NPUW: Enable PREFILL/GENERATE configs in LLMCompiledModel #15916

Job	Run time
Smart_CI	18s
Docker	47s
Build / Build	12m 20s
Conformance (OP)	9m 17s
dGPU Tests	0s
iGPU Tests	0s
Conformance (API)	2m 17s
NVIDIA plugin	2m 41s
CPU functional tests / CPU functional tests	21m 48s
C++ unit tests / C++ unit tests	14m 59s
Debian Packages / Debian Packages	1m 46s
JAX Layer Tests / JAX Layer Tests	1m 59s
OpenVINO JS API / OpenVINO JS API	1m 24s
ONNX Models Tests / ONNX Models tests	19m 57s
ONNX Runtime Integration / ONNX Runtime Integration	6m 7s
OpenVINO tokenizers extension / OpenVINO tokenizers extension	2m 52s
Python API tests / Python API tests	2m 22s
Python unit tests / Python unit tests	4m 22s
Pytorch Layer Tests / PyTorch Layer Tests	16m 39s
Samples / Samples	4m 32s
PyTorch Models Nightly Scope1 tests	0s
TensorFlow TF Hub Models tests	0s
TensorFlow Hugging Face Models tests	0s
PyTorch Models Nightly Scope2 tests	0s
TensorFlow Layer Tests / TensorFlow Layer Tests	13m 36s
TensorFlow Models tests / TensorFlow Models tests	14m 51s
PyTorch Models tests / PyTorch Models tests	32m 45s
JAX/Flax Models tests / JAX/Flax Models tests	1m 44s
ci/gha_overall_status	1s
	3h 9m 24s

Provide feedback