Commits

Breadcrumbs

History for

onr0.10

User selector

All users

All time

Commit History

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

Commits on Jul 11, 2016

• Merge changes from github.

  Show description for 6cd8b28
  

  andrewharp

  authored and

  tensorflower-gardener

  committed
  6cd8b28

  Copy full SHA for 6cd8b28

Commits on Jun 30, 2016

• Improved the gradients for tanh and sigmoid. This improves the speed of the ptb word model from 6800 to 7800 words per second.

  Show description for 5a8faef
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  5a8faef

  Copy full SHA for 5a8faef

• Merge changes from github.

  Show description for 533d891
  tensorflower-gardener

  committed
  533d891

  Copy full SHA for 533d891

Commits on Jun 25, 2016

• Merge changes from github.

  Show description for 67324b1
  tensorflower-gardener

  committed
  67324b1

  Copy full SHA for 67324b1

Commits on Jun 24, 2016

• Use float32 and float64 instead of float and double.

  Show description for 7d82281
  tensorflower-gardener

  committed
  7d82281

  Copy full SHA for 7d82281

Commits on Jun 17, 2016

• Remove an obsolete TODO for _ReductionDims(sparse_tensor, ..).

  Show description for 694cc06
  

  Zongheng Yang

  authored and

  tensorflower-gardener

  committed
  694cc06

  Copy full SHA for 694cc06

Commits on Jun 16, 2016

• Aggregate gradients in each device first.

  Show description for 88a50ea
  tensorflower-gardener

  committed
  88a50ea

  Copy full SHA for 88a50ea

• Merge changes from github.

  Show description for 37f56fe
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  37f56fe

  Copy full SHA for 37f56fe

• Fast path for reduction ops when reducing all dimensions.

  Show description for d57d349
  

  Zongheng Yang

  authored and

  tensorflower-gardener

  committed
  d57d349

  Copy full SHA for d57d349

Commits on Jun 15, 2016

• Make sure we never try to cast to a ref type.

  Show description for a00b14f
  tensorflower-gardener

  committed
  a00b14f

  Copy full SHA for a00b14f

Commits on Jun 14, 2016

• Execute TODOs to

  Show description for a558c6e
  

  itsmeolivia

  authored and

  tensorflower-gardener

  committed
  a558c6e

  Copy full SHA for a558c6e

Commits on Jun 3, 2016

• Merge changes from github.

  Show description for 60796d7
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  60796d7

  Copy full SHA for 60796d7

Commits on Jun 2, 2016

• Update copyright for 3p/tf/python.

  Show description for 0cf9ed3
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  0cf9ed3

  Copy full SHA for 0cf9ed3

Commits on May 24, 2016

• Register a new SparseDenseCwiseAdd op and add its Python sugar.

  Show description for ac78be6
  

  concretevitamin

  authored and

  tensorflower-gardener

  committed
  ac78be6

  Copy full SHA for ac78be6

Commits on May 23, 2016

• Merge changes from github.

  Show description for 892ca4d
  

  mrry

  authored and

  tensorflower-gardener

  committed
  892ca4d

  Copy full SHA for 892ca4d

Commits on May 19, 2016

• Support negative values in the reduction_indices argument of reduce_*

  Show description for cce41d3
  

  keveman

  authored and

  tensorflower-gardener

  committed
  cce41d3

  Copy full SHA for cce41d3

Commits on May 11, 2016

• Add cholesky_solve(chol, b), batch_cholesky_solve to linalg_ops to linalg_ops.

  Show description for 152a4f5
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  152a4f5

  Copy full SHA for 152a4f5

Commits on May 10, 2016

• Allow bfloat16 inputs to SparseMatMul.

  Show description for 6b373bb
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  6b373bb

  Copy full SHA for 6b373bb

Commits on May 9, 2016

• Override "/" and "*" for "sparse <op> dense" in Python.

  Show description for 32ecd83
  

  concretevitamin

  authored and

  tensorflower-gardener

  committed
  32ecd83

  Copy full SHA for 32ecd83

Commits on Apr 21, 2016

• Add gradients for SparseReduceSum.

  Show description for 7a4e9d3
  

  concretevitamin

  authored and

  tensorflower-gardener

  committed
  7a4e9d3

  Copy full SHA for 7a4e9d3

Commits on Apr 19, 2016

• Introduce tf.sparse_reduce_sum() and a CPU kernel.

  Show description for b3b58fc
  

  concretevitamin

  authored and

  tensorflower-gardener

  committed
  b3b58fc

  Copy full SHA for b3b58fc

Commits on Apr 17, 2016

• Starting special_math_ops.py. Made for math Ops that have more dependencies,

  Show description for 8db39c9
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  8db39c9

  Copy full SHA for 8db39c9

Commits on Apr 15, 2016

• fp16-enable all the componentwise ops.

  Show description for 31fd486
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  31fd486

  Copy full SHA for 31fd486

Commits on Apr 14, 2016

• Adds op for copying a diagonal band of a matrix or batch of matrices. Special cases include the functionality of numpy.tril and numpy.triu.

  Show description for 8aaed92
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  8aaed92

  Copy full SHA for 8aaed92

• Adds a simple fast-path to tf.accumulate_n for length 1 arguments.

  Show description for f6ebed4
  

  A. Unique TensorFlower

  authored and

  tensorflower-gardener

  committed
  f6ebed4

  Copy full SHA for f6ebed4

Commits on Apr 8, 2016

• Add batch_diag and batch_diag_part ops.

  Show description for b7eeb73
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  b7eeb73

  Copy full SHA for b7eeb73

Commits on Apr 4, 2016

• Add igamma & igammac TF ops + gradients.

  Show description for ac16ae0
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  ac16ae0

  Copy full SHA for ac16ae0

Commits on Apr 2, 2016

• PyLint updates.

  Show description for e01f6bf
  

  josh11b

  authored and

  tensorflower-gardener

  committed
  e01f6bf

  Copy full SHA for e01f6bf

Commits on Mar 30, 2016

• Merge changes from github.

  Show description for 80a5a3e
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  80a5a3e

  Copy full SHA for 80a5a3e