Commits

Branch selector

 master

User selector

fmassa

Datepicker

All time

Commit History

Commits on Feb 2, 2017

• Remove unused code from module

  

  fmassa

  authored and

  apaszke

  committed
  833b8cb

  Copy full SHA for 833b8cb

Commits on Jan 17, 2017

• Minor fixes to docs

  

  fmassa

  authored and

  soumith

  committed
  8d9f6c2

  Copy full SHA for 8d9f6c2

Commits on Jan 6, 2017

• Minor doc fixes

  

  fmassa

  authored and

  apaszke

  committed
  8369664

  Copy full SHA for 8369664

Commits on Dec 30, 2016

• Change container doc to assign child modules via attributes

  

  fmassa

  authored and

  apaszke

  committed
  e4a3aa9

  Copy full SHA for e4a3aa9

Commits on Nov 26, 2016

• Fix shapeCheck in Spatial Pooling modules

  fmassa

  committed
  a7f24cc

  Copy full SHA for a7f24cc

• Fix shapeCheck in Spatial Pooling modules

  fmassa

  committed
  08a1bc7

  Copy full SHA for 08a1bc7

Commits on Oct 23, 2016

• Fix for versions compiled without CUDA support (#155)

  Show description for b85fc35
  

  fmassa

  authored and

  apaszke

  committed
  b85fc35

  Copy full SHA for b85fc35

Commits on Oct 7, 2016

• Update LogSoftMax to work in spatial domain

  

  fmassa

  authored and

  soumith

  committed
  5db118e

  Copy full SHA for 5db118e

Commits on Sep 29, 2016

• Fix documentation for MaxUnpool2d (#68)

  

  fmassa

  authored and

  soumith

  committed
  57610a7

  Copy full SHA for 57610a7

Commits on Sep 21, 2016

• Update LogSoftMax to work in spatial domain

  

  fmassa

  authored and

  David Yamnitsky

  committed
  df77a8a

  Copy full SHA for df77a8a

Commits on May 19, 2016

• Fix baddbmm for non-contiguous results tensor

  fmassa

  committed
  92c5a38

  Copy full SHA for 92c5a38

• Fix addmm for non-contiguous results tensor

  fmassa

  committed
  a465b67

  Copy full SHA for a465b67

Commits on May 6, 2016

• Fix memory leak in addmm

  fmassa

  committed
  9def21a

  Copy full SHA for 9def21a

Commits on May 4, 2016

• Fix addmm for non-contiguous result tensor

  fmassa

  committed
  cf3884c

  Copy full SHA for cf3884c

Commits on May 2, 2016

• Remove THExpMinusApprox from SoftMax

  Show description for 06f1c8a
  fmassa

  committed
  06f1c8a

  Copy full SHA for 06f1c8a

Commits on Apr 29, 2016

• Fix CMakeLists for Intel compilers

  fmassa

  committed
  684f4e2

  Copy full SHA for 684f4e2

Commits on Mar 26, 2016

• Fix SpatialFullConvolutionMap bug #521

  fmassa

  committed
  29c91c8

  Copy full SHA for 29c91c8

Commits on Mar 8, 2016

• Move compilation flags from nn to THNN CMakeLists

  fmassa

  committed
  5450e44

  Copy full SHA for 5450e44

Commits on Mar 7, 2016

• Add missing declarations to THNN.h

  fmassa

  committed
  c07e0a6

  Copy full SHA for c07e0a6

Commits on Feb 27, 2016

• Add checks for convolution parameters

  fmassa

  committed
  708bfa9

  Copy full SHA for 708bfa9

• Add checks for convolution parameters

  fmassa

  committed
  d2e6c39

  Copy full SHA for d2e6c39

Commits on Feb 17, 2016

• Add THNN conversion for Spatial* modules

  Show description for e59dad9
  fmassa

  committed
  e59dad9

  Copy full SHA for e59dad9

• Move Spatial* to lib/THNN/generic

  fmassa

  committed
  959b634

  Copy full SHA for 959b634

Commits on Feb 11, 2016

• THNN conversion of Spatial* modules

  fmassa

  committed
  be88627

  Copy full SHA for be88627

• Moving Spatial* to lib/THCUNN

  fmassa

  committed
  6774055

  Copy full SHA for 6774055

Commits on Jan 31, 2016

• Unify C/Cuda signatures for SpatialConvolutionMM and Spatial(AdaptiveMax,Max)Pooling

  Show description for 95b39cd
  fmassa

  committed
  95b39cd

  Copy full SHA for 95b39cd

Commits on Jan 27, 2016

• Add THNN conversion of {Spatial(AdaptiveMax,Average,Max)Pooling}

  fmassa

  committed
  e62899b

  Copy full SHA for e62899b

• Move {Spatial(Adaptive,Average,Max)Pooling.c} to lib/THNN/generic

  fmassa

  committed
  ce427e3

  Copy full SHA for ce427e3

• Add functional convertion of SpatialConvolutionMM

  fmassa

  committed
  b5f91d5

  Copy full SHA for b5f91d5

• Move SpatialConvolutionMM.c -> lib/THNN/generic

  fmassa

  committed
  130ed2c

  Copy full SHA for 130ed2c

• Add THNN conversion of {Spatial(Adaptive,Average,Max)Pooling} and SpatialConvolutionMM

  fmassa

  committed
  3a22385

  Copy full SHA for 3a22385

• Move {SpatialConvolutionMM, Spatial(Adaptive,Average,Max)Pooling} to lib/TCUNN

  fmassa

  committed
  ec1d91b

  Copy full SHA for ec1d91b

Commits on Jul 31, 2015

• Fix for gemm with zero strides

  fmassa

  committed
  6e21043

  Copy full SHA for 6e21043

Commits on Jul 13, 2015

• Fix for gemm with zero strides plus add unit test for torch.mm

  fmassa

  committed
  50df65b

  Copy full SHA for 50df65b