C++ FakeTensor catches up to Python, and two ExecuTorch correctness fixes
- Fix the linear_q4gsw reference implementation for output channels not a multiple of 4 (#24930) (#24930) pytorch/executorch
- Do not delegate quantized linears that have no Vulkan kernel (#24971) (#24971) pytorch/executorch
- Phonon: streaming output and per-backend weight quantization (#24929) pytorch/executorch