Commits · 2f6c5a5ef86fc42ecec73837fd30bca6cfba7244 · ModelZoo / ResNet50_tensorflow

18 Dec, 2019 1 commit
- Refactor: Move transformer Estimator-only code to r1 folder. · 2f6c5a5e
  Hongkun Yu authored Dec 17, 2019
```
PiperOrigin-RevId: 286103069
```
  2f6c5a5e
17 Dec, 2019 2 commits
- Internal change · 7a69f962
  Chris Jones authored Dec 17, 2019
```
PiperOrigin-RevId: 285974579
```
  7a69f962
- Refactor: move transformer benchmark to benchmark folder. · 72deecc0
  Hongkun Yu authored Dec 16, 2019
```
PiperOrigin-RevId: 285905996
```
  72deecc0
16 Dec, 2019 1 commit
- Internal change · 913640d4
  Hongkun Yu authored Dec 16, 2019
```
PiperOrigin-RevId: 285765110
```
  913640d4
14 Dec, 2019 2 commits
- Internal change · 0788a23c
  Hongkun Yu authored Dec 13, 2019
```
PiperOrigin-RevId: 285533511
```
  0788a23c
- Clearly demarcate contrib symbols from standard tf symbols by importing them directly. · 357f30f4
  A. Unique TensorFlower authored Dec 13, 2019
```
PiperOrigin-RevId: 285503670
```
  357f30f4
12 Dec, 2019 2 commits
- Internal change · 0f5bdd0e
  Chen Chen authored Dec 12, 2019
```
PiperOrigin-RevId: 285285388
```
  0f5bdd0e
- Enabled performance related parameters for Transformer: all_reduce_alg, · 152adcbd
  A. Unique TensorFlower authored Dec 12, 2019
```
enable_eager, tf_gpu_thread_mode and datasets_num_private_threads

PiperOrigin-RevId: 285157129
```
  152adcbd
10 Dec, 2019 1 commit
- Internal change · 68104ce3
  A. Unique TensorFlower authored Dec 10, 2019
```
PiperOrigin-RevId: 284805626
```
  68104ce3
03 Dec, 2019 1 commit
- Internal change · 1d3cd3cf
  Hongkun Yu authored Dec 03, 2019
```
PiperOrigin-RevId: 283622345
```
  1d3cd3cf
02 Dec, 2019 1 commit
- Internal change · 494cf0b3
  Hongkun Yu authored Dec 01, 2019
```
PiperOrigin-RevId: 283266705
```
  494cf0b3
29 Nov, 2019 1 commit
- Fixed tf.compat.v1 bug · a5c96853
  Mirali Ahmadli authored Nov 29, 2019
  
  a5c96853
27 Nov, 2019 2 commits
- Enabled performance related parameters for Transformer: all_reduce_alg, · bd73fdfe
  Hongkun Yu authored Nov 27, 2019
```
enable_eager, tf_gpu_thread_mode and datasets_num_private_threads

PiperOrigin-RevId: 282804724
```
  bd73fdfe
- Enabled performance related parameters for Transformer: all_reduce_alg, · 18df1a09
  A. Unique TensorFlower authored Nov 27, 2019
```
enable_eager, tf_gpu_thread_mode and datasets_num_private_threads

PiperOrigin-RevId: 282797514
```
  18df1a09
25 Nov, 2019 2 commits

Inject enable_runtime_flags into benchmarks. · bcce419a

Sai Ganesh Bandiatmakuri authored Nov 25, 2019

This will help general debugging by enabling custom execution with --benchmark_method_steps.

E.g --benchmark_method_steps=train_steps=7 will run the benchmark for only 7 steps without modifying benchmark code.

PiperOrigin-RevId: 282396875

bcce419a

Internal change · 5b2575c2
Zongwei Zhou authored Nov 24, 2019
```
PiperOrigin-RevId: 282293668
```
5b2575c2

22 Nov, 2019 1 commit
- Transformer Nits: reshape -> set_shape · 22b84f74
  Hongkun Yu authored Nov 21, 2019
```
PiperOrigin-RevId: 281872406
```
  22b84f74
21 Nov, 2019 1 commit
- Move distribution stragegy init to beginning of run · a2a1b66f
  A. Unique TensorFlower authored Nov 20, 2019
```
PiperOrigin-RevId: 281641189
```
  a2a1b66f
18 Nov, 2019 1 commit
- Internal change · c1ac2bfc
  A. Unique TensorFlower authored Nov 18, 2019
```
PiperOrigin-RevId: 281086898
```
  c1ac2bfc
14 Nov, 2019 1 commit
- Update transformer's data_download.py to use TF 1.x compatibility mode. · 63d754ec
  Allen Wang authored Nov 14, 2019
```
PiperOrigin-RevId: 280455983
```
  63d754ec
11 Nov, 2019 1 commit
- [Clean up] Move TF1 only TPU utils to r1 folder. · 0c5919ef
  Hongkun Yu authored Nov 11, 2019
```
PiperOrigin-RevId: 279798675
```
  0c5919ef
29 Oct, 2019 1 commit
- Internal change · a71c9248
  Allen Wang authored Oct 29, 2019
```
PiperOrigin-RevId: 277305436
```
  a71c9248
24 Oct, 2019 2 commits

