Commits

Branch selector

 master

User selector

ifedan

Datepicker

All time

Commit History

Commits on Nov 12, 2019

• explicitly provide memory format when calling to clone() at prune.py (pytorch#29593)

  Show description for 7df854b
  

  ifedan

  authored and

  facebook-github-bot

  committed
  7df854b

  Copy full SHA for 7df854b

Commits on Nov 11, 2019

• explicitly provide memory format when calling to clone() at ProcessGroupGloo.cpp

  Show description for 65f3b98
  

  ifedan

  authored and

  facebook-github-bot

  committed
  65f3b98

  Copy full SHA for 65f3b98

Commits on Nov 8, 2019

• explicitly provide memory format when calling to clone() at SparseTensorUtils.h

  Show description for 740c9da
  

  ifedan

  authored and

  facebook-github-bot

  committed
  740c9da

  Copy full SHA for 740c9da

• explicitly provide memory format when calling to clone() at spectral_norm.py

  Show description for c69c243
  

  ifedan

  authored and

  facebook-github-bot

  committed
  c69c243

  Copy full SHA for c69c243

Commits on Nov 7, 2019

• explicitly provide memory format when calling to clone() at SparseTensor.cpp

  Show description for b008b34
  

  ifedan

  authored and

  facebook-github-bot

  committed
  b008b34

  Copy full SHA for b008b34

• explicitly provide memory format when calling to clone() at SparseTensorMath.cpp

  Show description for 09822a1
  

  ifedan

  authored and

  facebook-github-bot

  committed
  09822a1

  Copy full SHA for 09822a1

• explicitly provide memory format when calling to clone() at TensorShape.cpp

  Show description for d5d524d
  

  ifedan

  authored and

  facebook-github-bot

  committed
  d5d524d

  Copy full SHA for d5d524d

• explicitly provide memory format when calling to clone() at rprop.py

  Show description for 43d4d01
  

  ifedan

  authored and

  facebook-github-bot

  committed
  43d4d01

  Copy full SHA for 43d4d01

• explicitly provide memory format when calling to clone() at lbfgs.py

  Show description for b05e9d4
  

  ifedan

  authored and

  facebook-github-bot

  committed
  b05e9d4

  Copy full SHA for b05e9d4

• Fix the issue when NHWC Tensor has height or width larger then max cuda grid (pytorch#28931)

  Show description for 5d70b11
  

  ifedan

  authored and

  facebook-github-bot

  committed
  5d70b11

  Copy full SHA for 5d70b11

• explicitly provide memory format when calling to clone() at parameter.py

  Show description for 4926a51
  

  ifedan

  authored and

  facebook-github-bot

  committed
  4926a51

  Copy full SHA for 4926a51

• explicitly provide memory format when calling to clone() at batchnorm.py

  Show description for 9dcf519
  

  ifedan

  authored and

  facebook-github-bot

  committed
  9dcf519

  Copy full SHA for 9dcf519

• explicitly provide memory format when calling to clone() at Indexing.cpp

  Show description for 75309b4
  

  ifedan

  authored and

  facebook-github-bot

  committed
  75309b4

  Copy full SHA for 75309b4

Commits on Nov 6, 2019

• explicitly provide memory format when calling to clone() at TensorFactories.cpp

  Show description for 9ae6fd2
  

  ifedan

  authored and

  facebook-github-bot

  committed
  9ae6fd2

  Copy full SHA for 9ae6fd2

• explicitly provide memory format when calling to clone() at SortingKthValue.cu

  Show description for 50fa132
  

  ifedan

  authored and

  facebook-github-bot

  committed
  50fa132

  Copy full SHA for 50fa132

• explicitly provide memory format when calling to clone() at SpectralOps.cu

  Show description for af45801
  

  ifedan

  authored and

  facebook-github-bot

  committed
  af45801

  Copy full SHA for af45801

• explicitly provide memory format when calling to clone() at Unique.cu

  Show description for 4e53f3b
  

  ifedan

  authored and

  facebook-github-bot

  committed
  4e53f3b

  Copy full SHA for 4e53f3b

• explicitly provide memory format when calling to clone() at TensorTransformations.cpp

