Commits

Branch selector

 master

User selector

szagoruyko

Datepicker

All time

Commit History

Commits on Jul 25, 2017

• search for cudnn in conda

  

  szagoruyko

  authored and

  apaszke

  committed
  aa35be2

  Copy full SHA for aa35be2

Commits on Jun 26, 2017

• support more than 8 gpus (pytorch#774)

  

  szagoruyko

  authored and

  soumith

  committed
  b933423

  Copy full SHA for b933423

Commits on May 7, 2017

• Add F.normalize (pytorch#1467)

  

  szagoruyko

  authored and

  apaszke

  committed
  6d693fe

  Copy full SHA for 6d693fe

Commits on Mar 23, 2017

• export current_blas_handle

  

  szagoruyko

  authored and

  apaszke

  committed
  8dc5d2a

  Copy full SHA for 8dc5d2a

Commits on Mar 2, 2017

• boolean mode in module.train

  

  szagoruyko

  authored and

  soumith

  committed
  b5f7592

  Copy full SHA for b5f7592

Commits on Mar 1, 2017

• ConstantPad2d and F.pad (pytorch#856)

  

  szagoruyko

  authored and

  apaszke

  committed
  12efd53

  Copy full SHA for 12efd53

Commits on Feb 26, 2017

• Add finite differences gradcheck (pytorch#851)

  

  szagoruyko

  authored and

  soumith

  committed
  819d4b2

  Copy full SHA for 819d4b2

Commits on Feb 22, 2017

• contiguous view backward (pytorch#816)

  

  szagoruyko

  authored and

  soumith

  committed
  adb4cb2

  Copy full SHA for adb4cb2

Commits on Feb 13, 2017

• make networks converted with loadcaffe loadable

  

  szagoruyko

  authored and

  apaszke

  committed
  16d2c3d

  Copy full SHA for 16d2c3d

Commits on Jan 28, 2017

• Fix legacy load_lua for SpatialConvolution (pytorch#608)

  Show description for 3a704ff
  

  szagoruyko

  authored and

  apaszke

  committed
  3a704ff

  Copy full SHA for 3a704ff

Commits on Jan 25, 2017

• fallback to nn batchnorm on backward-evaluate (pytorch#589)

  

  szagoruyko

  authored and

  soumith

  committed
  2d01f38

  Copy full SHA for 2d01f38

Commits on Jan 24, 2017

• make adam have the same lr as lua torch (pytorch#576)

  

  szagoruyko

  authored and

  soumith

  committed
  2748b92

  Copy full SHA for 2748b92

Commits on Jan 18, 2017

• kl_div and docs (pytorch#429)

  

  szagoruyko

  authored and

  soumith

  committed
  9c218b4

  Copy full SHA for 9c218b4

Commits on Jan 5, 2017

• remove .zero() on grad_input conv and batch_norm

  

  szagoruyko

  authored and

  apaszke

  committed
  c976dd3

  Copy full SHA for c976dd3

Commits on Jan 1, 2017

• unsqueeze instead of view in dataloader

  

  szagoruyko

  authored and

  apaszke

  committed
  a0c614e

  Copy full SHA for a0c614e

Commits on Dec 29, 2016

• fix repr in legacy.nn.linear

  

  szagoruyko

  authored and

  soumith

  committed
  101950c

  Copy full SHA for 101950c

• fix in conv repr

  

  szagoruyko

  authored and

  soumith

  committed
  239ae94

  Copy full SHA for 239ae94

• test if modules can be printed with fixes

  

  szagoruyko

  authored and

  soumith

  committed
  55e850d

  Copy full SHA for 55e850d

• Basic functional interface (pytorch#354)

  

  szagoruyko

  authored and

  apaszke

  committed
  62af45d

  Copy full SHA for 62af45d

Commits on Dec 25, 2016

• remove unused buffer in avg pooling

  szagoruyko

  committed
  d2a93c3

  Copy full SHA for d2a93c3

• Move max pooling construction logic to functions (pytorch#343)

  

  szagoruyko

  authored and

  soumith

  committed
  bc475ca

  Copy full SHA for bc475ca

• default args for conv functions

  

  szagoruyko

  authored and

  soumith

  committed
  45d6212

  Copy full SHA for 45d6212

Commits on Dec 17, 2016

• add cudnn 5.0.5 to supported versions (pytorch#321)

  

  szagoruyko

  authored and

  soumith

  committed
  5586f48

  Copy full SHA for 5586f48

Commits on Dec 16, 2016

• ensure that legacy linear has gradWeight and gradBias fields (pytorch#319)

  

  szagoruyko

  authored and

  apaszke

  committed
  cc6e3c9

  Copy full SHA for cc6e3c9

Commits on Dec 12, 2016

• fix tests for GPU-less setup (pytorch#298)

  

  szagoruyko

  authored and

  soumith

  committed
  89d9303

  Copy full SHA for 89d9303

Commits on Dec 11, 2016

• fix gcc-6 build on os x (pytorch#297)

  

  szagoruyko

  authored and

  apaszke

  committed
  04393cd

  Copy full SHA for 04393cd

• legacy fixes (pytorch#287)

  

  szagoruyko

  authored and

  apaszke

  committed
  1031d67

  Copy full SHA for 1031d67

Commits on Jul 29, 2016

• gemm -> Sgemm

  

  szagoruyko

  authored and

  soumith

  committed
  9d1a0f3

  Copy full SHA for 9d1a0f3

Commits on Jul 17, 2016

• C-impl BCECriterion

  szagoruyko

  committed
  f09c8ab

  Copy full SHA for f09c8ab

Commits on Jun 18, 2016

• inplace HardTanh, subclass ReLU6

  szagoruyko

  committed
  f2b2ad6

  Copy full SHA for f2b2ad6

Commits on Jun 16, 2016

• inplace hardtanh, remove relu6

  szagoruyko

  committed
  d542253

  Copy full SHA for d542253

Commits on Jul 8, 2015

• check if torch is found before find

  szagoruyko

  committed
  6b1dfec

  Copy full SHA for 6b1dfec

Commits on Jun 9, 2015

• THC standalone compilation

  szagoruyko

  committed
  c3945c6

  Copy full SHA for c3945c6

Commits on Sep 24, 2014

• peer to peer access

  szagoruyko

  committed
  8cae227

  Copy full SHA for 8cae227