A guide on how to run Nemotron 3 Super 120B Thinking on 2 Nvidia DGX Spark

(corti.com)

2 points | by TechPreacher 5 hours ago

3 comments

orbanlevi 5 hours ago

I have 1 DGX Spark and running models with vLLM to, out of curiosity why not using Llama.cpp / TensorRT-LLM or any other alternatives?
awedisee 2 hours ago

Oh thank god. Finally a man of the people who can show us how to optimize 10k worth of equipment.
Because we all have at least two of these. Shout out to OP!!
TechPreacher 5 hours ago

[flagged]