Commits · cd00b9a7c367506397bc8752c95d27c31dffbd49 · ModelZoo / ResNet50_tensorflow

"...targets/git@developer.sourcefind.cn:gaoqiong/migraphx.git" did not exist on "5be87179c69447ad03d719e09fba80f82810655f"

24 Oct, 2019 3 commits
- Remove unnecessary flags · cd00b9a7
  Hongkun Yu authored Oct 24, 2019
```
PiperOrigin-RevId: 276518206
```
  cd00b9a7
- Internal change · 87d6459a
  Yeqing Li authored Oct 24, 2019
```
PiperOrigin-RevId: 276516912
```
  87d6459a
- Remove unnecessary test_input_fn. · 88c864f7
  Hongkun Yu authored Oct 23, 2019
```
PiperOrigin-RevId: 276394582
```
  88c864f7
23 Oct, 2019 3 commits
- Internal change · 6b0146f5
  Yeqing Li authored Oct 23, 2019
```
PiperOrigin-RevId: 276344518
```
  6b0146f5
- Adds files to utils folder. · 7d1cfc1e
  Yeqing Li authored Oct 23, 2019
```
PiperOrigin-RevId: 276317091
```
  7d1cfc1e
- Internal change · 638ba7a4
  Hongkun Yu authored Oct 22, 2019
```
PiperOrigin-RevId: 276210315
```
  638ba7a4
22 Oct, 2019 6 commits

Getting rid of experimental_allow_stateful now that we have... · e0460db7

Rohan Jain authored Oct 22, 2019

Getting rid of experimental_allow_stateful now that we have experimental_external_state_policy instead.
that option.

We've replaced experimental_allow_stateful with experimental_external_state_policy with a default of WARN in which we'll only print a warning message and not fail. As a result, we can now safely remove all client code that explicitly set the allow_stateful_ops flag to True.

PiperOrigin-RevId: 276115313

e0460db7

Internal change · 8a5702f2
Ruoxin Sang authored Oct 22, 2019
```
PiperOrigin-RevId: 276103495
```
8a5702f2
Fix EmbeddingPostprocessor · c34c439b
Dmitry Murygin authored Oct 22, 2019

c34c439b

Update XLNet classifier: · 444e8c79

Hongkun Yu authored Oct 21, 2019

- Support first/last summary type
- Support bert format input processing

PiperOrigin-RevId: 276005310

444e8c79

Adds save_steps to fix squad · 9616954e
Hongkun Yu authored Oct 21, 2019
```
PiperOrigin-RevId: 275994710
```
9616954e
train_data_size flag is deleted. · 128326ca
Hongkun Yu authored Oct 21, 2019
```
PiperOrigin-RevId: 275986990
```
128326ca

21 Oct, 2019 6 commits
- Pass the file patterns instead of flattened file paths into... · ad5a14f2
  Ruoxin Sang authored Oct 21, 2019
```
Pass the file patterns instead of flattened file paths into tf.data.Dataset.list_file. This helps Cloud DF4x4 Bert Startup from 11min34s to 6min53s.

PiperOrigin-RevId: 275953359
```
  ad5a14f2
- Move sptial_transforms to ops folder. · 2bceb1b7
  Yeqing Li authored Oct 21, 2019
```
PiperOrigin-RevId: 275940140
```
  2bceb1b7
- Consolidate DistributeOptions.auto_shard into DistributeOptions.auto_shard_policy. · bd0d5692
  Frank Chen authored Oct 21, 2019
```
PiperOrigin-RevId: 275930249
```
  bd0d5692
- Fix typo in utils/flags/_conventions.py · c3bf4a79
  minoring authored Oct 21, 2019
```
arparse -> argparse
```
  c3bf4a79
- implements more container methods for ParamsDict · 800d1dbb
  Hongkun Yu authored Oct 21, 2019
```
PiperOrigin-RevId: 275879424
```
  800d1dbb
- Internal change · 9ca59f8a
  Yeqing Li authored Oct 21, 2019
```
PiperOrigin-RevId: 275867562
```
  9ca59f8a
20 Oct, 2019 1 commit
- Refine readme · befbe0f9
  Hongkun Yu authored Oct 20, 2019
```
PiperOrigin-RevId: 275739207
```
  befbe0f9
19 Oct, 2019 4 commits
- Always pass **kwargs to __call__ override for custom layers. · d169c201
  Jing Li authored Oct 19, 2019
```
PiperOrigin-RevId: 275644913
```
  d169c201
- Internal change · f0f42c82
  Hongkun Yu authored Oct 19, 2019
```
PiperOrigin-RevId: 275610155
```
  f0f42c82
- Fix regression · 9ac54873
  Hongkun Yu authored Oct 18, 2019
```
PiperOrigin-RevId: 275603936
```
  9ac54873
- Internal change · 49ba237d
  A. Unique TensorFlower authored Oct 18, 2019
```
PiperOrigin-RevId: 275578662
```
  49ba237d
18 Oct, 2019 5 commits
- Fix xlnet save_steps and steps_per_epoch conflicts · e6750c5d
  Hongkun Yu authored Oct 18, 2019
```
Remove train_data_size flag

PiperOrigin-RevId: 275545035
```
  e6750c5d
- Update readme · d59bcd47
  Hongkun Yu authored Oct 18, 2019
```
PiperOrigin-RevId: 275428252
```
  d59bcd47
- Move CTL resnet example. · f2c61881
  Hongkun Yu authored Oct 18, 2019
```
PiperOrigin-RevId: 275417626
```
  f2c61881
- Support online masking for XLNet · 6cd426d9
  Jing Li authored Oct 17, 2019
```
PiperOrigin-RevId: 275408074
```
  6cd426d9
- Internal change · b0581d0a
  Hongkun Yu authored Oct 17, 2019
```
PiperOrigin-RevId: 275393975
```
  b0581d0a
17 Oct, 2019 6 commits
- Fix indent · 97f97d99
  minoring authored Oct 17, 2019
  
