test_conv_cudnn_nhwc_support_cuda_float32: AssertionError: Tensor.is_contiguous() unhandled kwargs

[desertfire]

Repro:

 PYTORCH_TEST_WITH_INDUCTOR=1 python test/test_nn.py -k test_conv_cudnn_nhwc_support_cuda_float32

Error:

[2022-10-27 23:30:54,172] torch._dynamo.convert_frame: [ERROR] WON'T CONVERT test_conv_cudnn_nhwc_support /fsx/users/binbao/pytorch/test/test_nn.py line 16716 
due to: 
Traceback (most recent call last):
  File "/fsx/users/binbao/pytorch/torch/_dynamo/variables/tensor.py", line 503, in call_method
    assert not kwargs, f"Tensor.{name}() unhandled kwargs"
AssertionError: Tensor.is_contiguous() unhandled kwargs

from user code:
   File "/fsx/users/binbao/pytorch/test/test_nn.py", line 16724, in test_conv_cudnn_nhwc_support
    self.assertTrue(o.is_contiguous(memory_format=torch.channels_last))

[yanboliang]

Updated command:

PYTORCH_TEST_WITH_INDUCTOR=1 python test/nn/test_convolution.py -k test_conv_cudnn_nhwc_support_cuda_float32