Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Dec 7, 2017

• Add missing gradient registration for ConjugateTranspose.

  Show description for afdb6c8
  tensorflower-gardener

  committed
  afdb6c8

  Copy full SHA for afdb6c8

Commits on Nov 11, 2017

• Changed gradient of GatherNd to use IndexedSlices when possible rather than producing a dense output.

  Show description for b138012
  

  tensorflower-gardener

  authored and

  aselle

  committed
  b138012

  Copy full SHA for b138012

Commits on Nov 6, 2017

• Fix DepthToSpaceGrad and SpaceToDepthGrad on data_format NCHW.

  Show description for e641c40
  

  Jingyue Wu

  authored and

  tensorflower-gardener

  committed
  e641c40

  Copy full SHA for e641c40

Commits on Oct 31, 2017

• Add TFE_Py_TensorShapeSlice function

  Show description for 8a09bbc
  

  iganichev

  authored and

  tensorflower-gardener

  committed
  8a09bbc

  Copy full SHA for 8a09bbc

Commits on Oct 18, 2017

• Replace as_gpu_tensor and as_cpu_tensor to gpu and cpu

  Show description for bba3957
  tensorflower-gardener

  committed
  bba3957

  Copy full SHA for bba3957

Commits on Oct 3, 2017

• Internal cleanup.

  Show description for 5123f29
  tensorflower-gardener

  committed
  5123f29

  Copy full SHA for 5123f29

Commits on Aug 29, 2017

• Fix bug with second derivatives in eager mode.

  Show description for f282bb1
  

  alextp

  authored and

  tensorflower-gardener

  committed
  f282bb1

  Copy full SHA for f282bb1

Commits on Aug 17, 2017

• ignore_existing should not be mandatory for eager colocate-with

  Show description for 436d3b8
  

  alextp

  authored and

  tensorflower-gardener

  committed
  436d3b8

  Copy full SHA for 436d3b8

Commits on Aug 16, 2017

• Merges eager and graph IndexedSlices.

  Show description for 36a693c
  

  alextp

  authored and

  tensorflower-gardener

  committed
  36a693c

  Copy full SHA for 36a693c

• Removes forced function definition for eager gradients.

  Show description for 1375866
  

  alextp

  authored and

  tensorflower-gardener

  committed
  1375866

  Copy full SHA for 1375866

Commits on Aug 4, 2017

• Added n-input, n-output identity op for CPU and GPU.

  Show description for 274a36e
  tensorflower-gardener

  committed
  274a36e

  Copy full SHA for 274a36e

Commits on Jul 11, 2017

• Add an axis parameter to tf.gather. Fixes GitHub issue #11223.

  Show description for b1f9e2c
  

  rryan

  authored and

  tensorflower-gardener

  committed
  b1f9e2c

  Copy full SHA for b1f9e2c

Commits on Jul 10, 2017

• Add a PadV2 op with support for specifying a pad value.

  Show description for a6773e9
  

  rryan

  authored and

  tensorflower-gardener

  committed
  a6773e9

  Copy full SHA for a6773e9

Commits on Jun 29, 2017

• Adds QuantizeAndDequantizeV3 op, which allows using a tensor for the number of quantization bits.

  Show description for 0a797c2
  tensorflower-gardener

  committed
  0a797c2

  Copy full SHA for 0a797c2

Commits on Jun 27, 2017

• Fix bugs in ScatterNd and add ScatterNdNonAliasingAdd.

  Show description for 63ac86a
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  63ac86a

  Copy full SHA for 63ac86a

Commits on May 19, 2017

• Merge changes from github.

  Show description for 53cb26d
  tensorflower-gardener

  committed
  53cb26d

  Copy full SHA for 53cb26d

Commits on Mar 31, 2017

• This is to address a long standing issue (probably from day 1 of TensorFlow) with gradients. When we have programs like this:

  Show description for fb56fc9
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  fb56fc9

  Copy full SHA for fb56fc9

Commits on Mar 23, 2017

• Add OnesLike op and kernel.

  Show description for 194e22d
  

  Suharsh Sivakumar

  authored and

  tensorflower-gardener

  committed
  194e22d

  Copy full SHA for 194e22d

Commits on Mar 6, 2017

• Add NoGradients for QuantizeAndDequantizeV2Op's input_min and input_max arguments.

  Show description for 5254f59
  

  Suharsh Sivakumar

  authored and

  tensorflower-gardener

  committed
  5254f59

  Copy full SHA for 5254f59

Commits on Feb 18, 2017

• Merge changes from github.

  Show description for 93a975e
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  93a975e

  Copy full SHA for 93a975e

Commits on Feb 15, 2017

• Make input_min and input_max Inputs of QuantizeAndDequantize rather than attributes.

  Show description for d3a4197
  

  Suharsh Sivakumar

  authored and

  tensorflower-gardener

  committed
  d3a4197

  Copy full SHA for d3a4197

Commits on Feb 7, 2017

• Better error message from trying to differentiate PreventGradient.

  Show description for eca81d4
  

  alextp

  authored and

  tensorflower-gardener

  committed
  eca81d4

  Copy full SHA for eca81d4

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 6, 2017

• Add support for gradient calculation for concat_v2 with negative axis to fix github

  Show description for f3d5990
  tensorflower-gardener

  committed
  f3d5990

  Copy full SHA for f3d5990

Commits on Jan 4, 2017

• Adds the following new ops:

  Show description for 6ec984e
  tensorflower-gardener

  committed
  6ec984e

  Copy full SHA for 6ec984e

Commits on Dec 22, 2016

• Allows int64 indices in gather_nd gradient.

  Show description for cec4b95
  tensorflower-gardener

  committed
  cec4b95

  Copy full SHA for cec4b95

Commits on Dec 16, 2016

• Replace array_ops.unpack with array_ops.unstack in third_party/tensorflow.

  Show description for 709ee95
  tensorflower-gardener

  committed
  709ee95

  Copy full SHA for 709ee95

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 14, 2016

• Switch leftover reference to tf.concat to tf.concat_v2 in array_grad.py.

  Show description for 2cf68e3
  tensorflower-gardener

  committed
  2cf68e3

  Copy full SHA for 2cf68e3

Commits on Dec 13, 2016

• Change implementation of concat grad to use new split. Results in up to a 5x

  Show description for 3e524a3
  tensorflower-gardener

  committed
  3e524a3

  Copy full SHA for 3e524a3

Commits on Dec 9, 2016

• Fix legacy Reverse gradient

  Show description for f3ef3cf
  

  aselle

  authored and

  tensorflower-gardener

  committed
  f3ef3cf

  Copy full SHA for f3ef3cf

• Switch tf.concat references to tf.concat_v2 in array_grad.py.

  Show description for b1ed447
  tensorflower-gardener

  committed
  b1ed447

  Copy full SHA for b1ed447

Commits on Dec 7, 2016

• Remove remaining uses of tf.reverse() and remove tf.reverse()

  Show description for 878f5fc
  

  aselle

  authored and

  tensorflower-gardener

  committed
  878f5fc

  Copy full SHA for 878f5fc

Commits on Dec 1, 2016

• Rename more argument names to be consistent with NumPy.

  Show description for 83cef3f
  

  aselle

  authored and

  tensorflower-gardener

  committed
  83cef3f

  Copy full SHA for 83cef3f

Commits on Nov 22, 2016

• Add concat_v2 to array_ops.py. Also, add gradient support for concat_v2 op.

  Show description for 95b55a9
  tensorflower-gardener

  committed
  95b55a9

  Copy full SHA for 95b55a9