llama: optimize RWKV7 inference by fusing some graph operators - #25206
Draft
MollySophia wants to merge 16 commits into
Draft
llama: optimize RWKV7 inference by fusing some graph operators#25206MollySophia wants to merge 16 commits into
MollySophia wants to merge 16 commits into
background
wait
wait-all
cancel
parallel
Loading