  Show description for 5e0cf05
  

  ifedan

  authored and

  facebook-github-bot

  committed
  5e0cf05

  Copy full SHA for 5e0cf05

• explicitly provide memory format when calling to clone() at LinearAlgebra.cpp

  Show description for 689599d
  

  ifedan

  authored and

  facebook-github-bot

  committed
  689599d

  Copy full SHA for 689599d

Commits on Nov 5, 2019

• explicitly provide memory format when calling to clone() at SobolEngineOps.cpp

  Show description for 35f8b45
  

  ifedan

  authored and

  facebook-github-bot

  committed
  35f8b45

  Copy full SHA for 35f8b45

• explicitly provide memory format when calling to clone() at Sorting.cpp

  Show description for 9232143
  

  ifedan

  authored and

  facebook-github-bot

  committed
  9232143

  Copy full SHA for 9232143

Commits on Oct 28, 2019

• provide memory format as Contiguous explicitly when calling to clone() (pytorch#28029)

  Show description for 5835ad0
  

  ifedan

  authored and

  facebook-github-bot

  committed
  5835ad0

  Copy full SHA for 5835ad0

Commits on Oct 24, 2019

• Fix compute_non_overlapping_and_dense() (pytorch#28551)

  Show description for 2020cc0
  

  ifedan

  authored and

  facebook-github-bot

  committed
  2020cc0

  Copy full SHA for 2020cc0

Commits on Oct 21, 2019

• max_pool2d cuda should have channel last optimized kernels[Performance improvement] (pytorch#24872)

  Show description for bc57967
  

  ifedan

  authored and

  facebook-github-bot

  committed
  bc57967

  Copy full SHA for bc57967

Commits on Oct 17, 2019

• cdist performance improvement for euclidean distance (pytorch#25799)

  Show description for 12dde7f
  

  ifedan

  authored and

  facebook-github-bot

  committed
  12dde7f

  Copy full SHA for 12dde7f

Commits on Oct 8, 2019

• Remove CPU_tensor_apply* from Normalization.cpp (pytorch#27327)

  Show description for ff03f9b
  

  ifedan

  authored and

  facebook-github-bot

  committed
  ff03f9b

  Copy full SHA for ff03f9b

• Remove CUDA_tensor_apply1 (pytorch#27313)

  Show description for 3f660cd
  

  ifedan

  authored and

  facebook-github-bot

  committed
  3f660cd

  Copy full SHA for 3f660cd

Commits on Oct 4, 2019

• Allow use cpu_serial_kernel with void-lambda (pytorch#27370)

  Show description for 50b3f9d
  

  ifedan

  authored and

  facebook-github-bot

  committed
  50b3f9d

  Copy full SHA for 50b3f9d

Commits on Oct 3, 2019

• Allow use cpu_serial_kernel with void-lambda (pytorch#27271)

  Show description for 6e2a8d0
  

  ifedan

  authored and

  facebook-github-bot

  committed
  6e2a8d0

  Copy full SHA for 6e2a8d0

Commits on Sep 28, 2019

• Migrate le/gt/ge/eq/ne from the TH to Aten. Added support of type promotion. (pytorch#27017)

  Show description for ee2c79d
  

  ifedan

  authored and

  facebook-github-bot

  committed
  ee2c79d

  Copy full SHA for ee2c79d

• Migrate le/gt/ge/eq/ne from the TH to Aten. Added support of type promotion. (pytorch#26981)

  Show description for 541de7e
  

  ifedan

  authored and

  facebook-github-bot

  committed
  541de7e

  Copy full SHA for 541de7e

Commits on Sep 26, 2019

• Migrate lt and lt_ from the TH to Aten (pytorch#25998)

  Show description for f99bc71
  

  ifedan

  authored and

  facebook-github-bot

  committed
  f99bc71

  Copy full SHA for f99bc71

• Migrate multinomial from the TH to Aten (CUDA) (pytorch#26481)

  Show description for 2eb5923
  

  ifedan

  authored and

  facebook-github-bot

  committed
  2eb5923

  Copy full SHA for 2eb5923

Commits on Sep 11, 2019

• compute common dtype based on inputs only (pytorch#25593)

  Show description for e69a6ba
  

  ifedan

  authored and

  facebook-github-bot

  committed
  e69a6ba

  Copy full SHA for e69a6ba

Commits on Sep 10, 2019

• Make torch checks same for both CPU and CUDA multinomial (pytorch#25595)

  Show description for bf04c2c
  

  ifedan

  authored and

  facebook-github-bot

  committed
  bf04c2c

  Copy full SHA for bf04c2c

Pagination

Previous

Next