Wenhao Hu

@coreylammie Thanks for contribution! I think the case output_dir doesn't exist an overwrite is whatever should not be asserted, just make output dir looks good to me.

`DequantizeLinear` is not supported yet. But contribution is well come. Just follow the formula `y = (x - x_zero_point) * x_scale` and add an OpCodeGenerator for it, could you make...

Hi, @juhan could you add support for this op? Contribution is welcome! Just follow the [PRelu](https://github.com/fumihwh/onnx-pytorch/blob/main/onnx_pytorch/op_code_generators/PRelu.py), I think it could help.

Upgrade onnx-pytorch by `pip install onnx-pytorch -U` and see if this problem still exists.

@jarmi-rm Which python version you use? My env is 3.9 and works well....

@alealv Thanks for contribution. Maybe conflict with earlier PR, just wait to merge.

@alealv Could you add test cases for pow and range op?

@GuardSkill Awesome! Thanks for contribution! Please use yapf to reformat code by using style file `.style.yapf`.

FYI，following test failed. ``` =========================== short test summary info ============================ FAILED onnx_pytorch/tests/test_base.py::TestBase::test_pad - assert False FAILED onnx_pytorch/tests/test_base.py::TestBase::test_pad_5D - assert False FAILED onnx_pytorch/tests/test_base.py::TestBase::test_pad_reflect - Assertio... FAILED onnx_pytorch/tests/test_onnx_model_zoo.py::TestModel::test_vision_style_transfer_fast_neural_style_candy FAILED onnx_pytorch/tests/test_onnx_model_zoo.py::TestModel::test_vision_style_transfer_fast_neural_style_mosaic FAILED onnx_pytorch/tests/test_onnx_model_zoo.py::TestModel::test_vision_style_transfer_fast_neural_style_pointilism...

@GuardSkill A hypothesis is that model's weight is not stored in `initializer` field. You can print `model.graph.initializer` to confirm it.