]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
CUDA: fix FA occupancy, optimize tile kernel (#15982)
authorJohannes Gäßler <redacted>
Wed, 17 Sep 2025 13:32:42 +0000 (15:32 +0200)
committerGitHub <redacted>
Wed, 17 Sep 2025 13:32:42 +0000 (15:32 +0200)
commitc959b676be29e93f8dbc3bd6056ceba812a9eb72
tree9bb2c7c424fff1c2531a197d2403f6aa317d8d7d
parentcd08fc3ecc0264b4414b68af3874a6c689ed60c1
CUDA: fix FA occupancy, optimize tile kernel (#15982)
ggml/src/ggml-cuda/common.cuh
ggml/src/ggml-cuda/fattn-common.cuh
ggml/src/ggml-cuda/fattn-tile.cu
ggml/src/ggml-cuda/vendors/hip.h