Code

[pull] master from ggerganov:master#119

Code

Closed

pull[bot] wants to merge 12 commits intoteleprint-me:masterteleprint-me/llama.cpp:masterfrom ggml-org:masterggml-org/llama.cpp:masterCopy head branch name to clipboard

+1,485-503

Commits

Commits on Jun 19, 2024

• un-ignore build-info.cmake and build-info.sh (#7996)

  Show description for a785474
  mdegans

  authored
  a785474

  Copy full SHA for a785474

• server : fix smart slot selection (#8020)

  sasha0552

  authored
  ba58993

  Copy full SHA for ba58993

Commits on Jun 20, 2024

• metal : fix ggml_metal_supports_op for BF16 (#8021)

  Show description for 2075a66
  mdegans

  authored
  2075a66

  Copy full SHA for 2075a66

• CUDA: stream-k decomposition for MMQ (#8018)

  Show description for d50f889
  JohannesGaessler

  authored
  d50f889

  Copy full SHA for d50f889

• [SYCL] Fix windows build and inference (#8003)

  Show description for de391e4
  luoyu-intel

  authored
  de391e4

  Copy full SHA for de391e4

• common: fix warning (#8036)

  Show description for abd894a
  

  JohannesGaessler

  and

  slaren

  authored
  abd894a

  Copy full SHA for abd894a

• convert-hf : Fix the encoding in the convert-hf-to-gguf-update.py (#8040)

  hamdoudhakem

  authored
  17b291a

  Copy full SHA for 17b291a

• requirements : Bump torch and numpy for python3.12 (#8041)

  hamdoudhakem

  authored
  b1ef562

  Copy full SHA for b1ef562

Commits on Jun 21, 2024

• swiftui : enable stream updating (#7754)

  shu223

  authored
  0e64591

  Copy full SHA for 0e64591

• llama : allow pooled embeddings on any model (#7477)

  Show description for 80ea089
  iamlemec

  authored
  80ea089

  Copy full SHA for 80ea089

• llama : optimize long word tokenization with WPM (#8034)

  Show description for a927b0f
  ggerganov

  authored
  a927b0f

  Copy full SHA for a927b0f

• ggml : AVX IQ quants (#7845)

  Show description for 7d5e877
  netrunnereve

  authored
  7d5e877

  Copy full SHA for 7d5e877