b11534: CUDA: Remove redundant CUDA copies after SSM_SCAN (#29807)
Mirrored from llama.cpp releases for archival readability. Support the source by reading on the original site.
-
CUDA: fuse copy of updated state snapshots into recurrent cache with ssm_scan
-
CUDA: remove redundant cuda copies with K==1 (non spec-dec) scenario as well
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.