Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 30, 2018

• Improve the GPU memory use discipline of CollectiveReduce.

  Show description for 729e39b
  tensorflower-gardener

  committed
  729e39b

  Copy full SHA for 729e39b

• Update ops-related pbtxt files.

  Show description for b7c2e78
  tensorflower-gardener

  committed
  b7c2e78

  Copy full SHA for b7c2e78

• [XLA] Enable fusing kIota for CPU & GPU

  Show description for fb65d4c
  

  majnemer

  authored and

  tensorflower-gardener

  committed
  fb65d4c

  Copy full SHA for fb65d4c

• [tf.data] Add special case for single dataset in `tf.contrib.data.sample_from_datasets()`.

  Show description for bea2cb3
  

  mrry

  authored and

  tensorflower-gardener

  committed
  bea2cb3

  Copy full SHA for bea2cb3

• Go: Update generated wrapper functions for TensorFlow ops.

  Show description for 98dd0cd
  tensorflower-gardener

  committed
  98dd0cd

  Copy full SHA for 98dd0cd

• [data-stats] Deletes deprecated `feature_stats` transformation to collect `features` and `feature-values` count statistics.

  Show description for 356433d
  

  shivaniag

  authored and

  tensorflower-gardener

  committed
  356433d

  Copy full SHA for 356433d

• Add verbose error message for capture_tpu_profile

  Show description for 3659ed9
  tensorflower-gardener

  committed
  3659ed9

  Copy full SHA for 3659ed9

• [XLA] Switch to using kIota from TF

  Show description for 7cda8c3
  

  majnemer

  authored and

  tensorflower-gardener

  committed
  7cda8c3

  Copy full SHA for 7cda8c3

• In keras distributed training, use strategy.num_towers instead of strategy._devices since the former is a public API and works for all strategies.

  Show description for e528493
  

  guptapriya

  authored and

  tensorflower-gardener

  committed
  e528493

  Copy full SHA for e528493

• Skip links to devices that do not participate in a collective instance.

  Show description for 7c0da47
  

  dubey

  authored and

  tensorflower-gardener

  committed
  7c0da47

  Copy full SHA for 7c0da47

Commits on Aug 29, 2018

• Fix docs.

  Show description for 1835903
  

  shashishekhar

  authored and

  tensorflower-gardener

  committed
  1835903

  Copy full SHA for 1835903

• Remove :c_api dependency from common_runtime/eager

  Show description for 3c55815
  

  jdduke

  authored and

  tensorflower-gardener

  committed
  3c55815

  Copy full SHA for 3c55815

• Automated rollback of commit 065f9b833ffbb3b2f03d63febb186275674ba133

  Show description for 24af42f
  

  mkuperst

  authored and

  tensorflower-gardener

  committed
  24af42f

  Copy full SHA for 24af42f

• Lookup ro.build.version.sdk property only when needed, rather than before call to main().

  Show description for 7db159e
  tensorflower-gardener

  committed
  7db159e

  Copy full SHA for 7db159e

• Use TFLite 1.10.1 CocoaPod in example apps.

  Show description for d68dabb
  

  miaout17

  authored and

  tensorflower-gardener

  committed
  d68dabb

  Copy full SHA for d68dabb

• In BaseGPUDeviceFactory::CreateDevices() move the check for no GPU devices a little earlier.

  Show description for d61350d
  tensorflower-gardener

  committed
  d61350d

  Copy full SHA for d61350d

• Pass V1 __init__.py file list when generating API. Also, check against both TENSORFLOW_API_INIT_FILES and TENSORFLOW_API_INIT_FILES_V1 in doc_srcs_test.

