashes when trying to recognize speakers #270

Open jcthompson1114 opened 1 month ago

jcthompson1114 commented 1 month ago

What happened?

Vibe crashes when I turn the recognize speakers option on. It works fine otherwise.

Steps to reproduce

turn on recognize speakers, download model, hit transcribe.

What OS are you seeing the problem on?

Window

Relevant log output

options: {
  "path": "C:\\Users\\XXX\\Documents\\Sound Recordings\\XXX + XXXYYY.m4a",
  "lang": "en",
  "verbose": false,
  "n_threads": 4,
  "init_prompt": "",
  "temperature": 0.4,
  "translate": null,
  "max_text_ctx": null,
  "word_timestamps": false,
  "max_sentence_len": 1
}

Caused by:
    Failed to create ONNX Runtime session: Load model from C:\Users\XXX\AppData\Local\github.com.thewh1teagle.vibe\segmentation-3.0.onnx failed:Protobuf parsing failed.

    Location:
        C:\Users\runneradmin/.cargo\registry\src\index.crates.io-6f17d22bba15001f\pyannote-rs-0.2.7\src\session.rs:7:19

Location:
    core\src\transcribe.rs:167:108
App Version: vibe 2.5.1
Commit Hash: bfa7a94b0fa46a50caa6f6a92a27ecb6de06b4bb
Arch: x86_64
Platform: windows
Kernel Version: 10.0.22631
OS: windows
OS Version: 10.0.22631
Cuda Version: n/a
Models: ggml-medium.bin
Default Model: "C:\\Users\\XXX\\AppData\\Local\\github.com.thewh1teagle.vibe\\ggml-medium.bin"
Cargo features: vulkan

{
    "avx": {
        "enabled": true,
        "support": true
    },
    "avx2": {
        "enabled": true,
        "support": true
    },
    "f16c": {
        "enabled": true,
        "support": true
    },
    "fma": {
        "enabled": true,
        "support": true
    }
}
thewh1teagle commented 1 month ago

Hey! thanks for report. It looks related to https://github.com/microsoft/onnxruntime/issues/13117 Please try to delete the files related the diarization in the models folder, it should be the segmentation-3.0.onnx and wespeaker_en_voxceleb_CAM++.onnx. Then disable and enable diarization option again.