Devotd Github

synchopate/nemotron-30b-cpu-256k

synchopate/nemotron-30b-cpu-256k presents an optimized configuration for running a 30B MoE model with a 256K context on a CPU.

View on GitHub ↗
stars
★ 21
Forks
2
Language
Python
License
—

Posts

1