Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit history

Commits on Aug 20, 2025

• make : remove make in favor of CMake (#15449)

  Show description for 37f10f9
  danbev

  authored
  37f10f9

  Copy full SHA for 37f10f9

Commits on Jun 9, 2025

• ggml-cpu : split arch-specific implementations (#13892)

  Show description for f470bc3
  

  xctan

  and

  ggerganov

  authored
  f470bc3

  Copy full SHA for f470bc3

Commits on May 7, 2025

• examples : remove infill (#13283)

  Show description for 4773d7a
  ggerganov

  authored
  4773d7a

  Copy full SHA for 4773d7a

Commits on May 5, 2025

• mtmd : rename llava directory to mtmd (#13311)

  Show description for 9b61acf
  ngxson

  authored
  9b61acf

  Copy full SHA for 9b61acf

Commits on May 2, 2025

• llama : move end-user examples to tools directory (#13249)

  Show description for 1d36b36
  

  slaren

  and

  ngxson

  authored
  1d36b36

  Copy full SHA for 1d36b36

Commits on Apr 15, 2025

• CUDA/HIP: Share the same unified memory allocation logic. (#12934)

  Show description for 84778e9
  hjc4869

  authored
  84778e9

  Copy full SHA for 84778e9

Commits on Mar 10, 2025

• musa: support new arch mp_31 and update doc (#12296)

  Show description for 2513645
  yeahdongcn

  authored
  2513645

  Copy full SHA for 2513645

Commits on Feb 22, 2025

• CUDA: app option to compile without FlashAttention (#12025)

  JohannesGaessler

  authored
  a28e0d5

  Copy full SHA for a28e0d5

Commits on Feb 21, 2025

• MUSA: support ARM64 and enable dp4a .etc (#11843)

  Show description for 0b3863f
  

  BodhiHu

  and

  Bodhi Hu

  authored
  0b3863f

  Copy full SHA for 0b3863f

Commits on Feb 18, 2025

• tool-call: refactor common chat / tool-call api (+ tests / fixes) (#11900)

  Show description for 63e489c
  ochafik

  authored
  63e489c

  Copy full SHA for 63e489c

Commits on Feb 15, 2025

• repo : update links to new url (#11886)

  Show description for 68ff663
  ggerganov

  authored
  68ff663

  Copy full SHA for 68ff663

Commits on Feb 2, 2025

• CUDA: use mma PTX instructions for FlashAttention (#11583)

  Show description for 864a0b6
  

  JohannesGaessler

  and

  slaren

  authored
  864a0b6

  Copy full SHA for 864a0b6

Commits on Jan 30, 2025

• Tool call support (generic + native for Llama, Functionary, Hermes, Mistral, Firefunction, DeepSeek) w/ lazy grammars (#9639)

  Show description for 8b576b6
  

  3 peopleauthored
  8b576b6

  Copy full SHA for 8b576b6

Commits on Jan 21, 2025

• Add Jinja template support (#11016)

  Show description for 6171c9d
  

  3 peopleauthored
  6171c9d

  Copy full SHA for 6171c9d

Commits on Dec 14, 2024

• llama : add Qwen2VL support + multimodal RoPE (#10361)

  Show description for ba1cb19
  

  HimariO

  and

  ggerganov

  authored
  ba1cb19

  Copy full SHA for ba1cb19

Commits on Dec 7, 2024

• ggml : refactor online repacking (#10446)

  Show description for 19d8762
  

  Djip007

  and

  ggerganov

  authored
  19d8762

  Copy full SHA for 19d8762

Commits on Dec 3, 2024

• server : (web ui) Various improvements, now use vite as bundler (#10599)

  Show description for 91c36c2
  ngxson

  authored
  91c36c2

  Copy full SHA for 91c36c2

Commits on Dec 2, 2024

• make : deprecate (#10514)

  Show description for 8648c52
  

  ggerganov

  and

  slaren

  authored
  8648c52

  Copy full SHA for 8648c52

Commits on Dec 1, 2024

• build: update Makefile comments for C++ version change (#10598)

  wangqin0

  authored
  43957ef

  Copy full SHA for 43957ef

Commits on Nov 29, 2024

• ggml : move AMX to the CPU backend (#10570)

  Show description for 7cc2d2c
  

  slaren

  and

  ggerganov

  authored
  7cc2d2c

  Copy full SHA for 7cc2d2c

Commits on Nov 26, 2024

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

  Show description for be0e350
  tristandruyen

  authored
  be0e350

  Copy full SHA for be0e350

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

  Show description for 249cd93
  yeahdongcn

  authored
  249cd93

  Copy full SHA for 249cd93

Commits on Nov 25, 2024

• Introduce llama-run (#10291)

  Show description for 0cc6375
  ericcurtin

  authored
  0cc6375

  Copy full SHA for 0cc6375

• ggml : add support for dynamic loading of backends (#10469)

  Show description for 5931c1f
  

  slaren

  and

  ggerganov

  authored
  5931c1f

  Copy full SHA for 5931c1f

• speculative : refactor and add a simpler example (#10362)

  Show description for d9d54e4
  ggerganov

  authored
  d9d54e4

  Copy full SHA for d9d54e4

Commits on Nov 19, 2024

• Fix missing file renames in Makefile due to changes in commit ae8de6d50a (#10413)

  avdg

  authored
  3952a22

  Copy full SHA for 3952a22

Commits on Nov 17, 2024

• metal : refactor kernel args into structs (#10238)

  Show description for cf32a9b
  ggerganov

  authored
  cf32a9b

  Copy full SHA for cf32a9b

• CUDA: remove DMMV, consolidate F16 mult mat vec (#10318)

  JohannesGaessler

  authored
  c3ea58a

  Copy full SHA for c3ea58a

• make : add ggml-opt (#0)

  Show description for a4200ca
  ggerganov

  committed
  a4200ca

  Copy full SHA for a4200ca

• tests : remove test-grad0

  ggerganov

  committed
  84274a1

  Copy full SHA for 84274a1

Commits on Nov 16, 2024

• make : auto-determine dependencies (#0)

  ggerganov

  committed
  8ee0d09

  Copy full SHA for 8ee0d09

Commits on Nov 15, 2024

• ggml : fix some build issues

  

  slaren

  authored and

  ggerganov

  committed
  883d206

  Copy full SHA for 883d206

• backend cpu: add online flow for aarch64 Q4_0 GEMV/GEMM kernels (#9921)

  Show description for 1607a5e
  

  chaxu01

  and

  slaren

  authored
  1607a5e

  Copy full SHA for 1607a5e

Commits on Nov 14, 2024

• ggml : build backends as libraries (#10256)

  Show description for ae8de6d
  

  3 peopleauthored
  ae8de6d

  Copy full SHA for ae8de6d

Commits on Nov 8, 2024

• metal : opt-in compile flag for BF16 (#10218)

  Show description for ec450d3
  ggerganov

  authored
  ec450d3

  Copy full SHA for ec450d3