[CaptainOfCoit]

There wasn't any instructions how the author got ollama/llama.cpp, could possibly be something nvidia shipped with the DGX Spark and is an old version?