[tmva][sofie] Add support for parsing onnx::Add operator in PyTorch parser

[MrViiV]

This Pull request:

Changes or fixes:

This PR adds support for parsing the onnx::Add operator in the SOFIE PyTorch parser (RModelParser_PyTorch.cxx).

The Add operator is a fundamental element-wise tensor operation and is widely used in modern neural network architectures, including residual connections (e.g. ResNet), transformer blocks, and other composite models exported from PyTorch.

Specifically, this PR:

• Implements the MakePyTorchAdd function to prepare the corresponding ROperator_BasicBinary with EBasicBinaryOperator::Add
• Registers the "onnx::Add" operator in the PyTorch parser operator map
• Reuses the existing SOFIE binary operator infrastructure, ensuring consistency with the ONNX parser implementation

This improves the PyTorch parser’s compatibility with real-world exported ONNX models.

Checklist:

• tested changes locally
• updated the docs (if necessary)

Testing:

A minimal PyTorch model containing an Add operation was exported to ONNX and verified to produce an onnx::Add node compatible with this parser extension.

[MrViiV]

@lmoneta @sanjibansg
This PR adds support for parsing onnx::Add operator in the PyTorch parser. I would appreciate your feedback.