🐛 [Bug] Encountered bug when compiling model with `acc_ops` conv converter

[gs-olive]

Bug Description

Encountered the following error when compiling model with acc_ops.conv2d converter:

  File "~/TensorRT/py/torch_tensorrt/fx/fx2trt.py", line 328, in call_function
    return converter(self.network, target, args, kwargs, self._cur_node_name)
  File "~/TensorRT/py/torch_tensorrt/fx/converters/aten_ops_converters.py", line 136, in aten_ops_convolution
    return acc_ops_converters.acc_ops_convnd(
  File "~/TensorRT/py/torch_tensorrt/fx/converters/acc_ops_converters.py", line 226, in acc_ops_convnd
    raise RuntimeError(
RuntimeError: linear convolution_79 has weight of type <class 'tensorrt.tensorrt.ITensor'>, Expect Optional[Tenosr]

To Reproduce

Steps to reproduce the behavior:

1. Compile Resnet18 with torch_tensorrt.dynamo.torch_compile

Expected behavior

Model should not throw errors due to acc_ops.conv2d layer.

Environment

• Torch-TensorRT Version (e.g. 1.0.0): 25db257
• PyTorch Version (e.g. 1.0): 2.1.0.dev20230419+cu117