Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 1, 2019

• Densenet Cleanup

  Show description for abd5369
  

  yashk2810

  authored and

  copybara-github

  committed
  abd5369

  Copy full SHA for abd5369

• internal

  Show description for 8b97abc
  

  MarkDaoust

  authored and

  copybara-github

  committed
  8b97abc

  Copy full SHA for 8b97abc

• Making sure tf.image.resize is run in tests.

  Show description for 267b143
  

  yashk2810

  authored and

  copybara-github

  committed
  267b143

  Copy full SHA for 267b143

• Resize Image API was changed on 2/28 and this is a fix for that so that the benchmarks don't fail.

  Show description for 47b8498
  

  yashk2810

  authored and

  copybara-github

  committed
  47b8498

  Copy full SHA for 47b8498

Commits on Feb 28, 2019

• Moving _nmt.ipynb to tensorflow_examples/community

  Show description for d4cfaee
  

  yashk2810

  authored and

  copybara-github

  committed
  d4cfaee

  Copy full SHA for d4cfaee

Commits on Feb 27, 2019

• Proportionally increase batch_size so that we get maximum performance out of the GPU's and scale the learning rate too.

  Show description for da48328
  

  yashk2810

  authored and

  copybara-github

  committed
  da48328

  Copy full SHA for da48328

• Add four TensorFlow Lite samples for Android and iOS.

  Show description for 29f3cb7
  

  tensorflower-gardener

  authored and

  copybara-github

  committed
  29f3cb7

  Copy full SHA for 29f3cb7

Commits on Feb 26, 2019

• Passing iterator as default instead of lambda

  Show description for e9c7a56
  

  yashk2810

  authored and

  copybara-github

  committed
  e9c7a56

  Copy full SHA for e9c7a56

• Densenet with Dist strat

  Show description for 35cb674
  

  yashk2810

  authored and

  copybara-github

  committed
  35cb674

  Copy full SHA for 35cb674

• Merge pull request tensorflow#11 from yashk2810:master

  Show description for 4a30d79
  copybara-github

  committed
  4a30d79

  Copy full SHA for 4a30d79

Commits on Feb 25, 2019

• Update README.md

  yashk2810

  authored
  51834b6

  Copy full SHA for 51834b6

• Renaming to models

  Show description for cf642c3
  

  yashk2810

  authored and

  copybara-github

  committed
  cf642c3

  Copy full SHA for cf642c3

• Add deprecation warning

  Show description for 0039492
  

  MarkDaoust

  authored and

  copybara-github

  committed
  0039492

  Copy full SHA for 0039492

• .result() returns a tensor where .numpy() will return a value.

  Show description for 8be2848
  

  yashk2810

  authored and

  lamberta

  committed
  8be2848

  Copy full SHA for 8be2848

Commits on Feb 24, 2019

• Removing dropout because data augmentation is set

  Show description for 30db916
  

  yashk2810

  authored and

  lamberta

  committed
  30db916

  Copy full SHA for 30db916

Commits on Feb 23, 2019

• prefetch to improve performance

  Show description for 0a89b65
  

  yashk2810

  authored and

  lamberta

  committed
  0a89b65

  Copy full SHA for 0a89b65

• channels_first

  Show description for c2ab486
  

  yashk2810

  authored and

  lamberta

  committed
  c2ab486

  Copy full SHA for c2ab486

Commits on Feb 22, 2019

• Shifting to resize_image_with_crop_or_pad. Also shifting to channels_last to test if V100 is faster with that.

  Show description for 10fd5a4
  

  yashk2810

  authored and

  lamberta

  committed
  10fd5a4

  Copy full SHA for 10fd5a4

• Adding more data augmentation and using SGD optimizer with nesterov momentum.

  Show description for 6419595
  

  yashk2810

  authored and

  lamberta

  committed
  6419595

  Copy full SHA for 6419595

• Removing some transforms so that NOOP warnings are not shown.

  Show description for 5c21d17
  

  yashk2810

  authored and

  lamberta

  committed
  5c21d17

  Copy full SHA for 5c21d17

• Adding kwargs so that the benchmark framework won't return errors when extras is empty.

  Show description for b906e49
  

  yashk2810

  authored and

  lamberta

  committed
  b906e49

  Copy full SHA for b906e49

• switching to channels_first

  Show description for c2dc997
  

  yashk2810

  authored and

  lamberta

  committed
  c2dc997

  Copy full SHA for c2dc997

• Adding data augmentation

  Show description for 4d1e457
  

  yashk2810

  authored and

  lamberta

  committed
  4d1e457

  Copy full SHA for 4d1e457

Commits on Feb 21, 2019

• Model running for 300 epochs with depth = 100 and k = 12. Also adding the regularization losses to custom training loops.

  Show description for 909357d
  

  yashk2810

  authored and

  MarkDaoust

  committed
  909357d

  Copy full SHA for 909357d

• Fixing logging

  Show description for 609ef62
  

  yashk2810

  authored and

  MarkDaoust

  committed
  609ef62

  Copy full SHA for 609ef62

• Changing the metric name to match compile.

  Show description for 79b4498
  

  yashk2810

  authored and

  MarkDaoust

  committed
  79b4498

  Copy full SHA for 79b4498

• Densenet benchmark

  Show description for f8994bb
  

  yashk2810

  authored and

  MarkDaoust

  committed
  f8994bb

  Copy full SHA for f8994bb

• Internal change

  Show description for 1140843
  

  yashk2810

  authored and

  MarkDaoust

  committed
  1140843

  Copy full SHA for 1140843

Commits on Feb 20, 2019

• Adding a warmup epoch to get accurate benchmarks

  Show description for 8a9e132
  

  yashk2810

  authored and

  MarkDaoust

  committed
  8a9e132

  Copy full SHA for 8a9e132

Commits on Feb 19, 2019

• Merge pull request tensorflow#6 from raskutti:master

  Show description for 13cb0ac
  MarkDaoust

  committed
  13cb0ac

  Copy full SHA for 13cb0ac

• Dcgan benchmark

  Show description for 73bcf46
  

  yashk2810

  authored and

  lamberta

  committed
  73bcf46

  Copy full SHA for 73bcf46

• Move deploy section to examples/community.

  Show description for 3a23533
  lamberta

  committed
  3a23533

  Copy full SHA for 3a23533

• Using class so that there is no need of using global train_step

  Show description for 0ce0f27
  

  yashk2810

  authored and

  lamberta

  committed
  0ce0f27

  Copy full SHA for 0ce0f27

• Update template.

  Show description for d4f9806
  

  MarkDaoust

  authored and

  lamberta

  committed
  d4f9806

  Copy full SHA for d4f9806

• Move attributions.

  MarkDaoust

  authored
  d373249

  Copy full SHA for d373249

Pagination

Previous

Next