Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 29, 2018

• Update bidirectional RNN to support state API.

  Show description for d5d02f0
  tensorflower-gardener

  committed
  d5d02f0

  Copy full SHA for d5d02f0

• Improve the implementation of HloSharding::GetSubSharding

  Show description for 8dfb753
  tensorflower-gardener

  committed
  8dfb753

  Copy full SHA for 8dfb753

• Expose default execution mode as a public symbol in context.

  Show description for a15f91d
  

  tomhennigan

  authored and

  tensorflower-gardener

  committed
  a15f91d

  Copy full SHA for a15f91d

• compat: Update forward compatibility horizon to 2018-08-29

  Show description for 5792a67
  tensorflower-gardener

  committed
  5792a67

  Copy full SHA for 5792a67

• Add support for hybrid evaluation in quantize weights transform.

  Show description for 47dee5b
  

  Suharsh Sivakumar

  authored and

  tensorflower-gardener

  committed
  47dee5b

  Copy full SHA for 47dee5b

• Use nccl if there is only one worker in multi-worker MirroredStrategies.

  Show description for 13d6aec
  

  Yuefeng Zhou

  authored and

  tensorflower-gardener

  committed
  13d6aec

  Copy full SHA for 13d6aec

• Allow TF_CONFIG when remote_cluster is specified in distribute coordinator's utility for estimator.

  Show description for 8299351
  

  Yuefeng Zhou

  authored and

  tensorflower-gardener

  committed
  8299351

  Copy full SHA for 8299351

• [XLA] Make explicit that negative interior_padding is not allowed.

  Show description for c010fec
  

  Justin Lebar

  authored and

  tensorflower-gardener

  committed
  c010fec

  Copy full SHA for c010fec

• Automated rollback of commit 069f808e5c0462819bcd6c73c75491b00cdd42c2

  Show description for 30a6b56
  

  k-w-w

  authored and

  tensorflower-gardener

  committed
  30a6b56

  Copy full SHA for 30a6b56

• [TF:XLA] Change group_assignment from 1d array attribute to 2d array input tensor with shape [num_groups, num_replica_per_group].

  Show description for 8012cf5
  

  toponado-zz

  authored and

  tensorflower-gardener

  committed
  8012cf5

  Copy full SHA for 8012cf5

• tfdbg: Add adjustable limit to total bytes dumped to disk

  Show description for 3d35a07
  

  caisq

  authored and

  tensorflower-gardener

  committed
  3d35a07

  Copy full SHA for 3d35a07

• Fix FTRL L2-shrinkage behavior: the gradient from the L2 shrinkage term should not end up in the accumulator.

  Show description for 20d5683
  tensorflower-gardener

  committed
  20d5683

  Copy full SHA for 20d5683

• Export Keras model to SavedModel.

  Show description for 069f808
  

  k-w-w

  authored and

  tensorflower-gardener

  committed
  069f808

  Copy full SHA for 069f808

• Avoid iterating the dict keys list for membership test.

  Show description for 4a83b95
  tensorflower-gardener

  committed
  4a83b95

  Copy full SHA for 4a83b95

• Add more model support to TocoConverter.

  Show description for 2e7352e
  

  Nupur Garg

  authored and

  tensorflower-gardener

  committed
  2e7352e

  Copy full SHA for 2e7352e

• Added support for categorical features.

  Show description for c4099e6
  tensorflower-gardener

  committed
  c4099e6

  Copy full SHA for c4099e6

• Increase the size of embeddings_test from small to medium to fix timeouts.

  Show description for 6eabd59
  

  pavithrasv

  authored and

  tensorflower-gardener

  committed
  6eabd59

  Copy full SHA for 6eabd59

• [XLA] In literal_comparison, set absolute error and relative error to infinity

  Show description for e105c10
  

  kayzhu

  authored and

  tensorflower-gardener

  committed
  e105c10

  Copy full SHA for e105c10

• [XLA] Add support for algebraic simplifications involving kIota

  Show description for de25bab
  

  majnemer

  authored and

  tensorflower-gardener

  committed
  de25bab

  Copy full SHA for de25bab

• Step 1 toward automatically assigning # of TPU cores.

  Show description for 4dcd000
  

  Jianwei Xie

  authored and

  tensorflower-gardener

  committed
  4dcd000

  Copy full SHA for 4dcd000

Commits on Aug 28, 2018

• [XLA:GPU] Create backwards-input convs when the filter is a constant.

  Show description for bb0f1e9
  

  Justin Lebar

  authored and

  tensorflower-gardener

  committed
  bb0f1e9

  Copy full SHA for bb0f1e9

• Make error message more readable.

  Show description for 0721433
  

  Sanjoy Das

  authored and

  tensorflower-gardener

  committed
  0721433

  Copy full SHA for 0721433

• Merge pull request #21702 from facaiy:ENH/assert_partial_shape

  Show description for ab4ae7e
  tensorflower-gardener

  committed
  ab4ae7e

  Copy full SHA for ab4ae7e

• [XLA:GPU] Print better error messages in stream_executor_util.

  Show description for 02163c5
  

  Justin Lebar

  authored and

  tensorflower-gardener

  committed
  02163c5

  Copy full SHA for 02163c5

• Make Graph::UpdateEdge() be O(e) instead of O(E)

  Show description for 7f52de1
  

  asimshankar

  authored and

  tensorflower-gardener

  committed
  7f52de1

  Copy full SHA for 7f52de1

• [XLA] Avoid implicit layout change for non-layout-changing instructions.

  Show description for bf7b20e
  

  bixia1

  authored and

  tensorflower-gardener

  committed
  bf7b20e

  Copy full SHA for bf7b20e

• Revert the numpy restrictions. Tested on CPU/GPU pip builds.

  Show description for 6729cd7
  

  Amit Patankar

  authored and

  tensorflower-gardener

  committed
  6729cd7

  Copy full SHA for 6729cd7

• Automated rollback of commit 683e21314a80ac6cb89eb959465ded41e381d23c

  Show description for 730b530
  

  shashishekhar

  authored and

  tensorflower-gardener

  committed
  730b530

  Copy full SHA for 730b530

• Merge pull request #21616 from karllessard:java-ops-default-type-attrs

  Show description for 85631dc
  tensorflower-gardener

  committed
  85631dc

  Copy full SHA for 85631dc

• Merge pull request #21808 from NervanaSystems:avijit/fix-license-files

  Show description for af94082
  tensorflower-gardener

  committed
  af94082

  Copy full SHA for af94082

• Change adadelta_test to large as adadelta_test_cpu is frequently timing out.

  Show description for 03bbc16
  tensorflower-gardener

  committed
  03bbc16

  Copy full SHA for 03bbc16

• _autopacking_helper casts tensors when it needs to.

  Show description for 1549491
  

  Akshay Modi

  authored and

  tensorflower-gardener

  committed
  1549491

  Copy full SHA for 1549491

• Triage cond tests that fail with cond_v2 enabled, + other small test fixes.

  Show description for 33b460d
  

  skye

  authored and

  tensorflower-gardener

  committed
  33b460d

  Copy full SHA for 33b460d

• Stop overriding destination device to nullptr (CPU:0) when its on CPU

  Show description for f8d9f1f
  

  Akshay Modi

  authored and

  tensorflower-gardener

  committed
  f8d9f1f

  Copy full SHA for f8d9f1f

• Update NNAPI delegate to support state API for RNN.

  Show description for 40cd08d
  tensorflower-gardener

  committed
  40cd08d

  Copy full SHA for 40cd08d

Pagination

Previous

Next