llama.cpp
llama_dsv4: write only used rows in state
#25325
Merged

llama_dsv4: write only used rows in state #25325

am17an merged 2 commits into ggml-org:master from am17an:dsv4-cache
am17an
am17an am17an requested a review from ggerganov ggerganov 132 days ago
ggerganov
ggerganov commented on 2026-07-06
am17an am17an requested a review from CISC CISC 132 days ago
am17an am17an force pushed 132 days ago
am17an am17an force pushed 132 days ago
ggerganov
am17an llama_dsv4: write only used rows in state
4b742ccb
am17an add TODO about conflating token pos with kv rows
1fc9b63e
am17an am17an force pushed to 1fc9b63e 118 days ago
fairydreaming
fairydreaming approved these changes on 2026-07-20
am17an
ruixiang63
ruixiang63 approved these changes on 2026-07-20
am17an am17an merged 91d2fc38 into master 117 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone