llama.cpp
vulkan: use smaller combined allocations to avoid fragmentation
#11551

Merged

vulkan: use smaller combined allocations to avoid fragmentation #11551

0cc4m merged 1 commit into ggml-org:master from jeffbolznv:smaller_allocations

jeffbolznv requested a review from

slaren 321 days ago

jeffbolznv requested a review from

0cc4m 321 days ago

slaren approved these changes on 2025-01-31

github-actions added Vulkan

github-actions added ggml

0cc4m commented on 2025-02-05

vulkan: use smaller combined allocations to avoid fragmentation

77009711

jeffbolznv force pushed to 77009711 316 days ago

0cc4m approved these changes on 2025-02-06

0cc4m merged 1b598b30 into master 315 days ago

Reviewers

0cc4m

slaren

Assignees

No one assigned

Labels

Vulkan ggml

Milestone

No milestone