sync : ggml #2991
ggerganov
merged
11 commits into
master
from
sync-ggml-25-04-02
cmake : fix whitespace (llama/0)
fc0b844e
Vulkan: Add DP4A MMQ and Q8_1 quantization shader (llama/12135)
ead4d064
ggml : faster ssm scan (llama/10558)
d9b53a01
SYCL: switch to SYCL namespace (llama/12674)
5014a789
SYCL: Rename oneMKL to oneMath (llama/12192)
959e437e
vulkan: fix build when glslc doesn't support coopmat (llama/12683)
690fa655
Fix clang warning in gguf_check_reserved_keys (llama/12686)
1f71018e
metal : use F32 prec in FA kernels (llama/12688)
ee44a0e3
opencl : fix memory allocation size (llama/12649)
e8c0a27e
get_rows and dup optimization (llama/12671)
91c0df58
sync : ggml
7ac68ebb
danbev
approved these changes
on 2025-04-02
Login to write a write a comment.
Login via GitHub