Support quantized Qwen3-VL / Qwen3.5-VL (dense + MoE) export from Megatron-Bridge and verify exported checkpoints - #2276

Support quantized Qwen3-VL / Qwen3.5-VL (dense + MoE) export from Megatron-Bridge and verify exported checkpoints#2276

kevalmorabia97 wants to merge 20 commits into

mainNVIDIA/Model-Optimizer:mainfrom

kmorabia/mbridge-qwen3vl-quantized-hf-exportNVIDIA/Model-Optimizer:kmorabia/mbridge-qwen3vl-quantized-hf-exportCopy head branch name to clipboard