feat(rm): Support loading FP4 and FP6 tensors
- Add support in readFromNpyFile.
- Add support in allocateTensor, which is used for constants read from the flatbuffers.
- Change IsFloat to take an input rather than a template parameter. This makes it possible to use the function in both constexpr and non-constexpr contexts.
Signed-off-by: Ian Tayler Lessa ian.taylerlessa@arm.com
Edited by Ian Tayler Lessa