@routineofaii: Looking to run local GraphRAG or other graph analytics use cases? With DGX Spark, you can prepare your local text files for graph use cases at your desk. In this demo, a Llama-3.1-70B model served by Ollama runs locally in DGX Spark’s 128GB of unified memory. After uploading local biomedical research papers into the LLM, the model extracts nodes and relationships that are embedded into a knowledge graph. By processing entire documents at once, the knowledge graph retains critical long-range relationships that make GraphRAG and other graph analytics more accurate. The entire demo runs on DGX Spark, enabling data privacy, security, and avoiding lag time with moving development to a data center. We use ArangoDB’s graph database to host our knowledge graph and three.js to explore it visually. #fyp #nvidia #graphicscard #graph #llama

Routine of AI

Open In TikTok:

Region: NP

Tuesday 04 November 2025 18:11:00 GMT