]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 0516e04)
vulkan: Fix ErrorOutOfHostMemory on Intel GPU when loading large models with --no...
authorMasato Nakasaka <redacted>
Thu, 12 Mar 2026 05:30:16 +0000 (22:30 -0700)
committerGitHub <redacted>
Thu, 12 Mar 2026 05:30:16 +0000 (06:30 +0100)
commit5866e3bbc8a2ede4236871ea73a3eb3337abf530
tree249d97458c233250f2ab993906642472c3c09d60tree
parent0516e04bf9c002bf190519af6795415ab3ebd059commit | diff
vulkan: Fix ErrorOutOfHostMemory on Intel GPU when loading large models with --no-mmap (#20059)

* Changed to reuse command buffers to fix crashing on Intel GPU

* Removed unused parameter

* Fixed compile error and minor mistake

* Fix logging

* Changing to use usage flag per command buffer

* fixed style

* added buffer reset

* Removed cmd_buffer_idx for reuse consistency

* Fixed style
ggml/src/ggml-vulkan/ggml-vulkan.cpp diff | blob | history
Packaging of ggml-org/llama.cpp