Skill
nemo-mbridge-perf-moe-long-context
Created by nvidia
Version 392ab8ab2587345a9d86240967f666bd59ac82ca
Overview
Long-context MoE training guidance for Megatron Bridge. Covers CP sizing, selective recompute, dispatcher choices, and practical patterns from DSV3, Qwen3, and Qwen3-Next long-context experiments.