llama.cpp
cuda : enable CUDA graphs for MMID 1 <= BS <= 4
#19645
Merged

cuda : enable CUDA graphs for MMID 1 <= BS <= 4 #19645

ggerganov merged 3 commits into master from gg/cuda-graphs-enable-bs-gt1
ggerganov
github-actions github-actions added Nvidia GPU
github-actions github-actions added ggml
ggerganov ggerganov requested a review from JohannesGaessler JohannesGaessler 4 days ago
ggerganov ggerganov requested a review from am17an am17an 4 days ago
am17an
am17an commented on 2026-02-15
Base automatically changed from gg/graph-fix-kq-mask-reuse to master 3 days ago
ggerganov ggerganov requested a review from CISC CISC 3 days ago
ggerganov cuda : enable CUDA graphs for MMID BS <= 4
7d0be2c4
ggerganov ggerganov force pushed from 15a68420 to 7d0be2c4 3 days ago
ORippler
ORippler commented on 2026-02-16
ORippler
ggerganov
ggerganov commented on 2026-02-16
ggerganov cont : add stream capture check
dd9af011
JohannesGaessler
JohannesGaessler approved these changes on 2026-02-17
ggerganov cont : add MMVQ_MMID_MAX_BATCH_SIZE
573b94ca
ggerganov
ggerganov ggerganov merged ad8207af into master 2 days ago
ggerganov ggerganov deleted the gg/cuda-graphs-enable-bs-gt1 branch 2 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone