run_model¶

dragon.vm.tensorrt.onnx.run_model(
  model,
  inputs,
  device='CUDA:0',
  **kwargs
)[source]¶

Execute an onnx model once.

Parameters:

• model (onnx.ModelProto) – The onnx model.
• inputs (Union[Sequence, Dict]) – The input arrays.
• device (str, optional) – The executing device.

Returns:

namedtuple – The model outputs.