Usage - Example

#2
by hooman650 - opened

Amazing work @Xenova - I have been using onnx-community/OuteTTS-0.2-500M and the quality was amazing (leveraging outetts npm package) - do you mind adding a small guide as how to use this model? I am super excited for this as the quality of the samples were awesome!

ONNX Community org

Usage example coming soon! :)

ONNX Community org

Example python code added to the README! :)

Example JS code coming soon.

Neat thanks alot looking forward to the JS code example

ONNX Community org

Awesome work @Xenova ! that was fast - as a suggestion it would be nice to have similar interface to outtets

export interface GenerateOutput {
    audioBuffer: AudioBuffer;
    save(filename: string): void;
    to_wav(filename: string): ArrayBuffer;
  }

It seems like kokoro-js exposes only save - cheers and thanks again

hooman650 changed discussion status to closed
ONNX Community org

The output is now a RawAudio object, which supports toWav, toBlob, and save functions. Hope that helps!

Nice, Yes, I just saw your example demo here thanks - this is amazing @Xenova . BTW, just a heads up, maybe its on my side only but if you pass there the device: 'webgpu' you get an exception not sure why. The demo also is not passing the device - is it supposed to work with GPU?

ONNX Community org

Indeed, we're working on fixing this bug in onnxruntime-web: https://github.com/microsoft/onnxruntime/issues/23403.

2025-01-17 02:57:22.385599 [E:onnxruntime:, sequential_executor.cc:505 ExecuteKernel] Non-zero status code returned while running Add node. Name:'/decoder/F0.1/Add' Status Message: Failed to run JSEP kernel

Sign up or log in to comment