Hello! A friend is helped me to installing Ollama to my local pc and ask tech questions from it. The idea is that since I have 48 gb ram on my Ada-6000 that allows me to use the larger knowledge / understanding model.
I found out that my Ollama with llama3.3:70b-instruct-q4_K_M is pretty wise! See how well it knows overall info about Omniverse Action Graph
There has been a Jensen official announcement for example that NVIDIA has teached USD file format to AI.
There is the “Nemotron” variant of LLama, built by NVidia. I wonder if that one has more Nvidia USD / Omniverse knowledge?
The model card states:
“Llama-3.1-Nemotron-70B-Instruct is a large language model customized by NVIDIA to improve the helpfulness of LLM generated responses to user queries.”
I am in AI Enterprise full pro version.
Currently I use my startup Cineshare.io dev team and skilled omniverse dev friends and team green members here at discord to solve my tech issues, maybe these tools will help me to solve some of my tasks, regarding the omniverse to build digital twins and my colorful character works like ACE.