Commits

Breadcrumbs

History for

onv1.60.1

User selector

All users

Datepicker

All time

Commit History

Commits on Jan 21, 2024

• Merge branch 'master' into concedo_experimental

  LostRuins

  committedJan 21, 2024
  1cb8a5e

Commits on Jan 20, 2024

• Merge branch 'master' into concedo_experimental

  LostRuins

  committedJan 20, 2024
  71e9a64

Commits on Jan 16, 2024

• ggml : introduce GGML_CALL function annotation (ggerganov#4850)

  jart

  committedJan 16, 2024
  a0b3ac8

Commits on Jan 12, 2024

• llama : ggml-backend integration (ggerganov#4766)

  

  3 peoplecommittedJan 12, 2024
  e7e4df0

Commits on Dec 7, 2023

• sync : ggml (new ops, tests, backend, etc.) (ggerganov#4359)

  

  ggerganov

  and

  slaren

  committedDec 7, 2023
  fe680e3

Commits on Nov 7, 2023

• cuda : supports running on CPU for GGML_USE_CUBLAS=ON build (ggerganov#3946)

  wsxiaoys

  committedNov 7, 2023
  46876d2

Commits on Oct 8, 2023

• sync : ggml (ggml-backend) (ggerganov#3548)

  ggerganov

  committedOct 8, 2023
  db3abcc

Commits on Sep 28, 2023

• llama : custom attention mask + parallel decoding + no context swaps (ggerganov#3228)

  

  ggerganov

  and

  slaren

  committedSep 28, 2023
  ec89379

Commits on Aug 25, 2023

• ROCm Port (#1087)

  

  8 peoplecommittedAug 25, 2023
  6bbc598

Commits on Aug 22, 2023

• ggml-cuda : use graph allocator (ggerganov#2684)

  

  slaren

  and

  JohannesGaessler

  committedAug 22, 2023
  1123f7f

Commits on Aug 18, 2023

• llama : add benchmark example (ggerganov#2626)

  

  slaren

  and

  JohannesGaessler

  committedAug 18, 2023
  097e121

Commits on Jul 31, 2023

• CUDA: mmq CLI option, fixed mmq build issues (ggerganov#2453)

  JohannesGaessler

  committedJul 31, 2023
  0728c5a

Commits on Jul 1, 2023

• Better CUDA synchronization logic (ggerganov#2057)

  JohannesGaessler

  committedJul 1, 2023
  0bc2cdf

Commits on Jun 28, 2023

• CUDA GPU acceleration for LoRAs + f16 models (ggerganov#1970)

  JohannesGaessler

  committedJun 28, 2023
  7f9753f

Commits on Jun 14, 2023

• CUDA full GPU acceleration, KV cache in VRAM (ggerganov#1827)

  JohannesGaessler

  committedJun 14, 2023
  254a7a7

Commits on Jun 12, 2023

• Leverage mmap for offloading tensors to GPU (ggerganov#1597)

  

  howard0su

  and

  JohannesGaessler

  committedJun 12, 2023
  58970a4

Commits on Jun 6, 2023

• Multi GPU support, CUDA refactor, CUDA scratch buffer (ggerganov#1703)

  JohannesGaessler

  committedJun 6, 2023
  17366df

Commits on May 20, 2023

• cuda : loading models directly into VRAM, norm calculation on GPU, broadcasting for ggml_mul (ggerganov#1483)

  

  15 peoplecommittedMay 20, 2023
  affc76e

Commits on May 13, 2023

• ggml : GPU-accelerated token generation (ggerganov#1412)

  

  JohannesGaessler

  and

  ggerganov

  committedMay 13, 2023
  905d87b

Commits on May 1, 2023

• cuBLAS: refactor and optimize f16 mat mul performance (ggerganov#1259)

  slaren

  committedMay 1, 2023
  58b367c

Commits on Apr 29, 2023

• cuBLAS: use host pinned memory and dequantize while copying (ggerganov#1207)

  slaren

  committedApr 29, 2023
  7fc50c0

Commits on Apr 28, 2023

• cuBLAS: non-contiguous tensor support (ggerganov#1215)

  SlyEcho

  committedApr 28, 2023
  b1ee8f5

• Remove Q4_3 which is no better than Q5 (ggerganov#1218)

  sw

  committedApr 28, 2023
  36d19a6

Commits on Apr 26, 2023

• ggml : add Q5_0 and Q5_1 quantization (ggerganov#1187)

  

  ggerganov

  and

  sw

  committedApr 26, 2023
  574406d

Commits on Apr 25, 2023

• ggml : add Q8_0 quantization format (rename the old one to Q8_1) (ARM NEON) (ggerganov#1179)

  ggerganov

  committedApr 25, 2023
  7a32fcb

Commits on Apr 21, 2023

• Improve cuBLAS performance by using a memory pool (#1094)

  slaren

  committedApr 21, 2023
  50cb666

Commits on Apr 20, 2023

• Add Q4_3 support to cuBLAS (#1086)

  slaren

  committedApr 20, 2023
  2005469

• Improve cuBLAS performance by dequantizing on the GPU (#1065)

  slaren

  committedApr 20, 2023
  02d6988

Loading