]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
vulkan: Allow up to 4096 elements for mul_mat_id row_ids (#13326)
authorJeff Bolz <redacted>
Fri, 9 May 2025 07:23:41 +0000 (02:23 -0500)
committerGitHub <redacted>
Fri, 9 May 2025 07:23:41 +0000 (09:23 +0200)
commit02115dcd9a6d0c03b527d5bee8c1493ab819ddbd
treef15e74144eb33e70ee14ebdbf2cd4d35182e3e6a
parentd9c4accaff30926e8a5fd8a2429e549158a845e0
vulkan: Allow up to 4096 elements for mul_mat_id row_ids (#13326)

This assert fired running Qwen_Qwen3-30B-A3B-Q2_K.gguf:

GGML_ASSERT(nei0 * nei1 <= 3072);

The tensor is 8 x 512. Increase this array size to accommodate.
ggml/src/ggml-vulkan/ggml-vulkan.cpp
ggml/src/ggml-vulkan/vulkan-shaders/mul_mm.comp
ggml/src/ggml-vulkan/vulkan-shaders/mul_mm_cm2.comp
ggml/src/ggml-vulkan/vulkan-shaders/mul_mmq.comp