Commits

Breadcrumbs

History for

onr0.10

User selector

All users

All time

Commit History

Commits on Aug 5, 2016

• Remove "for use_gpu" loop from image_ops_test.

  Show description for 3938bc3
  

  gunan

  authored and

  tensorflower-gardener

  committed
  3938bc3

  Copy full SHA for 3938bc3

• 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 4, 2016

• Added type information to variable scope

  Show description for 80947c0
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  80947c0

  Copy full SHA for 80947c0

• Fix sparse_merge() doc formatting.

  Show description for 230a955
  

  Zongheng Yang

  authored and

  tensorflower-gardener

  committed
  230a955

  Copy full SHA for 230a955

• Make attention_decoder in seq2seq work with state_is_tuple=True. (Fix #3056.)

  Show description for 6215671
  

  Lukasz Kaiser

  authored and

  tensorflower-gardener

  committed
  6215671

  Copy full SHA for 6215671

Commits on Aug 3, 2016

• Renames sparse_merge 'in_row_major_order' arg to 'already_sorted'

  Show description for eece529
  tensorflower-gardener

  committed
  eece529

  Copy full SHA for eece529

• Makes sparse_reorder optional in sparse_merge.

  Show description for ec73b4e
  tensorflower-gardener

  committed
  ec73b4e

  Copy full SHA for ec73b4e

• Update documentation about the requirement for sequence_length input to dynamic_rnn.

  Show description for 855d3b5
  tensorflower-gardener

  committed
  855d3b5

  Copy full SHA for 855d3b5

Commits on Aug 2, 2016

• Automated rollback of change 128393251

  Show description for 2823983
  

  aselle

  authored and

  tensorflower-gardener

  committed
  2823983

  Copy full SHA for 2823983

• Pipe infer_shape argument to tf.map_fn and tf.scan

  Show description for bc26328
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  bc26328

  Copy full SHA for bc26328

Commits on Aug 1, 2016

• Simplify tf.nn.sufficient_statistics.

  Show description for b1d9ef5
  tensorflower-gardener

  committed
  b1d9ef5

  Copy full SHA for b1d9ef5

• Don't consider control inputs in gradient backprop graph construction. I added

  Show description for 9d7c949
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  9d7c949

  Copy full SHA for 9d7c949

• 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

• Merge changes from github.

  Show description for abe9ab3
  tensorflower-gardener

  committed
  abe9ab3

  Copy full SHA for abe9ab3

Commits on Jul 29, 2016

• Use a division directly instead of multiplying by the inverse. Also use the rsqrt operation instead of taking the inverse of the square root.

  Show description for aebdf77
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  aebdf77

  Copy full SHA for aebdf77

• Fix a bug in handling while loop inside a "with tf.control_dependencies()". Also tidy up some error checking.

  Show description for f419dd1
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  f419dd1

  Copy full SHA for f419dd1

Commits on Jul 28, 2016

• Broken up changes toward enabling strided slice

  Show description for e3fdcca
  

  aselle

  authored and

  tensorflower-gardener

  committed
  e3fdcca

  Copy full SHA for e3fdcca

Commits on Jul 27, 2016

• Adding log poisson loss in order to support Poisson Regression.

  Show description for 6ab6800
  tensorflower-gardener

  committed
  6ab6800

  Copy full SHA for 6ab6800

• check_ops upgrade:

  Show description for b34cdec
  

  langmore

  authored and

  tensorflower-gardener

  committed
  b34cdec

  Copy full SHA for b34cdec

• If a tf.VariableScope is explicitly passes as an argument to

  Show description for 1f9b3f9
  tensorflower-gardener

  committed
  1f9b3f9

  Copy full SHA for 1f9b3f9

• Added a gradient for the check_numeric operation. This makes it possible to

  Show description for 3a1bf18
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  3a1bf18

  Copy full SHA for 3a1bf18

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

• Fix two striding slicing grad bugs

  Show description for 9b077c0
  

  aselle

  authored and

  tensorflower-gardener

  committed
  9b077c0

  Copy full SHA for 9b077c0

• Add custom_getter to variable scope.

  Show description for 904f76e
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  904f76e

  Copy full SHA for 904f76e

• Added 'stateful' parameter to tf.py_func and a stateless version of the op. This allows tf.py_func operations to be cached and optimized away in some cases.

  Show description for 2c5b766
  tensorflower-gardener

  committed
  2c5b766

  Copy full SHA for 2c5b766

Commits on Jul 25, 2016

• Added the ability to train the ptb_word_lm example using 16 bit floats

  Show description for 7d3f838
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  7d3f838

  Copy full SHA for 7d3f838

• Merge changes from github.

  Show description for 21716d8
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  21716d8

  Copy full SHA for 21716d8

• Automated rollback of change 128378997

  Show description for 1bd1c27
  

  aselle

  authored and

  tensorflower-gardener

  committed
  1bd1c27

  Copy full SHA for 1bd1c27

• Automated rollback of change 127492429

  Show description for d7a2571
  

  aselle

  authored and

  tensorflower-gardener

  committed
  d7a2571

  Copy full SHA for d7a2571

• Update the doc_string of tf.while_loop to make it clear that the return type differs when called with different lengths of loop_vars.

  Show description for 359d7fc
  tensorflower-gardener

  committed
  359d7fc

  Copy full SHA for 359d7fc

Commits on Jul 24, 2016

• Adds python_grad_func argument to function.Defun.

  Show description for 01925fe
  tensorflower-gardener

  committed
  01925fe

  Copy full SHA for 01925fe

Commits on Jul 23, 2016

• Extend gather_nd to support sliced indexing.

  Show description for 34883b6
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  34883b6

  Copy full SHA for 34883b6

Commits on Jul 22, 2016

• Use 32bit floats to compute cross entropies. 16 bit floats aren't accurate

  Show description for 0758686
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  0758686

  Copy full SHA for 0758686

• Fixing bug in which a Tensor was being checked as a Python bool, causing an error.

  Show description for 3acdd0a
  tensorflower-gardener

  committed
  3acdd0a

  Copy full SHA for 3acdd0a

• Translate shape inference functions for sparse_ops to C++.

  Show description for f3b6180
  tensorflower-gardener

  committed
  f3b6180

  Copy full SHA for f3b6180