onnx-community

/

Qwen2.5-Coder-1.5B-Instruct

like 5

Follow

ONNX Community 1.93k

const model = await AutoModel.from_pretrained('onnx-community/Qwen2.5-Coder-1.5B-Instruct', {
  dtype: 'q4',
  device: 'webgpu',
  session_options: {
    graphOptimizationLevel: 'all',
    executionMode: 'parallel',
    intraOpNumThreads: 4,    
    interOpNumThreads: 4, 
    logSeverityLevel: 1,
    enableMemoryPattern: true 
  },
  progress_callback: (progress) => {
  }
});

Using the above code results in the following error:

Uncaught 3587036944

Not really sure where to start, as that is the only error log I get. Perhaps I am out of memory? Does this model work for anyone else? Tried a lot of diff options for session_options.