Code

AVX IQ Quants#7845

Code

Merged

ggerganov merged 16 commits intoggerganov:masterggerganov/llama.cpp:masterfrom netrunnereve:avx_iqCopy head branch name to clipboardJun 21, 2024

+698-1

Commits

Commits on Jun 10, 2024

• initial iq4_xs

  netrunnereve

  committed
  0fd5a1b

Commits on Jun 11, 2024

• fix ci

  netrunnereve

  committed
  b7e1707

• Merge branch 'ggerganov:master' into avx_iq

  netrunnereve

  authored
  2f37328

Commits on Jun 12, 2024

• iq4_nl

  netrunnereve

  committed
  8d1d112

Commits on Jun 13, 2024

• iq1_m

  netrunnereve

  committed
  5ff64ad

Commits on Jun 14, 2024

• iq1_s

  netrunnereve

  committed
  75370d7

• iq2_xxs

  netrunnereve

  committed
  65765c9

• Merge branch 'ggerganov:master' into avx_iq

  netrunnereve

  authored
  520361f

Commits on Jun 15, 2024

• iq3_xxs

  netrunnereve

  committed
  5926186

• iq2_s

  netrunnereve

  committed
  dcfee06

• iq2_xs

  netrunnereve

  committed
  eccc609

• iq3_s before sllv

  netrunnereve

  committed
  39e816e

Commits on Jun 16, 2024

• iq3_s

  netrunnereve

  committed
  99f666c

• iq3_s small fix

  netrunnereve

  committed
  b57187f

Commits on Jun 18, 2024

• iq3_s sllv can be safely replaced with sse multiply

  netrunnereve

  committed
  29e2a96

Commits on Jun 21, 2024

• Merge branch 'ggerganov:master' into avx_iq

  netrunnereve

  authored
  a055767