Commits

Branch selector

 dskhudia-patch-1

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 17, 2019

• Update and rename ccpp.yml to onPRs.yml

  dskhudia

  authored
  5fea68b

  Copy full SHA for 5fea68b

Commits on Sep 16, 2019

• A bit more refactoring

  Show description for 53f0c0d
  

  zlateski

  authored and

  facebook-github-bot

  committed
  53f0c0d

  Copy full SHA for 53f0c0d

• Small refactoring of FBGEMM GenerateKernel class

  Show description for 96f2b9d
  

  zlateski

  authored and

  facebook-github-bot

  committed
  96f2b9d

  Copy full SHA for 96f2b9d

Commits on Sep 13, 2019

• Minor changes in initialization of dilation (pytorch#126)

  Show description for 2f1477d
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  2f1477d

  Copy full SHA for 2f1477d

• add missing instantiation for float bias for gconv (pytorch#127)

  Show description for c8cac64
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  c8cac64

  Copy full SHA for c8cac64

Commits on Sep 11, 2019

• fbgemmPacked and fbgemmConv apis with float bias + tests

  Show description for ea787e8
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  ea787e8

  Copy full SHA for ea787e8

• ReQuantization with FP32 bias

  Show description for 637288b
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  637288b

  Copy full SHA for 637288b

• API changes to take unquantized bias for depthwise conv

  Show description for 4150350
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  4150350

  Copy full SHA for 4150350

• Add assert to ensure the divisor is not 0 (#25960)

  Show description for 685b185
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  685b185

  Copy full SHA for 685b185

Commits on Sep 5, 2019

• CodeCache implemented with correct initialization of Static variables (pytorch#123)

  Show description for 9f096ab
  

  zlateski

  authored and

  facebook-github-bot

  committed
  9f096ab

  Copy full SHA for 9f096ab

Commits on Sep 4, 2019

• Revert D16968373: Introduced CodeCache container to share the microkernels among different threads.

  Show description for f9078fd
  

  ezyang

  authored and

  facebook-github-bot

  committed
  f9078fd

  Copy full SHA for f9078fd

• Modifying PackAWithIm2Col to support dilated convolution and adding test cases

  Show description for 823284e
  

  Protonu Basu

  authored and

  facebook-github-bot

  committed
  823284e

  Copy full SHA for 823284e

• remove dw conv refs and use conv_ref instead (pytorch#122)

  Show description for 624d098
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  624d098

  Copy full SHA for 624d098

• Introduced CodeCache container to share the microkernels among different threads.

  Show description for ab2d527
  

  zlateski

  authored and

  facebook-github-bot

  committed
  ab2d527

  Copy full SHA for ab2d527

• Modifying reference conv2d/3d, im2col2d.3d to support dilated convolutions

  Show description for 21782ff
  

  Protonu Basu

  authored and

  facebook-github-bot

  committed
  21782ff

  Copy full SHA for 21782ff

Commits on Sep 3, 2019

• Adding Support for dilations in the conv_param_t constructor

  Show description for 3ace43b
  

  Protonu Basu

  authored and

  facebook-github-bot

  committed
  3ace43b

  Copy full SHA for 3ace43b

• disable clang formatting in a few array definitions (pytorch#121)

  Show description for e55a596
  

  jspark1105

  authored and

  facebook-github-bot

  committed
  e55a596

  Copy full SHA for e55a596

Commits on Aug 30, 2019

• Adopt Contributor Covenant

  Show description for dc76fd4
  

  zpao

  authored and

  facebook-github-bot

  committed
  dc76fd4

  Copy full SHA for dc76fd4

Commits on Aug 29, 2019

• int8 specialization for AVX2 Quantize routine (pytorch#120)

  Show description for d4bfa96
  

  James Reed

  authored and

  facebook-github-bot

  committed
  d4bfa96

  Copy full SHA for d4bfa96

Commits on Aug 20, 2019

• Per channel support in fbgemmConv (pytorch#119)

  Show description for 280fa17
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  280fa17

  Copy full SHA for 280fa17

Commits on Aug 14, 2019

• Update asmjit to version that includes a bug fix (pytorch#118)

  Show description for a6d1d3e
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  a6d1d3e

  Copy full SHA for a6d1d3e

Commits on Aug 12, 2019

• fix error message (pytorch#117)

  Show description for 1be0815
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  1be0815

  Copy full SHA for 1be0815

• Update README.md with mentioning PyTorch (pytorch#116)

  Show description for aceefe3
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  aceefe3

  Copy full SHA for aceefe3

Commits on Aug 9, 2019

• Integrate VNNI into FBGEMM master branch (pytorch#114)

  Show description for 7b15607
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  7b15607

  Copy full SHA for 7b15607

Commits on Aug 8, 2019

• Add unpack to PackedGemmMatrixFP16 (pytorch#112)

  Show description for 122135c
  

  Yinghai Lu

  authored and

  facebook-github-bot

  committed
  122135c

  Copy full SHA for 122135c

Commits on Aug 6, 2019

• Back out "[fbgemm] Integrate VNNI into FBGEMM master branch"

  Show description for cf34b9a
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  cf34b9a

  Copy full SHA for cf34b9a

• Integrate VNNI into FBGEMM master branch (pytorch#113)

  Show description for d8b3323
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  d8b3323

  Copy full SHA for d8b3323

Commits on Aug 1, 2019

• Pass blocking param pointer into packedBufferSize() in PackBMatrix.cc

  Show description for 0d5d057
  

  evhunter

  authored and

  facebook-github-bot

  committed
  0d5d057

  Copy full SHA for 0d5d057

Commits on Jul 23, 2019

• Update OSS build instructions for submodule build (pytorch#110)

  Show description for f712cb2
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  f712cb2

  Copy full SHA for f712cb2

Commits on Jul 19, 2019

• Fix fbgemm OSS failure

  Show description for 1419a6e
  

  jianyuh

  authored and

  facebook-github-bot

  committed
  1419a6e

  Copy full SHA for 1419a6e

Commits on Jul 18, 2019

• Support pointwise with unified convolution interface as well (pytorch#108)

  Show description for 09493fd
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  09493fd

  Copy full SHA for 09493fd

• Fix missing blocking params in conv im2col code path.

  Show description for 28b7332
  

  evhunter

  authored and

  facebook-github-bot

  committed
  28b7332

  Copy full SHA for 28b7332

Commits on Jul 17, 2019

• While calling fbgemmConv with packed weights, packed weights should be compliant with convolution parameters

  Show description for 6e903d5
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  6e903d5

  Copy full SHA for 6e903d5

Commits on Jul 16, 2019

• changes to remove warnings when building in opt mode

  Show description for 931b3b7
  

  Protonu Basu

  authored and

  facebook-github-bot

  committed
  931b3b7

  Copy full SHA for 931b3b7

• Add functions needed for unpacking in PackWeightsForConv (pytorch#106)

  Show description for feca34d
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  feca34d

  Copy full SHA for feca34d

Pagination

Previous

Next