perf: optimize guided decoding with xgrammar upgrade, batched API, and async D2H overlap#4605
Open
windreamer wants to merge 6 commits into
Open
perf: optimize guided decoding with xgrammar upgrade, batched API, and async D2H overlap#4605windreamer wants to merge 6 commits into
windreamer wants to merge 6 commits into
background
wait
wait-all
cancel
parallel
Loading