llama.cpp
ggml-cuda: fixes for concurrent streams
#18496
Merged

ggml-cuda: fixes for concurrent streams #18496

am17an merged 4 commits into ggml-org:master from am17an:graph-opt-fix
am17an
github-actions github-actions added Nvidia GPU
github-actions github-actions added ggml
am17an am17an force pushed from 901c8243 to 14642162 56 days ago
am17an am17an force pushed from 7e2dcb71 to 1fff4fc6 56 days ago
am17an am17an force pushed from 1fff4fc6 to 918ebb95 56 days ago
am17an am17an force pushed from 918ebb95 to 0bb52944 56 days ago
ggerganov
am17an am17an force pushed from 0bb52944 to b3a9b4ca 56 days ago
am17an
am17an
am17an ggml-cuda: enable concurrent streams by default
25ae7986
am17an am17an force pushed from b3a9b4ca to 25ae7986 56 days ago
ggerganov
am17an make flag opt-in
93cfa8d1
am17an
ggerganov
ggerganov
ggerganov commented on 2026-01-02
am17an add todo about special casing
d405fa1c
am17an am17an requested a review from JohannesGaessler JohannesGaessler 52 days ago
JohannesGaessler
JohannesGaessler approved these changes on 2026-01-03
am17an am17an changed the title ggml-cuda: enable concurrent streams by default ggml-cuda: fixes for concurrent streams 52 days ago
am17an
am17an update comment
b423920f
am17an am17an force pushed from c44291b0 to b423920f 52 days ago
am17an am17an merged e57f5233 into master 52 days ago
thomasjfox
am17an
thomasjfox
tccybo
am17an
am17an
thomasjfox
am17an
am17an am17an deleted the graph-opt-fix branch 51 days ago
thomasjfox
am17an
thomasjfox

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone