]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 7398427)
test-backend-ops : use flops for some performance tests (#9657)
authorslaren <redacted>
Sat, 28 Sep 2024 12:32:46 +0000 (14:32 +0200)
committerGitHub <redacted>
Sat, 28 Sep 2024 12:32:46 +0000 (14:32 +0200)
commit1b2f992cd2cff0b69e5abe78bb8888d51ed19d67
treec1483b0e44b1d49fdfba04f102d452bd6faee23ctree
parent739842703e32cd43443c45e0b4f6647cc4e6b3d6commit | diff
test-backend-ops : use flops for some performance tests (#9657)

* test-backend-ops : use flops for some performance tests

- parallelize tensor quantization

- use a different set of cases for performance and correctness tests

- run each test for at least one second
tests/test-backend-ops.cpp diff | blob | history
Packaging of ggml-org/llama.cpp