Commits · 61aee86e4cc2a7c5aac82c01bef0173dda81a0aa · ModelZoo / ResNet50_tensorflow

14 Nov, 2019 2 commits
- Add pretrained checkpoints and TFHub links · 61aee86e
  Hongkun Yu authored Nov 13, 2019
```
PiperOrigin-RevId: 280321302
```
  61aee86e
- Fix segmentation fault on SSD tflite client destruction. (#7804) · 8475c0bb
  Henry authored Nov 13, 2019
```
* Fix segmentation fault on client destruction.

edgetpu::EdgeTpuContext needs to be created before tflite::Interpreter
```
  8475c0bb
13 Nov, 2019 5 commits

Internal changes. · 8861a624
Pengchong Jin authored Nov 13, 2019
```
PiperOrigin-RevId: 280262689
```
8861a624
Internal change · 13e18326
A. Unique TensorFlower authored Nov 13, 2019
```
PiperOrigin-RevId: 280253871
```
13e18326

Merged commit includes the following changes: (#7800) · b968a6ce

Mark Sandler authored Nov 13, 2019

280142968  by Zhichao Lu:

    Opensource MobilenetEdgeTPU + ssdlite into third-party object detection APIs on EdgeTPU.

--
280134001  by Zhichao Lu:

    Adds MobilenetEdgeTpu + ssdlite into internal object detection APIs on EdgeTPU.

--
278941778  by Zhichao Lu:

    Add support for fixed input shapes for 'encoded_image_string_tensor' and 'tf_example' inputs.

--
278933274  by Zhichao Lu:

      Adding fool proof check to avoid using 1x1 depthwise conv op.

--
278762192  by Zhichao Lu:

    Ensure correct number of iterations after training resumes.

--
278746440  by Zhichao Lu:

    Internal change.

--
278006953  by Zhichao Lu:

    Internal changes to tf.contrib symbols

--
278006330  by Zhichao Lu:

    Internal changes to tf.contrib symbols

--
277593959  by Zhichao Lu:

      Make the ssd_feature_extractor_test.py PY3 compatible. The "six.zip" will use "itertools.izip" in Python 2 and "zip" in Python 3....

b968a6ce

Internal change · c3bd5082
A. Unique TensorFlower authored Nov 12, 2019
```
PiperOrigin-RevId: 280123567
```
c3bd5082

Merged commit includes the following changes: (#7798) · 377c5285

Mark Sandler authored Nov 12, 2019

280061731  by Sergio Guadarrama:

    Fix table formatting in mobilenet readme

--
280059345  by Sergio Guadarrama:

    Internal change

PiperOrigin-RevId: 280061731

377c5285

12 Nov, 2019 9 commits

Merged commit includes the following changes: (#7797) · ba87e2c6

Mark Sandler authored Nov 12, 2019

279978375  by Sergio Guadarrama:

    Pass s=2 to the expanded_conv block so it can apply residual correctly in case of fused convolutions.  (Before it was relying on channel mismatch only)

--
279788358  by Sergio Guadarrama:

    Update README to add mobilenet-edgetpu details

--
279774392  by Sergio Guadarrama:

    Adds MobilenetV3-EdgeTpu definition.

--
278917344  by Sergio Guadarrama:

    Create visualwakewords dataset using slim scripts instead of custom scripts.

--
277940048  by Sergio Guadarrama:

    Internal changes to tf.contrib symbols

--

PiperOrigin-RevId: 279978375

ba87e2c6

Add the sample offset option in mask interpolation. · c2902cfb
Pengchong Jin authored Nov 12, 2019
```
PiperOrigin-RevId: 280031944
```
c2902cfb
Internal change · a76237da
Rajagopal Ananthanarayanan authored Nov 12, 2019
```
PiperOrigin-RevId: 280019807
```
a76237da

Fix MLM accuracy bug. We should not just take reduce mean. The sum of weights... · 95dc9045

Hongkun Yu authored Nov 12, 2019

Fix MLM accuracy bug. We should not just take reduce mean. The sum of weights should be the denominator.