  97f97d99
- people are commonly having issues configuring paths for ruinning python scripts. · 2517c61d
  marcussorealheis authored Oct 17, 2019
  
  2517c61d
- Internal change · 5375ecec
  Yeqing Li authored Oct 17, 2019
```
PiperOrigin-RevId: 275306010
```
  5375ecec
- Move ResNet CTL benchmark to benchmark folder. · d2f86342
  Hongkun Yu authored Oct 17, 2019
```
PiperOrigin-RevId: 275288636
```
  d2f86342
- Makes do_lower_case not trainable. · eabff6f0
  Hongkun Yu authored Oct 16, 2019
```
PiperOrigin-RevId: 275192365
```
  eabff6f0
- Add compat.v1 to support TF 2.0 in mnist · 9691ef7a
  minoring authored Oct 16, 2019
  
  9691ef7a
16 Oct, 2019 5 commits

Do not scale loss manually for BERT classifier when compile/fit() API is used. · 306e3e14
A. Unique TensorFlower authored Oct 16, 2019
```
PiperOrigin-RevId: 275142626
```
306e3e14
Internal change · 06412123
David Chen authored Oct 16, 2019
```
PiperOrigin-RevId: 275103426
```
06412123
Enables timer callback and disables checkpoint saving in retinanet benchmark test. · 1b77cd80
Yeqing Li authored Oct 16, 2019
```
PiperOrigin-RevId: 275080469
```
1b77cd80

Add support for the tf.keras.mixed_precision API in NCF · cb913691

Reed Wanderman-Milne authored Oct 16, 2019

To test, I did 50 fp32 runs and 50 fp16 runs. I used the following command:

python ncf_keras_main.py --dataset=ml-20m --num_gpus=1 --train_epochs=10 --clean --batch_size=99000 --learning_rate=0.00382059 --beta1=0.783529 --beta2=0.909003 --epsilon=1.45439e-7 --layers=256,256,128,64 --num_factors=64 --hr_threshold=0.635 --ml_perf --nouse_synthetic_data --data_dir ~/ncf_data_dir_python3 --model_dir ~/tmp_model_dir --keras_use_ctl

For the fp16 runs, I added --dtype=fp16. The average hit-rate for both fp16 and fp32 was 0.6365. I also did 50 runs with the mixed precision graph rewrite, and the average hit-rate was 0.6363. The difference is likely due to noise.

PiperOrigin-RevId: 275059871

cb913691

Internal change · e97e22df
Yeqing Li authored Oct 15, 2019
```
PiperOrigin-RevId: 274921478
```
e97e22df

15 Oct, 2019 1 commit
- Internal change · 32229eae
  Yeqing Li authored Oct 15, 2019
```
PiperOrigin-RevId: 274918820
```
  32229eae