Skip to content

[BugFix]Disable dispatch_gmm_combine_decode operator when mtp drafter model uses non-w8a8 while main model uses w8a8, or drafter model is eagle series#5293

Merged
wangxiyuan merged 1 commit into
vllm-project:mainfrom
wangqiankun13:guard_fp32_mtp
Jan 4, 2026
Merged