  Show description for 62bcfd9
  

  annarev

  authored and

  tensorflower-gardener

  committed
  62bcfd9

  Copy full SHA for 62bcfd9

• Temporarily disable QR decomposition on GPU due to a bug in the QR implementation in

  Show description for 82491c0
  tensorflower-gardener

  committed
  82491c0

  Copy full SHA for 82491c0

• [XLA] Update InstructionFusion::EffectivelyAtMostUnary for kIota

  Show description for 866b448
  

  majnemer

  authored and

  tensorflower-gardener

  committed
  866b448

  Copy full SHA for 866b448

• New XLA API to launch a program.

  Show description for 91c7cd5
  

  yunxing

  authored and

  tensorflower-gardener

  committed
  91c7cd5

  Copy full SHA for 91c7cd5

• Automated rollback of commit 150dee25d82589ca109957cc996efbd2a236e044

  Show description for 065f9b8
  

  mkuperst

  authored and

  tensorflower-gardener

  committed
  065f9b8

  Copy full SHA for 065f9b8

• Internal change.

  Show description for 52773e6
  tensorflower-gardener

  committed
  52773e6

  Copy full SHA for 52773e6

• [tf.data] Adds an optional label_key argument to `make_batch_featured_dataset()` for extracting the label from the feature dictionaries. If label_key is provided, returned dataset will be a tuple o…

  Show description for 8a0b579
  

  shivaniag

  authored and

  tensorflower-gardener

  committed
  8a0b579

  Copy full SHA for 8a0b579

• [tf.data] Whitelist the "Assert", "Print", and "Timestamp" logging ops for iterator checkpointing.

  Show description for c69f998
  

  mrry

  authored and

  tensorflower-gardener

  committed
  c69f998

  Copy full SHA for c69f998

• Add missing __init__.py file.

  Show description for b8ce1b1
  tensorflower-gardener

  committed
  b8ce1b1

  Copy full SHA for b8ce1b1

• [TF:XLA] Bump open source llvm revision to r340902

  Show description for 5f45866
  

  Sanjoy Das

  authored and

  tensorflower-gardener

  committed
  5f45866

  Copy full SHA for 5f45866

• [TF:XLA] Verify HLOs in buffer_assignment_test and hlo_instruction_test.

  Show description for 82ff2d8
  

  dimvar

  authored and

  tensorflower-gardener

  committed
  82ff2d8

  Copy full SHA for 82ff2d8

• Automated rollback of commit 97fb9a9c7c72c405a790eea698a91654d5184fed

  Show description for 4f10020
  

  jdduke

  authored and

  tensorflower-gardener

  committed
  4f10020

  Copy full SHA for 4f10020

• Fixed that in collective all reduce cross tower ops that batch all reduce is not properly used.

  Show description for 0a4353e
  

  Yuefeng Zhou

  authored and

  tensorflower-gardener

  committed
  0a4353e

  Copy full SHA for 0a4353e

• Fix a race condition in XlaLocalLaunchBase.

  Show description for a2d2ac2
  

  Sanjoy Das

  authored and

  tensorflower-gardener

  committed
  a2d2ac2

  Copy full SHA for a2d2ac2

• common_transformers should be a Python package

  Show description for 3df3d07
  tensorflower-gardener

  committed
  3df3d07

  Copy full SHA for 3df3d07

• Update unidirectional sequential RNN to support state API.

  Show description for 287e82b
  tensorflower-gardener

  committed
  287e82b

  Copy full SHA for 287e82b

• [XLA] Implement variadic reduce in the evaluator. It is currently supported only for the case where all of the inputs have the same element type.

  Show description for 150dee2
  

  mkuperst

  authored and

  tensorflower-gardener

  committed
  150dee2

  Copy full SHA for 150dee2

• Add new aggregation mode "ONLY_FIRST_TOWER" and use it for the global

  Show description for aca9336
  tensorflower-gardener

  committed
  aca9336

  Copy full SHA for aca9336

• Internal change.

  Show description for 8a22fa7
  

  jianlijianli

  authored and

  tensorflower-gardener

  committed
  8a22fa7

  Copy full SHA for 8a22fa7

Pagination

Previous

Next