Jeff Bolz ca31169a53 vulkan: Support Q2_0 (llama/25430) ...

* vulkan: Support Q2_0

The backend perf tests for mat-vec-mul weren't very good at first (worse than
q2_k), doubling the rows per workgroup made a big difference.

* reorder

* resolve merge conflict, adjust err threshold for f16->q2_0 set_rows

2026-07-30 16:31:44 +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: Support Q2_0 (llama/25430)

2026-07-30 16:31:44 +03:00

.gitignore

…

CMakeLists.txt

ggml-et: Initial ET backend (llama/24179)

2026-07-30 16:31:36 +03:00