]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/commit
vulkan: Support mul_mat_id with f32 accumulators (llama/15337)
authorJeff Bolz <redacted>
Sat, 16 Aug 2025 09:18:31 +0000 (04:18 -0500)
committerGeorgi Gerganov <redacted>
Mon, 18 Aug 2025 17:30:45 +0000 (20:30 +0300)
commitd14e626e6a4f1623b50f625a0628eb8c76296556
tree5f369b09ce66bc0d9b395a3d922c22197ccc3990
parent5b62995350955b4e5feb3dd74c9464caa2cc6db7
vulkan: Support mul_mat_id with f32 accumulators (llama/15337)

* vulkan: Add missing bounds checking to scalar/coopmat1 mul_mat_id

* vulkan: Support mul_mat_id with f32 accumulators, but they are not hooked up

- There's no explicit way to request f32 precision for mul_mat_id, but there
probably should be, and this gets the code in place for that.
- A couple fixes to check_results.
- Remove casts to fp16 in coopmat1 FA shader (found by inspection).
ggml/src/ggml-vulkan/ggml-vulkan.cpp
ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp