✨[Feature] Fully support `FakeTensor`s in Dynamo compile

[gs-olive]

Context

Currently, the torch.compile backend for Torch-TensorRT requires @fake_tensor_unsupported due to a few functions which cannot handle fake tensors. This support needs to be updated in order to support fake tensors on the path to full-scale dynamic shape support.

Key Consideration: Is it possible to support fake tensors while preserving the notion of "constant" tensors in TRT?

Proposed Changes

The to_numpy function will need to be updated to take a scalar:

TensorRT/py/torch_tensorrt/fx/converters/converter_utils.py

Line 154 in c7e79b2

def to_numpy(tensor: Optional[torch.Tensor]) -> Optional[np.ndarray]:

The create_constant function will need to be updated to not transform scalars into tensors:

TensorRT/py/torch_tensorrt/fx/converters/converter_utils.py

Line 226 in c7e79b2

def create_constant(

Special considerations must be made for constant tensors with more than one element.