[CPU] Disable ConvertGatherToGatherCompressed optimization for quantized models #24796

Job	Run time
Smart_CI	27s
Docker	49s
Build	7m 46s
Debian Packages	0s
PyTorch Models tests	0s
OpenVINO JS API / OpenVINO JS API	2m 18s
C++ unit tests / C++ unit tests	11m 56s
OpenVINO tokenizers extension / OpenVINO tokenizers extension	1m 51s
Python unit tests / Python unit tests	12m 31s
Samples / Samples	13m 39s
CPU functional tests / CPU functional tests	5m 11s
ONNX Runtime Integration / ONNX Runtime Integration	4m 27s
TensorFlow Models tests	0s
TensorFlow Layer Tests / TensorFlow Layer Tests	4m 17s
ci/gha_overall_status_linux_arm64	1s
	1h 5m 13s

Provide feedback