]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 5910ea9)
ggml : AVX2 support for Q4_0_8_8 (#8713)
authorSrihari-mcw <redacted>
Wed, 4 Sep 2024 16:51:22 +0000 (22:21 +0530)
committerGitHub <redacted>
Wed, 4 Sep 2024 16:51:22 +0000 (19:51 +0300)
commit581c305186a0ff93f360346c57e21fe16e967bb7
tree31a2b5b9a3efdd5988f3c6b0e78ad933a0e7a889tree
parent5910ea942772ab6cbc21d0ad2d1208750ba39e1dcommit | diff
ggml : AVX2 support for Q4_0_8_8 (#8713)

* Add AVX2 based implementations for quantize_q8_0_4x8, ggml_gemv_q4_0_8x8_q8_0 and ggml_gemm_q4_0_8x8_q8_0 functions

* Update code to fix issues occuring due to non alignment of elements to be processed as multiple of 16 in MSVC

* Update comments and indentation

* Make updates to reduce number of load instructions
ggml/src/ggml-aarch64.c diff | blob | history
Packaging of ggml-org/llama.cpp