Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 24, 2024

• add LLM-jp-3

  kojira

  committed
  1622a47

  Copy full SHA for 1622a47

• flake.lock: Update (#10470)

  Show description for cce5a90
  

  ggerganov

  and

  github-actions[bot]

  authored
  cce5a90

  Copy full SHA for cce5a90

• llama : fix op mul check with command-r-plus (#10476)

  slaren

  authored
  dc39012

  Copy full SHA for dc39012

• convert : XLMRoberta Type Vocab Size (#10458)

  Show description for 9336db4
  gabe-l-hart

  authored
  9336db4

  Copy full SHA for 9336db4

• fix gguf-py: Conversion error when multiple licenses are configured (#9807)

  Show description for 96fa2c5
  

  mmngays

  and

  mmnga

  authored
  96fa2c5

  Copy full SHA for 96fa2c5

Commits on Nov 23, 2024

• ggml : do not use ARM features not included in the build (#10457)

  slaren

  authored
  55ed008

  Copy full SHA for 55ed008

Commits on Nov 22, 2024

• ci: Update oneAPI runtime dll packaging (#10428)

  Show description for 6dfcfef
  MakeDecisionWorth

  authored
  6dfcfef

  Copy full SHA for 6dfcfef

• GitHub: ask for more info in issue templates (#10426)

  Show description for 599b3e0
  JohannesGaessler

  authored
  599b3e0

  Copy full SHA for 599b3e0

• CANN: Support Ascend310P to accelerate F32 and F16 Model (#10216)

  Show description for c18610b
  leo-pony

  authored
  c18610b

  Copy full SHA for c18610b

Commits on Nov 21, 2024

• cuda : optimize argmax (#10441)

  Show description for a5e4759
  

  slaren

  and

  JohannesGaessler

  authored
  a5e4759

  Copy full SHA for a5e4759

• llama : handle KV shift for recurrent models (#10402)

  Show description for 1bb30bf
  ggerganov

  authored
  1bb30bf

  Copy full SHA for 1bb30bf

• sync : ggml

  ggerganov

  committed
  87a533b

  Copy full SHA for 87a533b

• ggml/sched : do not skip views in pre-assignments

  

  slaren

  authored and

  ggerganov

  committed
  59b9172

  Copy full SHA for 59b9172

• ggml-opt: fix data corruption (ggml/1022)

  

  JohannesGaessler

  authored and

  ggerganov

  committed
  02e4eaf

  Copy full SHA for 02e4eaf

Commits on Nov 20, 2024

• vulkan: predicate max operation in soft_max shaders/soft_max (#10437)

  Show description for 9abe9ee
  jeffbolznv

  authored
  9abe9ee

  Copy full SHA for 9abe9ee

• cmake: add link dependencies to cmake find pkg (#10433)

  Show description for f95caa7
  bandoti

  authored
  f95caa7

  Copy full SHA for f95caa7

• llama : add .clang-format file (#10415)

  slaren

  authored
  fab5d30

  Copy full SHA for fab5d30

• vulkan: copy iq4_nl LUT into shared memory (#10409)

  jeffbolznv

  authored
  8fd4b7f

  Copy full SHA for 8fd4b7f

• vulkan: further optimize mul_mat_vec using larger loads (#10387)

  Show description for 1bacb9f
  jeffbolznv

  authored
  1bacb9f

  Copy full SHA for 1bacb9f

• update rel to 4040 (#10395)

  Show description for ad21c9e
  

  NeoZhangJianyu

  and

  arthw

  authored
  ad21c9e

  Copy full SHA for ad21c9e

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

• add cmake rvv support (#10411)

  lhpqaq

  authored
  42ae10b

  Copy full SHA for 42ae10b

• sync : ggml

  ggerganov

  committed
  9fe0fb0

  Copy full SHA for 9fe0fb0

• metal : fox offset integer overflows in im2col (ggml/1015)

  Show description for 611fabd
  

  pminev

  authored and

  ggerganov

  committed
  611fabd

  Copy full SHA for 611fabd

• metal : add `GGML_UNARY_OP_ELU` kernel (ggml/1018)

  

  PABannier

  authored and

  ggerganov

  committed
  12b0ad9

  Copy full SHA for 12b0ad9

• cmake: force MSVC compiler charset to utf-8 (#9989)

  MakeDecisionWorth

  authored
  342397d

  Copy full SHA for 342397d

• Add required ggml-base and backend libs to cmake pkg (#10407)

  bandoti

  authored
  2a11b6b

  Copy full SHA for 2a11b6b

• cuda : fix CUDA_FLAGS not being applied (#10403)

  slaren

  authored
  3ee6382

  Copy full SHA for 3ee6382

• llama : add check for KV cache shifts (#10401)

  Show description for 8e752a7
  ggerganov

  authored
  8e752a7

  Copy full SHA for 8e752a7

• llama : add OLMo November 2024 support (#10394)

  Show description for a88ad00
  2015aroras

  authored
  a88ad00

  Copy full SHA for a88ad00

• sycl : Add option to set the SYCL architecture for all targets (#10266)

  Show description for 2a1507c
  Rbiessy

  authored
  2a1507c

  Copy full SHA for 2a1507c

• vulkan: Optimize soft_max (#10301)

  Show description for b3e5859
  jeffbolznv

  authored
  b3e5859

  Copy full SHA for b3e5859

• sycl: Revert MUL_MAT_OP support changes (#10385)

  Alcpz

  authored
  557924f

  Copy full SHA for 557924f

Commits on Nov 18, 2024

• cuda : only use native when supported by cmake (#10389)

  slaren

  authored
  d3481e6

  Copy full SHA for d3481e6

• Skip searching root path for cross-compile builds (#10383)

  bandoti

  authored
  531cb1c

  Copy full SHA for 531cb1c

Pagination

Previous

Next