Knowledge Agent
Air-gapped RAG stack (vLLM + Dify + Qdrant) built to run Qwen2.5-72B locally and scale from one A100 to a 4-GPU cluster with no code changes.
No README content was published for this project.
Search hackers, navigate pages
Air-gapped RAG stack (vLLM + Dify + Qdrant) built to run Qwen2.5-72B locally and scale from one A100 to a 4-GPU cluster with no code changes.