Remove unnecessary flags · cd00b9a7
Hongkun Yu authored Oct 24, 2019
```
PiperOrigin-RevId: 276518206
```
cd00b9a7

Tensorboard summary writing for few params · 1f4936f1

Madhura Pande authored Oct 24, 2019

Using this code on GPU, was not logging "learning_rate" and "gradient norm" etc. We need to use simple "tf.summary.scalar" instead of contrib one to log these params. Tested on Tensorflow 1.14.

1f4936f1

11 Oct, 2019 3 commits
- Revert "Revert "Update usage of tf.contrib.data to tf.data.experimental" (#7654)" · 35daa566
  Hongkun Yu authored Oct 11, 2019
```
This reverts commit b4e560dc.
```
  35daa566
- Revert "Update usage of tf.contrib.data to tf.data.experimental" (#7654) · b4e560dc
  Hongkun Yu authored Oct 11, 2019
```
* Revert "Update tf.contrib.data to tf.data.experimental. (#7650)"

This reverts commit faf4bbb3.

* revert research
```
  b4e560dc
- Update tf.contrib.data to tf.data.experimental. (#7650) · faf4bbb3
  Derek Murray authored Oct 11, 2019
  
  faf4bbb3
10 Oct, 2019 2 commits

change benchmark's log verbosity to logging.INFO. it seems to me that DEBUG... · 1f9efe10

A. Unique TensorFlower authored Oct 10, 2019

change benchmark's log verbosity to logging.INFO. it seems to me that DEBUG map to ---v=1 internally, which is way to verbose for the purpose of benchmarking.

PiperOrigin-RevId: 274040907

1f9efe10

Internal change · cf4cae61
Hongkun Yu authored Oct 10, 2019
```
PiperOrigin-RevId: 273966871
```
cf4cae61

09 Oct, 2019 1 commit

Simply LayerNorm mixed precision logic. · 0257b276

Reed Wanderman-Milne authored Oct 09, 2019

Instead of needing to ensure variables are float32, casting inputs to float32, etc, instead dtype="float32" is passed to the layer constructor, which will do all that logic automatically.

The only difference is the output of LayerNorm is now float32 instead of float16, so an extra cast is needed elsewhere.

PiperOrigin-RevId: 273833286

0257b276

07 Oct, 2019 1 commit
- Internal change · 41293260
  A. Unique TensorFlower authored Oct 07, 2019
```
PiperOrigin-RevId: 273371605
```
  41293260
18 Sep, 2019 1 commit
- Internal change · 32dadfc2
  David Chen authored Sep 18, 2019
```
PiperOrigin-RevId: 269914707
```
  32dadfc2
17 Sep, 2019 1 commit
- Move Bert to NLP. Tasks are moved to nlp/bert/ · 1862b9c3
  Hongkun Yu authored Sep 17, 2019
```
Refactor basic utils to modeling/

PiperOrigin-RevId: 269600561
```
  1862b9c3
16 Sep, 2019 1 commit
- Update readme to emphasize 'global batch size'. · 841bf60b
  Hongkun Yu authored Sep 16, 2019
```
PiperOrigin-RevId: 269376599
```
  841bf60b
09 Sep, 2019 2 commits

Internal change · 5ceb3acf
David Chen authored Sep 09, 2019
```
PiperOrigin-RevId: 268057019
```
5ceb3acf

Unexpose some flags from models which do not use them. · e91c41c2

Reed Wanderman-Milne authored Sep 09, 2019

--stop_threshold, --num_gpu, --hooks, --export_dir, and --distribution_strategy have been unexposed from models which do not use them

PiperOrigin-RevId: 268032080

e91c41c2

05 Sep, 2019 1 commit
- Internal change · 4bd15fa6
  A. Unique TensorFlower authored Sep 05, 2019
```
PiperOrigin-RevId: 267435985
```
  4bd15fa6
04 Sep, 2019 1 commit

Unexpose some flags from models which do not use them. · a85c40e3

Reed Wanderman-Milne authored Sep 03, 2019

--clean, --train_epochs, and --epochs_between_evals have been unexposed from models which do not use them

PiperOrigin-RevId: 267065651

a85c40e3

03 Sep, 2019 1 commit
- Avoid importing private ObjectIdentitySet class · bd211e3e
  Gaurav Jain authored Sep 02, 2019
```
PiperOrigin-RevId: 266848625
```
  bd211e3e
30 Aug, 2019 1 commit
- Use tf.compat.v2 when accessing policies. · a6f9945a
  Reed Wanderman-Milne authored Aug 30, 2019
```
I soon plan on making mixed precision policies exposed in TF2 only.

PiperOrigin-RevId: 266433372
```
  a6f9945a