Commits

Branch selector

 master

User selector

colesbury

Datepicker

All time

Commit History

Commits on Jul 3, 2017

• Handle None in modules list.

  Show description for b4414c0
  

  colesbury

  authored and

  soumith

  committed
  b4414c0

  Copy full SHA for b4414c0

Commits on Jun 30, 2017

• Use torch.matmul in nn.Linear (pytorch#1935)

  Show description for da0fad8
  

  colesbury

  authored and

  soumith

  committed
  da0fad8

  Copy full SHA for da0fad8

• Add weight normalization implementation (pytorch#1945)

  Show description for 2c038f2
  

  colesbury

  authored and

  soumith

  committed
  2c038f2

  Copy full SHA for 2c038f2

Commits on Jun 29, 2017

• Add ignore_index to nnl_loss and cross_entropy (pytorch#1937)

  

  colesbury

  authored and

  soumith

  committed
  4d5075a

  Copy full SHA for 4d5075a

• Improve error message when accessing attributes that don't exist (pytorch#1936)

  Show description for 0a95613
  colesbury

  authored
  0a95613

  Copy full SHA for 0a95613

Commits on Jun 28, 2017

• Speed up torch.matmul for 3D+ x 2D/1D tensors (pytorch#1931)

  Show description for 8a4eb50
  colesbury

  authored
  8a4eb50

  Copy full SHA for 8a4eb50

Commits on Jun 26, 2017

• Fix assertEquals for lists and tuples (pytorch#1913)

  Show description for 7cdd018
  colesbury

  authored
  7cdd018

  Copy full SHA for 7cdd018

Commits on Jun 22, 2017

• Add batch sampler to DataLoader (pytorch#1867)

  

  colesbury

  authored and

  apaszke

  committed
  f09027b

  Copy full SHA for f09027b

Commits on Jun 14, 2017

• Fix errors and warnings when building docs (pytorch#1806)

  

  colesbury

  authored and

  soumith

  committed
  9c53c6d

  Copy full SHA for 9c53c6d

• Added GLU (gated linear unit)

  Show description for 38b9598
  

  colesbury

  authored and

  soumith

  committed
  38b9598

  Copy full SHA for 38b9598

• Added GLU (gated linear unit)

  Show description for 244af06
  

  colesbury

  authored and

  soumith

  committed
  244af06

  Copy full SHA for 244af06

• Added GLU (gated linear unit)

  Show description for 1cf105d
  

  colesbury

  authored and

  soumith

  committed
  1cf105d

  Copy full SHA for 1cf105d

Commits on Jun 7, 2017

• Make sure the number of MKL and OpenMP threads match

  Show description for d81da41
  

  colesbury

  authored and

  soumith

  committed
  d81da41

  Copy full SHA for d81da41

• Make sure the number of MKL and OpenMP threads match

  Show description for 62835fc
  

  colesbury

  authored and

  soumith

  committed
  62835fc

  Copy full SHA for 62835fc

Commits on Jun 5, 2017

• Fix sharing of CUDA tensors on non-current devices

  Show description for 85a95d8
  

  colesbury

  authored and

  soumith

  committed
  85a95d8

  Copy full SHA for 85a95d8

• Fix sharing of CUDA tensors on non-current devices

  

  colesbury

  authored and

  soumith

  committed
  6422ea3

  Copy full SHA for 6422ea3

Commits on Jun 3, 2017

• Fix gc_refs assertion failure (pytorch#1705)

  Show description for eba3dc8
  

  colesbury

  authored and

  soumith

  committed
  eba3dc8

  Copy full SHA for eba3dc8

Commits on Jun 2, 2017

• Fix compilation with cuDNN 5 (pytorch#1703)

  colesbury

  authored
  ea05d6a

  Copy full SHA for ea05d6a

Commits on May 26, 2017

• Fix decodeUInt64BE

  Show description for 45f665d
  colesbury

  committed
  45f665d

  Copy full SHA for 45f665d

Commits on May 25, 2017

• Check for released variables in SavedVariable::unpack() (pytorch#1648)

  Show description for 036c3f9
  

  colesbury

  authored and

  soumith

  committed
  036c3f9

  Copy full SHA for 036c3f9

• Fix conv1d segfault when weight doesn't require grad (pytorch#1646)

  Show description for 98581b9
  

  colesbury

  authored and

  soumith

  committed
  98581b9

  Copy full SHA for 98581b9

Commits on May 24, 2017

• Fix segfault in autograd: (pytorch#1644)

  Show description for e1d257b
  

  colesbury

  authored and

  soumith

  committed
  e1d257b

  Copy full SHA for e1d257b

Commits on May 19, 2017

• Fix error message in CUDA forked subprocess (pytorch#1585)

  Show description for 7f6cd7c
  colesbury

  authored
  7f6cd7c

  Copy full SHA for 7f6cd7c

• Check cuDNN version at runtime (pytorch#1586)

  Show description for 625850c
  

  colesbury

  authored and

  soumith

  committed
  625850c

  Copy full SHA for 625850c

• Check for required non-None arguments in C++ autograd functions (pytorch#1589)

  

  colesbury

  authored and

  soumith

  committed
  9b34477

  Copy full SHA for 9b34477

• Fix call to Tensor.set_ in rnn.py (pytorch#1592)

  colesbury

  authored
  e6c9509

  Copy full SHA for e6c9509

Commits on May 10, 2017

• Make torch.cat not synchronize the host and device

  

  colesbury

  authored and

  soumith

  committed
  d5e8210

  Copy full SHA for d5e8210

Commits on Apr 28, 2017

• Refactor Python string utility function

  

  colesbury

  authored and

  apaszke

  committed
  4c1cdb6

  Copy full SHA for 4c1cdb6

Commits on Apr 27, 2017

• Use fully qualified name as tp_name for tensors and storages (pytorch#1379)

  colesbury

  authored
  27990fe

  Copy full SHA for 27990fe

Commits on Apr 26, 2017

• Check argument types in 'checkTypes' (pytorch#1363)

  Show description for 8ca7bf2
  colesbury

  authored
  8ca7bf2

  Copy full SHA for 8ca7bf2

Commits on Apr 24, 2017

• Fix torch.stack() with Variable inputs (pytorch#1345)

  colesbury

  authored
  3ab074b

  Copy full SHA for 3ab074b

Commits on Apr 23, 2017

• Fix errors when no CUDA devices are available (pytorch#1334)

  Show description for aab30d4
  

  colesbury

  authored and

  soumith

  committed
  aab30d4

  Copy full SHA for aab30d4

Commits on Apr 22, 2017

• Concatenate directly into shared memory when constructing batches (pytorch#1323)

  Show description for 24d92b5
  

  colesbury

  authored and

  soumith

  committed
  24d92b5

  Copy full SHA for 24d92b5

Commits on Apr 21, 2017

• Document torchvision members

  colesbury

  committed
  1375694

  Copy full SHA for 1375694

Commits on Apr 5, 2017

• Use thrust::inclusive_scan for 1D cumsum/cumprod (pytorch#742)

  Show description for 6bd4ecd
  

  colesbury

  authored and

  soumith

  committed
  6bd4ecd

  Copy full SHA for 6bd4ecd

Pagination

Previous

Next