Code

[pull] master from ggerganov:master#155

Code

Closed

pull[bot] wants to merge 30 commits intoteleprint-me:masterteleprint-me/llama.cpp:masterfrom ggerganov:masterggerganov/llama.cpp:masterCopy head branch name to clipboard

+2,816-3,333

Commits

Commits on Nov 25, 2024

• ci : build docker images only once daily (#10503)

  slaren

  authored
  50d5cec

  Copy full SHA for 50d5cec

• Introduce llama-run (#10291)

  Show description for 0cc6375
  ericcurtin

  authored
  0cc6375

  Copy full SHA for 0cc6375

Commits on Nov 26, 2024

• vulkan: Fix a vulkan-shaders-gen arugment parsing error (#10484)

  Show description for 0eb4e12
  sparkleholic

  authored
  0eb4e12

  Copy full SHA for 0eb4e12

• CANN: RoPE and CANCAT operator optimization (#10488)

  Show description for 7066b4c
  

  noemotiovon

  and

  noemotiovon

  authored
  7066b4c

  Copy full SHA for 7066b4c

• CANN: Improve the Inferencing Performance for Ascend NPU Device (#10454)

  Show description for 9a4b79b
  

  3 peopleauthored
  9a4b79b

  Copy full SHA for 9a4b79b

• speculative : simplify the implementation (#10504)

  Show description for 811872a
  ggerganov

  authored
  811872a

  Copy full SHA for 811872a

• server : fix parallel speculative decoding (#10513)

  Show description for 84e1c33
  ggerganov

  authored
  84e1c33

  Copy full SHA for 84e1c33

• ggml-cpu: cmake add arm64 cpu feature check for macos (#10487)

  Show description for 25669aa
  chaxu01

  authored
  25669aa

  Copy full SHA for 25669aa

• ci : add ubuntu cuda build, build with one arch on windows (#10456)

  slaren

  authored
  c6807b3

  Copy full SHA for c6807b3

• ci : publish the docker images created during scheduled runs (#10515)

  slaren

  authored
  7db3846

  Copy full SHA for 7db3846

• cmake : enable warnings in llama (#10474)

  Show description for ab96610
  ggerganov

  authored
  ab96610

  Copy full SHA for ab96610

• restore the condistion to build & update pacakge when merge (#10507)

  Show description for 0bbd226
  

  NeoZhangJianyu

  and

  arthw

  authored
  0bbd226

  Copy full SHA for 0bbd226

• server : replace behave with pytest (#10416)

  Show description for 45abe0f
  ngxson

  authored
  45abe0f

  Copy full SHA for 45abe0f

• vulkan: fix group_norm (#10496)

  Show description for 904109e
  jeffbolznv

  authored
  904109e

  Copy full SHA for 904109e

• mtgpu: Add MUSA_DOCKER_ARCH in Dockerfiles && update cmake and make (#10516)

  Show description for 249cd93
  yeahdongcn

  authored
  249cd93

  Copy full SHA for 249cd93

• Fix HIP flag inconsistency & build docs (#10524)

  Show description for be0e350
  tristandruyen

  authored
  be0e350

  Copy full SHA for be0e350

• llama : disable warnings for 3rd party sha1 dependency (#10527)

  slaren

  authored
  30ec398

  Copy full SHA for 30ec398

• ci : remove nix workflows (#10526)

  slaren

  authored
  5a349f2

  Copy full SHA for 5a349f2

• Add OLMo 2 model in docs (#10530)

  Show description for de50973
  2015aroras

  authored
  de50973

  Copy full SHA for de50973

• ci : fix cuda releases (#10532)

  slaren

  authored
  c9b00a7

  Copy full SHA for c9b00a7

Commits on Nov 27, 2024

• vulkan: optimize Q2_K and Q3_K mul_mat_vec (#10459)

  jeffbolznv

  authored
  4a57d36

  Copy full SHA for 4a57d36

• vulkan: skip integer div/mod in get_offsets for batch_idx==0 (#10506)

  jeffbolznv

  authored
  71a6498

  Copy full SHA for 71a6498

• vulkan: further optimize q5_k mul_mat_vec (#10479)

  jeffbolznv

  authored
  249a790

  Copy full SHA for 249a790

• vulkan: Handle GPUs with less shared memory (#10468)

  Show description for 5b3466b
  jeffbolznv

  authored
  5b3466b

  Copy full SHA for 5b3466b

• vulkan: define all quant data structures in types.comp (#10440)

  jeffbolznv

  authored
  c31ed2a

  Copy full SHA for c31ed2a

• Do not include arm_neon.h when compiling CUDA code (ggml/1028)

  

  frankier

  authored and

  ggerganov

  committed
  9150f8f

  Copy full SHA for 9150f8f

• sync : ggml

  ggerganov

  committed
  fee824a

  Copy full SHA for fee824a

• metal : fix group_norm support condition (#0)

  ggerganov

  committed
  9e2301f

  Copy full SHA for 9e2301f

• ci : faster CUDA toolkit installation method and use ccache (#10537)

  Show description for 46c69e0
  slaren

  authored
  46c69e0

  Copy full SHA for 46c69e0

• Add some minimal optimizations for CDNA (#10498)

  Show description for 3ad5451
  IMbackK

  authored
  3ad5451

  Copy full SHA for 3ad5451