]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
ggml : support AVX512VNNI (llama/6280)
authorJustine Tunney <redacted>
Mon, 25 Mar 2024 05:39:56 +0000 (01:39 -0400)
committerGeorgi Gerganov <redacted>
Wed, 27 Mar 2024 11:20:00 +0000 (13:20 +0200)
commit2b9042a364c97703d3a4cb892e7524e3a20b7499
tree0b92da52062575ca1e46b07582edf7ffe02d11fa
parentfc815abac49a8d23d14eeb1060c5c87f1a0b0255
ggml : support AVX512VNNI (llama/6280)

This change causes some quants (e.g. Q4_0, Q8_0) to go faster on some
architectures (e.g. AMD Zen 4).
src/ggml-quants.c