add openvino support for igpu acceleation

[jsapede]

Is your feature request related to a problem? Please describe.
only SYCL acceleraiton is available but intel CPU, even old (> gen6) includes openvino acceleration witch works pretty well (equivalent to google coral for what i've tested on frigate)

enabling openvino could strongly accelerate models even in old / small / cheap systems, no need for cuda

Describe the solution you'd like
docker images including openvino and /dev/dri/renderD128 passthrough to access igpu

Describe alternatives you've considered

Additional context
buildind accelerated local hosted voice assistants for homeassistant including distil-whisper or whisper.cpp / wyioming protocol

[mudler]

Thanks for showing interest in LocalAI and OpenVINO , I'll close this one as there is already another one already opened here: #1722

[jsapede]

ok thanks !