llama.cpp
02115dcd - vulkan: Allow up to 4096 elements for mul_mat_id row_ids (#13326)

Commit

97 days ago

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.

References

#13326 - vulkan: Allow up to 4096 elements for mul_mat_id row_ids

Author

jeffbolznv

Parents

d9c4acca

Files4

ggml/src/ggml-vulkan
- ggml-vulkan.cpp
- vulkan-shaders
  - mul_mm.comp
  - mul_mm_cm2.comp
  - mul_mmq.comp

llama.cpp 02115dcd - vulkan: Allow up to 4096 elements for mul_mat_id row_ids (#13326)

llama.cpp
02115dcd - vulkan: Allow up to 4096 elements for mul_mat_id row_ids (#13326)