Traceback (most recent call last):
  File "../trt/perf.py", line 23, in <module>
    import trtutils
  File "/content/DeepLearningExamples/PyTorch/SpeechRecognition/Jasper/trt/trtutils.py", line 18, in <module>
    import onnxruntime as ort
ModuleNotFoundError: No module named 'onnxruntime'

Getting component
Traceback (most recent call last):
  File "../trt/perf.py", line 132, in <module>
    main(args)
  File "../trt/perf.py", line 38, in main
    pyt_components, saved_onnx = perfutils.get_pytorch_components_and_onnx(args)
  File "/content/DeepLearningExamples/PyTorch/SpeechRecognition/Jasper/trt/perfutils.py", line 225, in get_pytorch_components_and_onnx
    onnx_path = get_onnx(args.onnx_path, model, args)
  File "/content/DeepLearningExamples/PyTorch/SpeechRecognition/Jasper/trt/perfutils.py", line 118, in get_onnx
    phony_signal = torch.zeros(signal_shape, dtype=torch.float, device=torch.device("cuda"))
TypeError: zeros(): argument 'size' must be tuple of ints, but found element of type Tensor at pos 3
bash: line 6:  1294 Segmentation fault      (core dumped) python ../trt/perf.py --ckpt_path ./jasper_fp16.pt --wav=example1.wav --model_toml=../configs/jasper10x5dr_nomask.toml --make_onnx --onnx_path jasper.onnx --engine_path jasper.plan