I tested it with the DGX Spark in my home-lab and it works well; however, for agentic tasks, I didn't get output as optimal (in terms of speed/quality) as I did

I tested it with the DGX Spark in my home-lab and it works well; however, for agentic tasks, I didn't get output as optimal (in terms of speed/quality) as I did with Nemotron. Also ref: https://developer.nvidia.com/blog/run-local-agentic-ai-workflows-with-metas-muse-glimmer-on-nvidia/ https://developer.nvidia.com/blog/run-local-agentic-ai-workf...