Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on May 11, 2017

• Merge changes from github.

  Show description for ee112cf
  

  benoitsteiner

  authored and

  tensorflower-gardener

  committed
  ee112cf

  Copy full SHA for ee112cf

• Decode raw op should return when input is empty.

  Show description for 79c7dfa
  

  Yuefeng Zhou

  authored and

  tensorflower-gardener

  committed
  79c7dfa

  Copy full SHA for 79c7dfa

• get_attr returns dtype objects instead of raw ints

  Show description for 15f32d9
  

  Vijay Vasudevan

  authored and

  tensorflower-gardener

  committed
  15f32d9

  Copy full SHA for 15f32d9

• Fix a bug when applying stop_gradient() on the gradient of a while loop. It turned out that I really needed to have tf.gradients() to backprop correctly on the gradient loop even if we have a stop_…

  Show description for 96980d5
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  96980d5

  Copy full SHA for 96980d5

Commits on May 10, 2017

• Copy sparse_feature_cross to sparse_ops. Two ops will be exposed as sparse_ops._sparse_cross and sparse_ops._sparse_cross_hashed. This cl does not touch the kernels and ops in contrib.

  Show description for fd85d42
  tensorflower-gardener

  committed
  fd85d42

  Copy full SHA for fd85d42

• Don't eval in integer softplus test, since it breaks asan

  Show description for 96acbc0
  

  girving

  authored and

  tensorflower-gardener

  committed
  96acbc0

  Copy full SHA for 96acbc0

• Add initial read-only lookup table ops into core that uses the V2 kernels.

  Show description for b1820e0
  

  ysuematsu

  authored and

  tensorflower-gardener

  committed
  b1820e0

  Copy full SHA for b1820e0

• Avoid checking for shape consistency in split, when the dimension is unknown.

  Show description for 0a77fc7
  tensorflower-gardener

  committed
  0a77fc7

  Copy full SHA for 0a77fc7

• * Wrap cuBlas high-level solvers into CudaSolver.

  Show description for 2de7a9e
  tensorflower-gardener

  committed
  2de7a9e

  Copy full SHA for 2de7a9e

Commits on May 7, 2017

• Internal change.

  Show description for 93572de
  

  annarev

  authored and

  tensorflower-gardener

  committed
  93572de

  Copy full SHA for 93572de

Commits on May 6, 2017

• Internal change.

  Show description for be2a9e5
  

  annarev

  authored and

  tensorflower-gardener

  committed
  be2a9e5

  Copy full SHA for be2a9e5

• Fix multiple minor split_v related issues.

  Show description for 53f6845
  tensorflower-gardener

  committed
  53f6845

  Copy full SHA for 53f6845

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 May 4, 2017

• [TF] optimization: SparseTensorDenseMatMul GPU kernel rewritten in pure cuda.

  Show description for ffc5e6f
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  ffc5e6f

  Copy full SHA for ffc5e6f

• [tf distributions] Move conditional back to contrib (not slated for core); move transformed into core.

  Show description for d1001f4
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  d1001f4

  Copy full SHA for d1001f4

Commits on May 3, 2017

• sparse_ops: Preserving static shape info in sparse_reshape, sparse_reorder, sparse_add, sparse_reset_shape in the cases where all input shapes are known and do not contain implicit "-1" dimensions…

  Show description for 1457d7f
  

  langmore

  authored and

  tensorflower-gardener

  committed
  1457d7f

  Copy full SHA for 1457d7f

• Internal change.

  Show description for 965d620
  tensorflower-gardener

  committed
  965d620

  Copy full SHA for 965d620

• In the CUDA path of depthwise_conv2d, use compile-time constants if the filter size and depth_multiplier match the xception model.

  Show description for 7828637
  tensorflower-gardener

  committed
  7828637

  Copy full SHA for 7828637

Commits on May 2, 2017

• [tf distributions] Move the remaining whitelisted distributions to core.

  Show description for a16a92f
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  a16a92f

  Copy full SHA for a16a92f

Commits on May 1, 2017

• [tf distributions] Move kullback_leibler, special_math, and Normal into TF core.

  Show description for 68c514f
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  68c514f

  Copy full SHA for 68c514f

• Add should-use for commonly misused ops.

  Show description for dbdc9fc
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  dbdc9fc

  Copy full SHA for dbdc9fc

• Move bucketize op to math_ops. This exposes the method as math_op._bucketize, but does not expose it in tf public interface. The method is still exposed as tf.contrib.layers.bucketize.

  Show description for 65283e2
  tensorflower-gardener

  committed
  65283e2

  Copy full SHA for 65283e2

Commits on Apr 28, 2017

• Remove or warn about nn_ops accidentally registered to ints

  Show description for be88359
  

  girving

  authored and

  tensorflower-gardener

  committed
  be88359

  Copy full SHA for be88359

Commits on Apr 27, 2017

• Initial move of tf.contrib.distributions to core:

  Show description for 0db6371
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  0db6371

  Copy full SHA for 0db6371

Commits on Apr 26, 2017

• Add tf.log_sigmoid

  Show description for 9845d0e
  

  girving

  authored and

  tensorflower-gardener

  committed
  9845d0e

  Copy full SHA for 9845d0e

• Add eager (static) checking to assert_equal when constant values are available.

  Show description for cfbeafe
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  cfbeafe

  Copy full SHA for cfbeafe

Commits on Apr 24, 2017

• Add a test.

  Show description for ec1672f
  

  yuanbyu

  authored and

  tensorflower-gardener

  committed
  ec1672f

  Copy full SHA for ec1672f

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 21, 2017

• Switch DirectSession to use _Arg and _Retval ops for feeding and fetching.

  Show description for 858e0af
  

  mrry

  authored and

  tensorflower-gardener

  committed
  858e0af

  Copy full SHA for 858e0af

Commits on Apr 20, 2017

• Add unreduced NONE, and reduced MEAN options for losses.

  Show description for 20a5213
  tensorflower-gardener

  committed
  20a5213

  Copy full SHA for 20a5213

• Added missing support for uint16 type in the queues.

  Show description for 1872e57
  tensorflower-gardener

  committed
  1872e57

  Copy full SHA for 1872e57

Commits on Apr 19, 2017

• Fix bug in resource variable to/from proto

  Show description for 242e0fc
  

  alextp

  authored and

  tensorflower-gardener

  committed
  242e0fc

  Copy full SHA for 242e0fc

• Disable invalid buffer forwarding in GPU kernel for matrix_triangular_solve.

  Show description for b4ca33d
  tensorflower-gardener

  committed
  b4ca33d

  Copy full SHA for b4ca33d

• Speed up non-eigen codepath for SparseTensorDenseMatmul by using FastBoundsCheck instead of CHECK_LT.

  Show description for c58d0d7
  tensorflower-gardener

  committed
  c58d0d7

  Copy full SHA for c58d0d7

Commits on Apr 18, 2017

• [TF] Optimization of tf.tile: Skip allocation if multiples is all ones.

  Show description for 1c3e6c7
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  1c3e6c7

  Copy full SHA for 1c3e6c7

Pagination

Previous

Next