Commits

Breadcrumbs

History for

onskleinfeld-patch-1

User selector

All users

All time

Commit History

Commits on Sep 22, 2016

• Adding note to doc string comment for tf.accumulate_n, so to let clients know this op is not differentiable and could use tf.add_n instead.

  Show description for c488a9f
  tensorflower-gardener

  committed
  c488a9f

  Copy full SHA for c488a9f

• Reference math_ops symbols in the documentation and remove from whitelist.

  Show description for 150f44c
  

  Patrick Nguyen

  authored and

  tensorflower-gardener

  committed
  150f44c

  Copy full SHA for 150f44c

Commits on Sep 21, 2016

• Merge changes from github.

  Show description for 999b794
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  999b794

  Copy full SHA for 999b794

Commits on Sep 15, 2016

• Consolidate fft ops: Replace fft* ops with their batch_fft* equivalents and deprecate the latter.

  Show description for 9287c2f
  tensorflower-gardener

  committed
  9287c2f

  Copy full SHA for 9287c2f

Commits on Sep 14, 2016

• Automated rollback of change 133050375

  Show description for b22f921
  tensorflower-gardener

  committed
  b22f921

  Copy full SHA for b22f921

Commits on Sep 13, 2016

• Automated rollback of change 133033018

  Show description for 805b235
  tensorflower-gardener

  committed
  805b235

  Copy full SHA for 805b235

• Allow applying Mul and Div operators to uint16 values.

  Show description for 0034ec8
  tensorflower-gardener

  committed
  0034ec8

  Copy full SHA for 0034ec8

• 1. Rename batch_matrix_* ops to matrix_*. Specifically, this applies to

  Show description for 892deb2
  tensorflower-gardener

  committed
  892deb2

  Copy full SHA for 892deb2

• Update documentation script to not omit operators

  Show description for 250b9bf
  

  aselle

  authored and

  tensorflower-gardener

  committed
  250b9bf

  Copy full SHA for 250b9bf

Commits on Sep 12, 2016

• Remove _SelectShape function that is no longer needed - missed this when

  Show description for e1e311f
  tensorflower-gardener

  committed
  e1e311f

  Copy full SHA for e1e311f

• Change Select to use C++ shape function.

  Show description for 6901a0b
  tensorflower-gardener

  committed
  6901a0b

  Copy full SHA for 6901a0b

Commits on Sep 9, 2016

• Replace most linear algebra ops with their batch versions and deprecate

  Show description for fa9c7b7
  tensorflower-gardener

  committed
  fa9c7b7

  Copy full SHA for fa9c7b7

• Switch ops in functional_ops, math_ops, state_ops, string_ops,

  Show description for 29faf36
  tensorflower-gardener

  committed
  29faf36

  Copy full SHA for 29faf36

Commits on Sep 8, 2016

• Switch several ops in control_flow_ops, math_ops, state_ops, and string_ops to

  Show description for 3501c2f
  tensorflower-gardener

  committed
  3501c2f

  Copy full SHA for 3501c2f

Commits on Sep 7, 2016

• When calling C++ shape functions from python shape functions,

  Show description for 7e538db
  tensorflower-gardener

  committed
  7e538db

  Copy full SHA for 7e538db

Commits on Sep 2, 2016

• Enable C++ shape function for math_ops.py and array_ops.cc shape functions that

  Show description for 92052d1
  tensorflower-gardener

  committed
  92052d1

  Copy full SHA for 92052d1

• Fix gradients of numerous complex ops

  Show description for 821063d
  

  girving

  authored and

  tensorflower-gardener

  committed
  821063d

  Copy full SHA for 821063d

Commits on Aug 31, 2016

• Automated rollback of change 131739513

  Show description for e11b997
  tensorflower-gardener

  committed
  e11b997

  Copy full SHA for e11b997

• I'm requesting a documentation revision for `range`.

  Show description for 5aee733
  tensorflower-gardener

  committed
  5aee733

  Copy full SHA for 5aee733

Commits on Aug 30, 2016

• Enable C++ shape function for math_ops.py and array_ops.cc shape functions that

  Show description for a7fc9f5
  tensorflower-gardener

  committed
  a7fc9f5

  Copy full SHA for a7fc9f5

Commits on Aug 27, 2016

• Optimized the gradients of the sqrt, rsqrt, and inv functions

  Show description for f2f582b
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  f2f582b

  Copy full SHA for f2f582b

Commits on Aug 26, 2016

• Add betainc (regularized incomplete beta integral) TF op.

  Show description for af28f96
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  af28f96

  Copy full SHA for af28f96

Commits on Aug 25, 2016

• Adding reduce_logsumexp to core math_ops.py.

  Show description for e1a89ef
  tensorflower-gardener

  committed
  e1a89ef

  Copy full SHA for e1a89ef

Commits on Aug 23, 2016

• Allow a python shape inference fn to delegate to the cpp shape

  Show description for 011402d
  tensorflower-gardener

  committed
  011402d

  Copy full SHA for 011402d

Commits on Aug 10, 2016

• TensorFlow: change shape function for reduction ops, validating

  Show description for 92a05c4
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  92a05c4

  Copy full SHA for 92a05c4

• Move _BroadcastShape to a separate function in tensor_shape.

  Show description for fddaa6b
  tensorflower-gardener

  committed
  fddaa6b

  Copy full SHA for fddaa6b

Commits on Aug 9, 2016

• Deprecate op_scope by extending name_scope and move all callers to the new name_scope.

  Show description for 1df3fb0
  

  Lukasz Kaiser

  authored and

  tensorflower-gardener

  committed
  1df3fb0

  Copy full SHA for 1df3fb0

Commits on Aug 8, 2016

• Change accumulate_n() to no longer require the output shape to be known in advance, and to no longer blocked until the first input is available.

  Show description for 8e58388
  tensorflower-gardener

  committed
  8e58388

  Copy full SHA for 8e58388

• Extract shape information and encode it as constants. This improves the step time of the ptb_word_lm model by 5 to 10%, and the training speed of the inception model by 5%.

  Show description for 8f9a91a
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  8f9a91a

  Copy full SHA for 8f9a91a

Commits on Aug 5, 2016

• Changes {batch_}self_adjoint_eig ops to return eigenvalues and eigenvectors as separate outputs. This is implemented as a new set of kernels SelfAdjointEigV2 and BatchSelfAdjointEigV2. {batch_}self…

  Show description for bea8097
  tensorflower-gardener

  committed
  bea8097

  Copy full SHA for bea8097

Commits on Aug 1, 2016

• Add an op for singular value decomposition (SVD) of a dense matrix or batches of dense matrices. This calls Eigen::JacobiSVD<Matrix, Eigen::HouseholderQRPreconditioner> which is known to be rather …

  Show description for c0944a3
  tensorflower-gardener

  committed
  c0944a3

  Copy full SHA for c0944a3

Commits on Jul 26, 2016

• scalar_mul: avoid conflicts between float32 and float32_ref

  Show description for 91d4044
  

  goat000

  authored and

  tensorflower-gardener

  committed
  91d4044

  Copy full SHA for 91d4044

Commits on Jul 25, 2016

• Merge changes from github.

  Show description for 21716d8
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  21716d8

  Copy full SHA for 21716d8

Commits on Jul 14, 2016

• batch_matrix_transpose added to array_ops. This transposes the last two

  Show description for d307fa8
  

  langmore

  authored and

  tensorflower-gardener

  committed
  d307fa8

  Copy full SHA for d307fa8

• Add BatchMatrixWithDiag op & gradient

  Show description for 5d516c8
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  5d516c8

  Copy full SHA for 5d516c8