openvino

9e83aea1 - [NPU] Acknowledge OV model compression in weightless compilation (#34609)

[NPU] Acknowledge OV model compression in weightless compilation (#34609) ### Details: - OV model compression may produce models with multiple constants pointing to the same binary data. If this is the case, rely on compiler to provide actual constant description (shape and precision), keeping the plugin code simple. Additionally, introduce several checks to ensure that the buffer sizes are the same. They are relatively cheap computation-wise, yet allow to ensure model invariants. - Changes and description by @andrey-golubev, small addition by me. ### Tickets: - *EISW-204968* ### AI Assistance: - *AI assistance used: no* - *If yes, summarize how AI was used and what human validation was performed (build/tests/manual checks).* --------- Co-authored-by: Golubev, Andrey <andrey.golubev@intel.com>