Pytorch 1.13 onnx export is incompatible with TensorRT conversion

Pytorch 1.13 APIs for onnx export seem to generate onnx graphs that are incompatible with TensorRT conversion tools we use internally. The bias vectors seem to be added as an input to conv nodes instead of being added as an initializer. For example: Torchvision resnet18 when exported to onnx using torch.onnx.export results in the following graph:

Pytorch 1.13

The bias nodes are initialized as part of Conv operator in Pytorch.1.11

Using Pytorch1.13 onnx graph for TensorRT conversion results in the following error:
Errors out with [TRT] ModelImporter.cpp:779: ERROR: builtin_op_importers.cpp:647 In function importConv: [8] Assertion failed: && "The bias tensor is required to be an initializer for the Conv operator."

Code to generate onnx graph:

   import torch
   import torch.onnx as onnx
   import torchvision.models as models
   resnet18 = models.resnet18()
   x = torch.randn(1, 3, 224, 224, requires_grad=True)
   onnx_file = "./resnet18.onnx"
   onnx.export(resnet18, x, onnx_file)

Is this a known regression?

I had the same problem with my network but managed to solve it by using onnx simplifier. This removes the extra Identity nodes again, which makes TensorRT able to use the ONNX graph again.

I would say, this should rather be reported as bug on GitHub - pytorch/pytorch: Tensors and Dynamic neural networks in Python with strong GPU acceleration