Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Sep 27, 2017

• Add float16 support to tf.nn.fused_batch_norm on the GPU.

  Show description for 759690f
  

  reedwm

  authored and

  tensorflower-gardener

  committed
  759690f

  Copy full SHA for 759690f

Commits on Sep 15, 2017

• Update the minimum requirement of espsilon for batch norm.

  Show description for 2dcdd6c
  

  Yao Zhang

  authored and

  tensorflower-gardener

  committed
  2dcdd6c

  Copy full SHA for 2dcdd6c

Commits on Aug 15, 2017

• Merge changes from github.

  Show description for 28ce1d1
  tensorflower-gardener

  committed
  28ce1d1

  Copy full SHA for 28ce1d1

Commits on Jul 5, 2017

• Fix moments calculation to never result in negative variance and avoid

  Show description for eccd162
  tensorflower-gardener

  committed
  eccd162

  Copy full SHA for eccd162

Commits on Jun 30, 2017

• Remove unused internal facility for loss functions to colocate logit computations with the weights.

  Show description for cd15cb8
  tensorflower-gardener

  committed
  cd15cb8

  Copy full SHA for cd15cb8

Commits on Jun 22, 2017

• For candidate sampling, add facility to colocate the logit computation with the sharded embeddings.

  Show description for a1e054f
  tensorflower-gardener

  committed
  a1e054f

  Copy full SHA for a1e054f

Commits on May 24, 2017

• Allow uses of over-parameterized separable_conv.

  Show description for 1822073
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  1822073

  Copy full SHA for 1822073

Commits on May 5, 2017

• Merge changes from github.

  Show description for 692fad2
  

  ringw

  authored and

  tensorflower-gardener

  committed
  692fad2

  Copy full SHA for 692fad2

Commits on Apr 22, 2017

• Merge changes from github.

  Show description for 3269423
  

  caisq

  authored and

  tensorflower-gardener

  committed
  3269423

  Copy full SHA for 3269423

Commits on Apr 10, 2017

• Support partitioned bias variable in sampled_softmax_loss.

  Show description for a996942
  tensorflower-gardener

  committed
  a996942

  Copy full SHA for a996942

Commits on Mar 22, 2017

• Fix some documentation formatting errors.

  Show description for 540a812
  

  Patrick Nguyen

  authored and

  tensorflower-gardener

  committed
  540a812

  Copy full SHA for 540a812

Commits on Mar 8, 2017

• Add initial support for NCHW format for DepthwiseConv2D (and separable_conv2d).

  Show description for ce016c8
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  ce016c8

  Copy full SHA for ce016c8

Commits on Mar 1, 2017

• Add documentation in nce_loss and sampled_softmax_loss to prevent a common bug.

  Show description for 5de3363
  tensorflower-gardener

  committed
  5de3363

  Copy full SHA for 5de3363

Commits on Feb 27, 2017

• Stop the gradients on the sampled_values of the compute_sampled_logits.

  Show description for 6e9d1c8
  tensorflower-gardener

  committed
  6e9d1c8

  Copy full SHA for 6e9d1c8

Commits on Feb 22, 2017

• Fixing the broken Candidate Sampling PDF link.

  Show description for bd7f1f4
  tensorflower-gardener

  committed
  bd7f1f4

  Copy full SHA for bd7f1f4

Commits on Feb 14, 2017

• Manual documentation fixes.

  Show description for fa3d574
  tensorflower-gardener

  committed
  fa3d574

  Copy full SHA for fa3d574

• Fix links according to new convention.

  Show description for e7f333b
  

  Patrick Nguyen

  authored and

  tensorflower-gardener

  committed
  e7f333b

  Copy full SHA for e7f333b

Commits on Feb 2, 2017

• Break PartitionedVariable into shards so that gather ops are placed on parameter server.

  Show description for ddcbe34
  tensorflower-gardener

  committed
  ddcbe34

  Copy full SHA for ddcbe34

Commits on Jan 11, 2017

• Switch tf.concat_v2 references in third_party/tensorflow to tf.concat.

  Show description for 0e226af
  tensorflower-gardener

  committed
  0e226af

  Copy full SHA for 0e226af

Commits on Jan 10, 2017

• Make moments numerically stable by default. Added tests for moments.

  Show description for fdbd02c
  

  sguada

  authored and

  tensorflower-gardener

  committed
  fdbd02c

  Copy full SHA for fdbd02c

Commits on Jan 5, 2017

• Change arg order for {softmax,sparse_softmax,sigmoid}_cross_entropy_with_logits to be (labels, predictions), and force use of named args to avoid accidents.

  Show description for 333dc32
  

  martinwicke

  authored and

  tensorflower-gardener

  committed
  333dc32

  Copy full SHA for 333dc32

Commits on Dec 20, 2016

• Rename usages of tf.mul, tf.neg, tf.sub that are used internally

  Show description for cb4acf5
  

  aselle

  authored and

  tensorflower-gardener

  committed
  cb4acf5

  Copy full SHA for cb4acf5

Commits on Dec 15, 2016

• Switch array_ops.pack/unpack to array_ops.stack/unstack. Also switch a few remaining references to tf.pack/unpack to tf.stack/unstack.

  Show description for c5dc750
  tensorflower-gardener

  committed
  c5dc750

  Copy full SHA for c5dc750

Commits on Dec 13, 2016

• Swap the order of inputs and labels of the sampled_softmax_loss.

  Show description for a467759
  

  Jianwei Xie

  authored and

  tensorflower-gardener

  committed
  a467759

  Copy full SHA for a467759

Commits on Dec 9, 2016

• Swaps order of input/label arguments for weighted_cross_entropy_with_logits

  Show description for 7aafb97
  tensorflower-gardener

  committed
  7aafb97

  Copy full SHA for 7aafb97

• Merge changes from github.

  Show description for 1cb9689
  

  andrewharp

  authored and

  tensorflower-gardener

  committed
  1cb9689

  Copy full SHA for 1cb9689

Commits on Dec 7, 2016

• Add atrous support to depthwise_conv2d.

  Show description for dd45ad2
  tensorflower-gardener

  committed
  dd45ad2

  Copy full SHA for dd45ad2

• Switch all tf.concat(concat_dim, value, name) calls in third_party/tensorflow to tf.concat_v2(value, axis, name).

  Show description for d4eb834
  tensorflower-gardener

  committed
  d4eb834

  Copy full SHA for d4eb834

• Swap the input and label arguments in nce_loss

  Show description for 761b12e
  

  Jianwei Xie

  authored and

  tensorflower-gardener

  committed
  761b12e

  Copy full SHA for 761b12e

Commits on Dec 2, 2016

• Swaps the predictions/labels arguments in log_poisson_loss.

  Show description for ee36760
  tensorflower-gardener

  committed
  ee36760

  Copy full SHA for ee36760

Commits on Nov 30, 2016

• Migrate tf summary ops to use tf.contrib.deprecated endpoints.

  Show description for c532a5b
  

  Dan Mané

  authored and

  tensorflower-gardener

  committed
  c532a5b

  Copy full SHA for c532a5b

• Deprecate tf.select since it is getting replaced by tf.where.

  Show description for fe558b0
  tensorflower-gardener

  committed
  fe558b0

  Copy full SHA for fe558b0

Commits on Nov 23, 2016

• Seal tf.nn's interface.

  Show description for af6c3f0
  

  Patrick Nguyen

  authored and

  tensorflower-gardener

  committed
  af6c3f0

  Copy full SHA for af6c3f0

Loading