GPU usage

Infrastructure includes GPU usage as a monitored specification. GPU usage monitoring depends on the elin server. The DataLens DS-STAR Implementation Plan includes the GPU Infrastructure as a requirement. GPU Infrastructure requires deployment of vLLM with Qwen2.5-Coder-14B-AWQ model. GPU Infrastructure requires deployment of Qdrant vector database. GPU Infrastructure requires installation of DuckDB database system. GPU Infrastructure requires Python environment setup with all dependencies. GPU Infrastructure uses vLLM for large language model execution on elin. GPU Infrastructure includes the use of Qdrant vector database for semantic search capabilities. GPU Infrastructure uses vLLM for large language model execution on elin. GPU Infrastructure includes the use of Qdrant vector database for semantic search capabilities. GPU Infrastructure uses vLLM for large language model execution on elin. GPU Infrastructure includes the use of Qdrant vector database for semantic search capabilities. The plan considers GPU usage on elin especially for Ollama calls and embedding models.