]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 3962fc1)
HIP: force max threads per block to be 1024 (#11621)
authorfxzjshm <redacted>
Tue, 4 Feb 2025 18:18:38 +0000 (02:18 +0800)
committerGitHub <redacted>
Tue, 4 Feb 2025 18:18:38 +0000 (19:18 +0100)
commit3ec9fd4b77b6aca03a3c2bf678eae3f9517d6904
tree0c191b0c43a42b5d1d0d97071848030477888a37tree
parent3962fc1a7956dd0afacfbce10fd1a3ffd3ad857ecommit | diff
HIP: force max threads per block to be 1024 (#11621)

Some old/vendor forked version of llvm still use 256. Explicitly set it to 1024 to align with upstream llvm.

Signed-off-by: fxzjshm <redacted>
ggml/src/ggml-hip/CMakeLists.txt diff | blob | history
Packaging of ggml-org/llama.cpp