Commits

Breadcrumbs

History for

onr0.8

User selector

All users

All time

Commit History

Commits on Apr 19, 2016

• Merge pull request #1902 from martinwicke/branch_119712558

  Show description for 44a6b91
  

  martinwicke

  authored and

  caisq

  committed
  44a6b91

  Copy full SHA for 44a6b91

Commits on Apr 11, 2016

• Restrict the use of OUT_OF_RANGE to what it was intended to.

  Show description for cc7f05f
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  cc7f05f

  Copy full SHA for cc7f05f

• Make the TF_CALL_*() macros drop the semicolon, instead putting that burden

  Show description for 6e8e7ca
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  6e8e7ca

  Copy full SHA for 6e8e7ca

Commits on Apr 10, 2016

• This is another step to make TensorFlow more interactive and flexible to users. It allows a tensor produced by a run call to stay "in-place" so that a future run call can use it in-place. To achiev…

  Show description for 098f930
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  098f930

  Copy full SHA for 098f930

Commits on Apr 9, 2016

• Temporary rollback of half support in TF_CALL_REAL_NUMBER_TYPES()

  Show description for e47ddeb
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  e47ddeb

  Copy full SHA for e47ddeb

• More comprehensively enforcing Tensor MaxDimensions limit.

  Show description for 4d9ec5e
  

  David G. Andersen

  authored and

  tensorflower-gardener

  committed
  4d9ec5e

  Copy full SHA for 4d9ec5e

Commits on Apr 8, 2016

• Fixing additional int64->32 implicit conversion warnings.

  Show description for ffd8892
  

  David G. Andersen

  authored and

  tensorflower-gardener

  committed
  ffd8892

  Copy full SHA for ffd8892

• Make TF_CALL_REAL_NUMBER_TYPES() and related macros include Eigen::half

  Show description for 86dfcdc
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  86dfcdc

  Copy full SHA for 86dfcdc

Commits on Apr 6, 2016

• Various bugfixes to the core float16 support. (These started showing up

  Show description for ec58489
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  ec58489

  Copy full SHA for ec58489

Commits on Apr 2, 2016

• Enable constant folding in L0 optimization level.

  Show description for 0cd025b
  

  keveman

  authored and

  tensorflower-gardener

  committed
  0cd025b

  Copy full SHA for 0cd025b

Commits on Mar 31, 2016

• Add half support for the first basic ops, namely Cast and Const.

  Show description for 9b71f96
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  9b71f96

  Copy full SHA for 9b71f96

Commits on Mar 30, 2016

• Added a config to enable the use of a tracking allocator on CPU

  Show description for 7b7b7fb
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  7b7b7fb

  Copy full SHA for 7b7b7fb

Commits on Mar 29, 2016

• Fixing some implicit int64->32 downcast errors. (In most cases, by

  Show description for dff1e63
  

  David G. Andersen

  authored and

  tensorflower-gardener

  committed
  dff1e63

  Copy full SHA for dff1e63

• Add half support for the first basic ops, namely Cast and Const.

  Show description for 160ac73
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  160ac73

  Copy full SHA for 160ac73

• Add half support for the first basic ops, namely Cast and Const.

  Show description for b6d66ff
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  b6d66ff

  Copy full SHA for b6d66ff

Commits on Mar 28, 2016

• Change selective registration of op kernels to work off of the class name

  Show description for e30d80d
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  e30d80d

  Copy full SHA for e30d80d

Commits on Mar 25, 2016

• Mention OpKernel thread-safety requirements.

  Show description for c9b94f0
  

  rryan

  authored and

  tensorflower-gardener

  committed
  c9b94f0

  Copy full SHA for c9b94f0

• Changes to reduce code size:

  Show description for f5a9af1
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  f5a9af1

  Copy full SHA for f5a9af1

Commits on Mar 24, 2016

• Several changes to reduce TensorFlow code size (especially important

  Show description for 821920c
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  821920c

  Copy full SHA for 821920c

Commits on Mar 23, 2016

• Allows explicitly setting a gradient function for a user defined function.

  Show description for f10637b
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  f10637b

  Copy full SHA for f10637b

• Fix a case where DT_HALF would not be part of the right type lists

  Show description for d79bc31
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  d79bc31

  Copy full SHA for d79bc31

Commits on Mar 22, 2016

• Add a half type to TensorFlow core, based on Eigen::half.

  Show description for 9d03824
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  9d03824

  Copy full SHA for 9d03824

• Adding more types for __ANDROID_TYPES_FULL__ profile.

  Show description for 59e04a0
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  59e04a0

  Copy full SHA for 59e04a0

• Apply selective registration to op registration, to remove the OpDefBuilder

  Show description for 03993ff
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  03993ff

  Copy full SHA for 03993ff

Commits on Mar 18, 2016

• Change Cuda pinned memory allocation to BFC allocator.

  Show description for 2b7ca5c
  

  zheng-xq

  authored and

  tensorflower-gardener

  committed
  2b7ca5c

  Copy full SHA for 2b7ca5c

• Extend allocator_test to ensure that AllocatorAttributes values are all stored

  Show description for 6f19265
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  6f19265

  Copy full SHA for 6f19265

Commits on Mar 16, 2016

• Rollforward of "Merge changes from github."

  Show description for 4c85a08
  

  girving

  authored and

  tensorflower-gardener

  committed
  4c85a08

  Copy full SHA for 4c85a08

• Rollback of: "Merge changes from github."

  Show description for 9a4878c
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  9a4878c

  Copy full SHA for 9a4878c

• Merge changes from github.

  Show description for 3ae663c
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  3ae663c

  Copy full SHA for 3ae663c

• Avoid race conditions in TensorShape and PartialTensorShape

  Show description for 127d17a
  

  girving

  authored and

  tensorflower-gardener

  committed
  127d17a

  Copy full SHA for 127d17a

Commits on Mar 15, 2016

• Remove race conditions from TensorShapeUtils::MakeShape

  Show description for 725e968
  

  girving

  authored and

  tensorflower-gardener

  committed
  725e968

  Copy full SHA for 725e968

• When selective registration is used, only register gradient functions if there

  Show description for a6e174f
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  a6e174f

  Copy full SHA for a6e174f

Commits on Mar 14, 2016

• Remove race conditions from TensorShapeUtils::MakeShape

  Show description for ca250b2
  

  girving

  authored and

  tensorflower-gardener

  committed
  ca250b2

  Copy full SHA for ca250b2

• Remove race conditions from TensorShapeUtils::MakeShape

  Show description for 80344f3
  

  girving

  authored and

  tensorflower-gardener

  committed
  80344f3

  Copy full SHA for 80344f3

Commits on Mar 12, 2016

• Change safe_strto32 and safe_strto64 to accept StringPiece. Updates callers to

  Show description for 3b55e1f
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  3b55e1f

  Copy full SHA for 3b55e1f