feat: Add Intel GPU Support via SYCL on Linux platform

[xiangyang-95]

Problem
Add support for Intel GPU via Intel OneAPI SYCL on Linux

Success Criteria

• Add Linux Runner with Intel OneAPI container
• Add build steps using Intel OneAPI container
• Add app running steps with Intel OneAPI container

Additional context
N/A

[dan-menlo]

I am merging this issue into janhq/cortex.cpp#470. Thanks @xiangyang-95

• llama.cpp also supports sycl now
• We may work more closely with Intel on Lunar Lake support