Commits

Branch selector

 master

User selector

XapaJIaMnu

Datepicker

All time

Commit History

Commits on Sep 11, 2023

• Fix compilation with latest clang (#1007)

  XapaJIaMnu

  authored
  53b0b0d

  Copy full SHA for 53b0b0d

Commits on Aug 17, 2023

• Add an option to not encode sentencepiece during training/decoding al… (#1003)

  Show description for 961a728
  XapaJIaMnu

  authored
  961a728

  Copy full SHA for 961a728

Commits on Nov 2, 2022

• best-deep alias broken (#968)

  Show description for 07a2ac8
  XapaJIaMnu

  authored
  07a2ac8

  Copy full SHA for 07a2ac8

Commits on Sep 2, 2022

• Fix clang 13.0.1 (#939)

  Show description for 3bd281c
  

  XapaJIaMnu

  and

  snukky

  authored
  3bd281c

  Copy full SHA for 3bd281c

Commits on Feb 15, 2022

• update the intgemm version to upstream (#924)

  Show description for 8a9580b
  

  XapaJIaMnu

  and

  snukky

  authored
  8a9580b

  Copy full SHA for 8a9580b

Commits on Dec 16, 2021

• Use apple accelerate on MacOs by default (#897)

  XapaJIaMnu

  authored
  e26e5b6

  Copy full SHA for e26e5b6

Commits on Dec 7, 2021

• Fix AVX2+ detection on Mac (#895)

  Show description for e8a1a25
  

  XapaJIaMnu

  and

  snukky

  authored
  e8a1a25

  Copy full SHA for e8a1a25

Commits on Nov 23, 2021

• Add GCC 11 support (#888)

  Show description for ab6b826
  

  XapaJIaMnu

  and

  snukky

  authored
  ab6b826

  Copy full SHA for ab6b826

Commits on Nov 22, 2021

• Task alias validation during training mode (#886)

  Show description for 1adf80b
  

  XapaJIaMnu

  and

  snukky

  authored
  1adf80b

  Copy full SHA for 1adf80b

Commits on May 4, 2021

• Enable compute86 where supported (#863)

  Show description for 379212b
  XapaJIaMnu

  authored
  379212b

  Copy full SHA for 379212b

Commits on Mar 23, 2021

• Properly copy the entire vector in the int16_t case (#845)

  Show description for ffd997e
  XapaJIaMnu

  authored
  ffd997e

  Copy full SHA for ffd997e

Commits on Mar 19, 2021

• Fix model loading on architectures where size_t is 32bits (#825)

  Show description for d780082
  

  XapaJIaMnu

  and

  snukky

  authored
  d780082

  Copy full SHA for d780082

Commits on Jan 25, 2021

• Integrate intgemm into marian (#595)

  Show description for 600f5cb
  

  5 peopleauthored
  600f5cb

  Copy full SHA for 600f5cb

Commits on Oct 26, 2020

• Fix MPI on GCC8+ (#742)

  XapaJIaMnu

  authored
  f3c99bf

  Copy full SHA for f3c99bf

Commits on Sep 9, 2020

• Compute75 for NCCL (#711)

  Show description for 9c6ad01
  XapaJIaMnu

  authored
  9c6ad01

  Copy full SHA for 9c6ad01

Commits on Sep 1, 2020

• Enable compute75 when using cuda10 (#698)

  Show description for 4d9d156
  XapaJIaMnu

  authored
  4d9d156

  Copy full SHA for 4d9d156

Commits on Aug 27, 2020

• Partial CUDA11 Support (#696)

  Show description for 96738fe
  XapaJIaMnu

  authored
  96738fe

  Copy full SHA for 96738fe

Commits on May 14, 2020

• Batched gemm (#633)

  Show description for c1facda
  XapaJIaMnu

  authored
  c1facda

  Copy full SHA for c1facda

Commits on Feb 11, 2020

• Remove unused variables (#571)

  XapaJIaMnu

  authored
  64a67d5

  Copy full SHA for 64a67d5

Commits on Dec 5, 2019

• Fix compilation on CPUs that don't support AVX and some white space i… (#561)

  Show description for 34e99da
  

  XapaJIaMnu

  authored and

  emjotde

  committed
  34e99da

  Copy full SHA for 34e99da

Commits on Jun 24, 2019

• Fix compilation on gcc 9

  XapaJIaMnu

  committed
  2dedcf8

  Copy full SHA for 2dedcf8

Commits on Jun 19, 2019

• Fix compilation on gcc 8

  XapaJIaMnu

  committed
  a6052e4

  Copy full SHA for a6052e4

Commits on Feb 24, 2019

• Add the changes to CMakeLists.txt and add also support for finding the include directory

  XapaJIaMnu

  committed
  f8943a7

  Copy full SHA for f8943a7

• A more intelligent way to detect cblas if it's not included in openblas

  XapaJIaMnu

  committed
  a2d381d

  Copy full SHA for a2d381d

• Merge branch 'master' into cblas

  XapaJIaMnu

  committed
  0aded98

  Copy full SHA for 0aded98

Commits on Jun 23, 2018

• Remove unused ThreadPool and destructor, causing a crash in mini-batch fit

  XapaJIaMnu

  committed
  d80f587

  Copy full SHA for d80f587

Commits on May 24, 2018

• Merge pull request #243 from marian-nmt/multinode-hotfix

  Show description for 335fc66
  XapaJIaMnu

  authored
  335fc66

  Copy full SHA for 335fc66

• Merge pull request #241 from marian-nmt/opt_delay_multi_node

  Show description for 4e7b81f
  XapaJIaMnu

  authored
  4e7b81f

  Copy full SHA for 4e7b81f

Commits on May 14, 2018

• Don't require thread-multiple for sync multinode. Proper finalization

  XapaJIaMnu

  committed
  e2651e6

  Copy full SHA for e2651e6

• Properly finalize MPI

  XapaJIaMnu

  committed
  309158d

  Copy full SHA for 309158d

• Cudamemcpy also needs to be smaller

  XapaJIaMnu

  committed
  34ed548

  Copy full SHA for 34ed548

• Fix communication size

  XapaJIaMnu

  committed
  2a839bf

  Copy full SHA for 2a839bf

• Removed unimplemented function

  XapaJIaMnu

  committed
  ae487e0

  Copy full SHA for ae487e0

Commits on May 13, 2018

• Implement sychronous graph group multi-node

  XapaJIaMnu

  committed
  5484c2f

  Copy full SHA for 5484c2f

Commits on May 12, 2018

• Begin implementing local optimizers. Fix device initialization

  XapaJIaMnu

  committed
  ea79da6

  Copy full SHA for ea79da6

Pagination

Previous

Next