Oliver Simons a84dfd3e10 CUDA: Add Cooperative-Groups-based parallelization of ncols in softmax ...

Old implementation parallelizes rows across SMs, which does not fit the
needs of backend-sampling (where we have ncols >> nrows and thus want to
parallelize ncols across SMs)

2025-12-09 12:58:56 +01:00

..

cmake

ggml: Skip backend library linking code when GGML_BACKEND_DL=ON (#15094)

2025-08-07 13:45:41 +02:00

include

ggml-zendnn : add ZenDNN backend for AMD CPUs (#17690)

2025-12-07 00:13:33 +08:00

src

CUDA: Add Cooperative-Groups-based parallelization of ncols in softmax

2025-12-09 12:58:56 +01:00

.gitignore

vulkan : cmake integration (#8119)

2024-07-13 18:12:39 +02:00

CMakeLists.txt

ggml-cpu: add ggml_thread_cpu_relax with Zihintpause support (#17784)

2025-12-08 10:41:34 +02:00