Code

Add support for AVX VNNI #4589

Code

Merged

ggerganov merged 9 commits intoggml-org:masterggml-org/llama.cpp:masterfrom janhq:masterjanhq/llama.cpp-avx-vnni:masterCopy head branch name to clipboardDec 30, 2023

+33-8

Commits

Commits on Dec 22, 2023

• feat: add avx_vnni based on intel documents

  tikikun

  committed
  5537026

  Copy full SHA for 5537026

• ggml: add avx vnni based on intel document

  tikikun

  committed
  32ca569

  Copy full SHA for 32ca569

• llama: add avx vnni information display

  tikikun

  committed
  621af1b

  Copy full SHA for 621af1b

• docs: add more details about using oneMKL and oneAPI for intel processors

  tikikun

  committed
  0c35ed7

  Copy full SHA for 0c35ed7

• docs: add more details about using oneMKL and oneAPI for intel processors

  tikikun

  committed
  47f0c7b

  Copy full SHA for 47f0c7b

• docs: add more details about using oneMKL and oneAPI for intel processors

  tikikun

  committed
  08d9475

  Copy full SHA for 08d9475

• docs: add more details about using oneMKL and oneAPI for intel processors

  tikikun

  committed
  4caa9e2

  Copy full SHA for 4caa9e2

• docs: add more details about using oneMKL and oneAPI for intel processors

  tikikun

  committed
  66ae732

  Copy full SHA for 66ae732

Commits on Dec 30, 2023

• Update ggml.c

  Show description for e92aceb
  

  tikikun

  and

  ggerganov

  authored
  e92aceb

  Copy full SHA for e92aceb