fixing the DTensor import for igpu platforms

Author: apbose

py/torch_tensorrt/dynamo/backend/backends.py

@@ -10,7 +10,6 @@
 from torch._dynamo.backends.common import aot_autograd
 from torch._dynamo.utils import detect_fake_mode
 from torch._functorch.aot_autograd import aot_export_joint_simple
-from torch.distributed.tensor import DTensor
 from torch_tensorrt.dynamo import CompilationSettings
 from torch_tensorrt.dynamo._compiler import compile_module
 from torch_tensorrt.dynamo.lowering import (