NXP backend: Add support for optimizing Conv+BN during QAT#16246

roman-janik-nxp merged 6 commits intopytorch:mainpytorch/executorch:mainfrom

nxp-upstream:EIEX-650-fix-native-conv-batchnorm-fusing-leaving-artefactsnxp-upstream/executorch:EIEX-650-fix-native-conv-batchnorm-fusing-leaving-artefactsCopy head branch name to clipboard