Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 1, 2024

• Merge branch 'ggerganov:master' into master

  salaxieb

  authored
  1df9cef

  Copy full SHA for 1df9cef

• cann: support q8_0 for Ascend backend (#8805)

  wangshuai09

  authored
  c8a0090

  Copy full SHA for c8a0090

Commits on Jul 31, 2024

• server : update llama-server embedding flag documentation (#8779)

  Show description for afbbcf3
  okigan

  authored
  afbbcf3

  Copy full SHA for afbbcf3

• Build: Fix potential race condition (#8781)

  Show description for ed9d285
  HanClinto

  authored
  ed9d285

  Copy full SHA for ed9d285

• Adding Gemma 2 2B configs (#8784)

  Show description for 398ede5
  

  pculliton

  and

  slaren

  authored
  398ede5

  Copy full SHA for 398ede5

• cmake : fix use of external ggml (#8787)

  iboB

  authored
  44d28dd

  Copy full SHA for 44d28dd

Commits on Jul 30, 2024

• nix: cuda: rely on propagatedBuildInputs (#8772)

  Show description for 268c566
  SomeoneSerge

  authored
  268c566

  Copy full SHA for 268c566

• py: add_array() will not add to kv store if value is an empty array (#8774)

  Show description for 7e72aa7
  

  mofosyne

  and

  compilade

  authored
  7e72aa7

  Copy full SHA for 7e72aa7

• added android implementation of ggml_print_backtrace_symbols (#8751)

  Show description for 7c27a19
  

  l3utterfly

  and

  slaren

  authored
  7c27a19

  Copy full SHA for 7c27a19

• flake.lock: Update (#8729)

  ggerganov

  authored
  140074b

  Copy full SHA for 140074b

• cann: update cmake (#8765)

  wangshuai09

  authored
  6e2b600

  Copy full SHA for 6e2b600

• [SYCL] Add `TIMESTEP_EMBEDDING` OP (#8707)

  Show description for c887d8b
  zhentaoyu

  authored
  c887d8b

  Copy full SHA for c887d8b

Commits on Jul 29, 2024

• ggml: bugfix: fix the inactive elements is agnostic for risc-v vector (#8748)

  Show description for 75af08c
  

  CarterLi999

  and

  carter.li

  authored
  75af08c

  Copy full SHA for 75af08c

• cuda : organize vendor-specific headers into vendors directory (#8746)

  Show description for 439b3fc
  yeahdongcn

  authored
  439b3fc

  Copy full SHA for 439b3fc

• [SYCL] add conv support (#8688)

  airMeng

  authored
  0832de7

  Copy full SHA for 0832de7

Commits on Jul 28, 2024

• cmake: use 1 more thread for non-ggml in CI (#8740)

  JohannesGaessler

  authored
  6eeaeba

  Copy full SHA for 6eeaeba

• chore : Fix vulkan related compiler warnings, add help text, improve CLI options (#8477)

  Show description for 4730fac
  

  teleprint-me

  and

  0cc4m

  authored
  4730fac

  Copy full SHA for 4730fac

• llama : refactor session file management (#8699)

  Show description for 4c676c8
  compilade

  authored
  4c676c8

  Copy full SHA for 4c676c8

Commits on Jul 27, 2024

• feat: Support Moore Threads GPU (#8383)

  Show description for e54c35e
  yeahdongcn

  authored
  e54c35e

  Copy full SHA for e54c35e

• scripts : sync vulkan-shaders (#0)

  ggerganov

  committed
  5e2727f

  Copy full SHA for 5e2727f

• scripts : sync ggml-aarch64 sources

  ggerganov

  committed
  56f20aa

  Copy full SHA for 56f20aa

• ggml : add missing semicolon (#0)

  Show description for 345c8c0
  ggerganov

  committed
  345c8c0

  Copy full SHA for 345c8c0

• sync : ggml

  Show description for ae7985c
  ggerganov

  committed
  ae7985c

  Copy full SHA for ae7985c

• ggml : loop tiling optimizations for scalar path (ggml/898)

  Show description for a05ca93
  

  heshpdx

  authored and

  ggerganov

  committed
  a05ca93

  Copy full SHA for a05ca93

• ggml: add support for float16 input tensors in pooling operations (ggml/895)

  Show description for 9f77d89
  

  2 peopleauthored and

  ggerganov

  committed
  9f77d89

  Copy full SHA for 9f77d89

• vulkan : initialize vk_buffer_struct members to VK_NULL_HANDLE (ggml/893)

  Show description for 203b7f1
  

  neobrain

  authored and

  ggerganov

  committed
  203b7f1

  Copy full SHA for 203b7f1

• cmake : only enable GGML_NATIVE and x86 flags if not crosscompiling (ggml/885)

  

  iboB

  authored and

  ggerganov

  committed
  d2b851b

  Copy full SHA for d2b851b

• ggml : remove unnecessary UNUSED macro call (ggml/880)

  Show description for c12b6e8
  

  danbev

  authored and

  ggerganov

  committed
  c12b6e8

  Copy full SHA for c12b6e8

• llama : add support for llama 3.1 rope scaling factors (#8676)

  Show description for b5e9546
  

  jmorganca

  and

  compilade

  authored
  b5e9546

  Copy full SHA for b5e9546

• llama : add function for model-based max number of graph nodes (#8622)

  Show description for 92090ec
  ggerganov

  authored
  92090ec

  Copy full SHA for 92090ec

• common : add --no-warmup option for main/llama-cli (#8712)

  Show description for 9d03d08
  danbev

  authored
  9d03d08

  Copy full SHA for 9d03d08

• cann: Fix Multi-NPU execution error (#8710)

  Show description for bfb4c74
  wangshuai09

  authored
  bfb4c74

  Copy full SHA for bfb4c74

• ggml : reduce hash table reset cost (#8698)

  Show description for 2b1f616
  slaren

  authored
  2b1f616

  Copy full SHA for 2b1f616

Commits on Jul 26, 2024

• llama : fix order of parameters (#8706)

  Show description for 01245f5
  

  foldl

  and

  Judd

  authored
  01245f5

  Copy full SHA for 01245f5

Commits on Jul 25, 2024

• server : add Speech Recognition & Synthesis to UI (#8679)

  Show description for 01aec4a
  ElYaiko

  authored
  01aec4a

  Copy full SHA for 01aec4a

Pagination

Previous

Next