Jeff Bolz cb8a0ad83c vulkan: add iq4_nl support back to FA (llama/24585) ...

* vulkan: add iq4_nl support back to FA

I was originally concerned about wasting shared memory on the LUT, but it's small
and unlikely to matter in practice.

Also support q1_0 for non-coopmat2.

Fixes #23681

* remove q1_0 FA support

2026-07-30 16:31:50 +03:00

..

cmake

ggml : Parallelize quant LUT init (llama/23595)

2026-05-25 12:26:07 +03:00

include

DeepseekV4: Add fused hyper-connection ops (llama/25585)

2026-07-30 16:31:43 +03:00

src

vulkan: add iq4_nl support back to FA (llama/24585)

2026-07-30 16:31:50 +03:00

.gitignore

…

CMakeLists.txt

HIP: remove rocWMMA FlashAttention (llama/26046)

2026-07-30 16:31:48 +03:00