The Conv and MatMul templates do not support batch

What commit version of aidge do you use

• convolution on aidge_export_cpp on main: 838aa704
• convolution on aidge_export_cpp on dev: e98d3eee
• matmul on aidge_export_cpp on main and dev: d68ceb49

Problem description

The issue was identified during the development of a C-based exporter, comparing the CPU backend results with the CPP and C export results. A code review highlighted that neither the MatMul nor the Conv2D kernels in the CPP export use the batch parameter. The batch parameter is explicitly mentioned in the documentation for the MatMul operator.