Commits

Branch selector

 a60dd98

User selector

All users

Datepicker

All time

Commit History

Commits on Apr 13, 2018

• Refactor model_tpu_main.py files and move continuous eval loop into model_lib.py

  Show description for a60dd98
  pkulzc

  committed
  a60dd98

  Copy full SHA for a60dd98

• Update create_pascal_tf_record.py and create_pet_tf_record.py

  Show description for f98f000
  pkulzc

  committed
  f98f000

  Copy full SHA for f98f000

• Enforcing that all eval metric ops have valid python strings.

  Show description for 2151c44
  pkulzc

  committed
  2151c44

  Copy full SHA for 2151c44

• Set regularize_depthwise to true in mobilenet_v1_argscope.

  Show description for 0940134
  pkulzc

  committed
  0940134

  Copy full SHA for 0940134

• Refactoring of Mask-RCNN to put all mask prediction code in third stage.

  Show description for b47ca97
  pkulzc

  committed
  b47ca97

  Copy full SHA for b47ca97

• Returning eval_on_train_input_fn from create_estimator_and_inputs(), rather than using train_input_fn in EVAL mode (which will still have data augmentation).

  Show description for 227f41e
  pkulzc

  committed
  227f41e

  Copy full SHA for 227f41e

• Access TPUEstimator and CrossShardOptimizer from tf namesspace.

  Show description for 7e81000
  pkulzc

  committed
  7e81000

  Copy full SHA for 7e81000

• Fix test that was flaky due to randomness

  Show description for b0c5c3b
  pkulzc

  committed
  b0c5c3b

  Copy full SHA for b0c5c3b

• Minor fixes to match git repo.

  Show description for 2624a23
  pkulzc

  committed
  2624a23

  Copy full SHA for 2624a23

• use shape utils for assertion in feature extractor.

  Show description for e25c014
  pkulzc

  committed
  e25c014

  Copy full SHA for e25c014

• Class agnostic masks for mask_rcnn

  Show description for a4d9c3a
  pkulzc

  committed
  a4d9c3a

  Copy full SHA for a4d9c3a

• Add learning rate summary in EVAL mode in model.py

  Show description for bfd15ec
  pkulzc

  committed
  bfd15ec

  Copy full SHA for bfd15ec

• Migrating away from Experiment class, as it is now deprecated. Also, refactoring into a separate model library and binaries.

  Show description for 90cc9ba
  pkulzc

  committed
  90cc9ba

  Copy full SHA for 90cc9ba

• Add classification_loss and localiztion_loss metrics for TPU jobs.

  Show description for 8deba73
  pkulzc

  committed
  8deba73

  Copy full SHA for 8deba73

• Add an option to skip the last striding in mobilenet. The modified network has nominal output stride 16 instead of 32.

  Show description for aedfa2e
  pkulzc

  committed
  aedfa2e

  Copy full SHA for aedfa2e

• Add option to override base feature extractor hyperparams in SSD models. This would allow us to use the same set of hyperparams for the complete feature extractor (base + new layers) if desired.

  Show description for decbad8
  pkulzc

  committed
  decbad8

  Copy full SHA for decbad8

• Adding an attribute to SSD model to indicate which fields in prediction dictionary have a batch dimension. This will be useful for future video models.

  Show description for 45ecc0f
  pkulzc

  committed
  45ecc0f

  Copy full SHA for 45ecc0f

• Internal change.

  Show description for a0c3c44
  pkulzc

  committed
  a0c3c44

  Copy full SHA for a0c3c44

• Sync to latest master.

  pkulzc

  committed
  b668f59

  Copy full SHA for b668f59

Commits on Apr 12, 2018

• Resnet distribution strategies (tensorflow#3887)

  Show description for 32aa656
  

  Taylor Robie

  authored
  32aa656

  Copy full SHA for 32aa656

• Add link to the paper and the Tratz dataset (tensorflow#3745)

  

  vered1986

  authored and

  waterson

  committed
  ad7755c

  Copy full SHA for ad7755c

• Merge pull request tensorflow#3957 from franckverrot/patch-1

  Show description for f20fac8
  MarkDaoust

  authored
  f20fac8

  Copy full SHA for f20fac8

• Correct a small typo

  Show description for b87f5a0
  franckverrot

  authored
  b87f5a0

  Copy full SHA for b87f5a0

Commits on Apr 11, 2018

• Merge pull request tensorflow#3942 from asimshankar/mnist_sequential_and_estimator

  Show description for 2c74e41
  asimshankar

  authored
  2c74e41

  Copy full SHA for 2c74e41

• Fix order of variable update and moving average update

  Show description for 720d336
  iganichev

  authored
  720d336

  Copy full SHA for 720d336

• Fix order of variable update and moving average update

  Show description for f6e4dfe
  iganichev

  committed
  f6e4dfe

  Copy full SHA for f6e4dfe

• official/mnist: Address comment in PR

  asimshankar

  committed
  cfe945e

  Copy full SHA for cfe945e

Commits on Apr 10, 2018

• change reference_data.py to use tf.gfile (tensorflow#3921)

  Show description for 2661eb9
  

  Taylor Robie

  authored
  2661eb9

  Copy full SHA for 2661eb9

• official/mnist: Another formatting tweak

  asimshankar

  committed
  a763e45

  Copy full SHA for a763e45

• official/mnist: Linter fixes

  asimshankar

  committed
  d171851

  Copy full SHA for d171851

• Add copyright header for testing script. (tensorflow#3943)

  qlzh727

  authored
  c93409c

  Copy full SHA for c93409c

• Merge pull request tensorflow#3941 from qlzh727/3885-fix

  Show description for 748ed3b
  YknZhu

  authored
  748ed3b

  Copy full SHA for 748ed3b

• Merge branch 'master' into mnist_sequential_and_estimator

  asimshankar

  committed
  4e0ca75

  Copy full SHA for 4e0ca75

• official/mnist: Use tf.keras.Sequential to simplify network definition.

  asimshankar

  committed
  e48a403

  Copy full SHA for e48a403

• Fix string to bytes conversion in deeplab data input. tensorflow#3885

  Show description for d4aa1a2
  qlzh727

  committed
  d4aa1a2

  Copy full SHA for d4aa1a2

Pagination

Previous

Next