Commits

Branch selector

 master

User selector

benoitsteiner

Datepicker

All time

Commit History

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

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

Commits on Jul 21, 2016

• Fixed the initialization of the gradients for the strided slice op on GPU

  Show description for 691085d
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  691085d

  Copy full SHA for 691085d

• Support float16 in tf.contrib.losses

  Show description for 9f592d0
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  9f592d0

  Copy full SHA for 9f592d0

Commits on Jul 20, 2016

• Fixed incorrect debug assertion

  Show description for eccab96
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  eccab96

  Copy full SHA for eccab96

Commits on Jul 19, 2016

• Improved support for fp16

  Show description for 9223380
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  9223380

  Copy full SHA for 9223380

Commits on Jul 14, 2016

• Backported some changes to the reduction code from upstream Eigen to keep the code in sync.

  Show description for dd410bc
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  dd410bc

  Copy full SHA for dd410bc

• Made sure that philox arrays are initialized properly.

  Show description for 529e297
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  529e297

  Copy full SHA for 529e297

Commits on Jul 13, 2016

• Switched to the latest version of Eigen that provides significant performance

  Show description for 10211a6
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  10211a6

  Copy full SHA for 10211a6

Commits on Jul 12, 2016

• Switched to the latest version of Eigen that provides significant performance

  Show description for 518bd57
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  518bd57

  Copy full SHA for 518bd57

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

Commits on Jun 29, 2016

• Updated the batch normalization code to make it work with fp16

  Show description for 3a48e6a
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  3a48e6a

  Copy full SHA for 3a48e6a

Commits on Jun 27, 2016

• Added the ability to restore checkpoints containing 16 bit floats.

  Show description for 8abce2b
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  8abce2b

  Copy full SHA for 8abce2b

• Updated the draw_bounding_box operation to work properly with fp16

  Show description for 79a5188
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  79a5188

  Copy full SHA for 79a5188

Commits on Jun 23, 2016

• Added an option to train cifar10 using fp16.

  Show description for bcb8f33
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  bcb8f33

  Copy full SHA for bcb8f33

• Upgraded Eigen to the latest version that provides new scan operations. This will enable the implementation of the cumsum operation in TensorFlow

  Show description for bb40163
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  bb40163

  Copy full SHA for bb40163

Commits on Jun 20, 2016

• Merge pull request tensorflow#2914 from oxtopus/fixup-mnist-url

  Show description for 2dde4e3
  benoitsteiner

  authored
  2dde4e3

  Copy full SHA for 2dde4e3

• Merge pull request tensorflow#2956 from benoitsteiner/branch_125246026

  Show description for ebfefd2
  benoitsteiner

  authored
  ebfefd2

  Copy full SHA for ebfefd2

• Merge commit for internal changes

  benoitsteiner

  committed
  410633e

  Copy full SHA for 410633e

Commits on Jun 18, 2016

• Merge pull request tensorflow#2941 from benoitsteiner/branch_125215862

  Show description for c43a32d
  benoitsteiner

  authored
  c43a32d

  Copy full SHA for c43a32d

• Merge commit for internal changes

  benoitsteiner

  committed
  7764e8f

  Copy full SHA for 7764e8f

Commits on Jun 17, 2016

• Merge pull request tensorflow#2893 from terrytangyuan/deprecate

  Show description for b327647
  benoitsteiner

  authored
  b327647

  Copy full SHA for b327647

• Merge pull request tensorflow#2918 from benoitsteiner/branch_125120308

  Show description for b8652a2
  benoitsteiner

  authored
  b8652a2

  Copy full SHA for b8652a2

• Merge commit for internal changes

  benoitsteiner

  committed
  0530e73

  Copy full SHA for 0530e73

Commits on Jun 16, 2016

• Merge commit for internal changes

  benoitsteiner

  committed
  51323d6

  Copy full SHA for 51323d6

• Tied a few loose ends:

  Show description for fe70d8f
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  fe70d8f

  Copy full SHA for fe70d8f

• Merge changes from github.

  Show description for 37f56fe
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  37f56fe

  Copy full SHA for 37f56fe

• Added support for fp16 to TensorFlow queue operations used by the inception

  Show description for 179a756
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  179a756

  Copy full SHA for 179a756

• Merge pull request tensorflow#2892 from benoitsteiner/branch_124994976

  Show description for 966505c
  benoitsteiner

  authored
  966505c

  Copy full SHA for 966505c

Commits on Jun 15, 2016

• Merge commit for internal changes

  benoitsteiner

  committed
  406db88

  Copy full SHA for 406db88

• Merge pull request tensorflow#2884 from siddharth-agrawal/sparse_tensor_sign

  Show description for 05e406b
  benoitsteiner

  authored
  05e406b

  Copy full SHA for 05e406b

• Merge pull request tensorflow#2872 from terrytangyuan/batch_norm

  Show description for f452a84
  benoitsteiner

  authored
  f452a84

  Copy full SHA for f452a84

• Fixed uninitialized variable error.

  Show description for 1077988
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  1077988

  Copy full SHA for 1077988

• Merge pull request tensorflow#2886 from benoitsteiner/branch_124956736

  Show description for 8df836f
  benoitsteiner

  authored
  8df836f

  Copy full SHA for 8df836f

• Merge commit for internal changes

  benoitsteiner

  committed
  71855d9

  Copy full SHA for 71855d9

Pagination

Previous

Next