🐛Issue with brats_mri_segmentation bundle Torch_Tensorrt converting

[apbose]

Works in pytorch 22.10 docker but not in pytorch 23.01 docker.
See NVBugs for more details

[apbose]

The error comes in evaluator aten::slice.t(t[] l, int start, int end=9223372036854775807, int step=1) -> (t[]) where end is %self.convInit.conv.bias.27 : NoneType = prim::Constant() . It expects true in IValue isInt() unwrapping in unwraptoInt(), but the none type returns false.