[FEATURE] Add SYCL feature flags to rllm-llamacpp build (To add support for Intel GPUs)

[TheMrCodes]

The goal of this pull request is to add support for Intel GPU to the build script of rllm llama.cpp This is done by

• expanding the cargo feature flags of rllm-llamacpp and rllm-llamacpp_low and incoporate
• adding the needed build arguments in the main.rs from rllm-llamacpp_low
• and adding appropriate new options to server.sh

The code will be tested on a machine with Intel 13Gen Processor & Intel Arc A770 GPU

[TheMrCodes]

@microsoft-github-policy-service agree

[TheMrCodes]

The basic build works now. 🚀 Want to test the feature again if the llama.cpp version got bumped

Known Issues:

• llama.cpp hangs after loading model into VRAM on Intel ARC -> This is a bug of the current llama.cpp version and has to be tested again after an upgrade
• currently only FP32 build works -> This also could depend on the current version but further testing has to be done