Documentation, code, application and tutorial material for cloud-HPC integration with initial LLM use in OSS

[K123AsJ0k1]

As per request, this merge provides cloud-HPC integration changes into the main that were developed in the https://github.com/K123AsJ0k1/cloud-hpc-oss-mlops-platform fork. The changes are mostly additive with exceptions being deployment image updates and new setup script suggestion. The main changes are the following:

• NVIDIA GPU support with GPU operator
• Sharing 1 GPU between 10 pods via nvshare
• Ray GPU configuration
• LLM related inference and storage deployments
• RAG Ray code with pararellism using multiple CPU workers and a GPU actor
• Cloud-HPC integration documentation, applications and tutorial

These were developed and tested in a Ubuntu 22.04 VM run in CSC CPouta