PiperOrigin-RevId: 280002181

95dc9045

Fix export model script (#7791) · 0177deeb
Yukun Zhu authored Nov 11, 2019

0177deeb
Fix init_fn in train_utils. (#7790) · 87bc52c5
Yukun Zhu authored Nov 11, 2019

87bc52c5
Support to use KerasBERT for squad task. · d48d036a
Chen Chen authored Nov 11, 2019
```
PiperOrigin-RevId: 279873276
```
d48d036a

Update new models and internal changes (#7788) · 146a37c6

Yukun Zhu authored Nov 11, 2019

* Update for py3 and some internal changes

* \nInternal refactor\n

PiperOrigin-RevId: 279809827

* clean up

146a37c6

Update for py3 and some internal changes (#7786) · 1498d941
Yukun Zhu authored Nov 11, 2019

1498d941

11 Nov, 2019 5 commits
- Internal change · 42c3b8f0
  A. Unique TensorFlower authored Nov 11, 2019
```
PiperOrigin-RevId: 279834138
```
  42c3b8f0
- [Clean up] Move TF1 only TPU utils to r1 folder. · 0c5919ef
  Hongkun Yu authored Nov 11, 2019
```
PiperOrigin-RevId: 279798675
```
  0c5919ef
- Internal change · 0ae21b98
  Hongkun Yu authored Nov 11, 2019
```
PiperOrigin-RevId: 279782659
```
  0ae21b98
- Fix checkpoint path in bert_benchmark.py. · 0faa2ff3
  Hongkun Yu authored Nov 11, 2019
```
PiperOrigin-RevId: 279768972
```
  0faa2ff3
- Release keras bert: · f1d35b4e
  Hongkun Yu authored Nov 11, 2019
```
- Update classifier example.
- Add new converted checkpoints.
- Update benchmark,

PiperOrigin-RevId: 279762797
```
  f1d35b4e
10 Nov, 2019 2 commits
- Update bert readme: uses correct tf-nightly pip package · 0351cb87
  Hongkun Yu authored Nov 10, 2019
```
PiperOrigin-RevId: 279631484
```
  0351cb87
- Remove guptapriya from code owners · a5e7c134
  guptapriya authored Nov 10, 2019
  
  a5e7c134
08 Nov, 2019 1 commit
- Internal change · 4972c084
  A. Unique TensorFlower authored Nov 08, 2019
```
PiperOrigin-RevId: 279361336
```
  4972c084
07 Nov, 2019 6 commits
- Updated code to Tensorflow 2.0 (#7770) · 426b2c6e
  Silas-Asamoah authored Nov 07, 2019
  
  426b2c6e
- Clean up: remove resnet folder from github. · aa031d91
  Hongkun Yu authored Nov 07, 2019
```
PiperOrigin-RevId: 279117906
```
  aa031d91
- Check NaN during training loop. · 14efeaa5
  Yeqing Li authored Nov 07, 2019
```
PiperOrigin-RevId: 279116219
```
  14efeaa5
- Adds support for separable convolution. · 98db9b25
  Yeqing Li authored Nov 07, 2019
```
PiperOrigin-RevId: 279113063
```
  98db9b25
- Update losses. · b2bf29cf
  Yeqing Li authored Nov 07, 2019
```
PiperOrigin-RevId: 279111017
```
  b2bf29cf
- Internal change · 833e6939
  A. Unique TensorFlower authored Nov 06, 2019
```
PiperOrigin-RevId: 278969067
```
  833e6939
06 Nov, 2019 3 commits
- Internal change · bf0dc049
  Ayush Dubey authored Nov 06, 2019
```
PiperOrigin-RevId: 278902210
```
  bf0dc049
- Internal change · 94833324
  Yeqing Li authored Nov 05, 2019
```
PiperOrigin-RevId: 278756245
```
  94833324
- Internal change · 02cc984e
  A. Unique TensorFlower authored Nov 05, 2019
```
PiperOrigin-RevId: 278754959
```
  02cc984e
05 Nov, 2019 3 commits
- Fix a comment. · 0203278a
  Hongkun Yu authored Nov 05, 2019
```
PiperOrigin-RevId: 278672795
```
  0203278a
- Move official/mnist folder to official/r1/mnist. · b4488b30
  Will Cromar authored Nov 05, 2019
```
PiperOrigin-RevId: 278649613
```
  b4488b30
- Internal change · 17d26cb7
  Will Cromar authored Nov 04, 2019
```
PiperOrigin-RevId: 278508058
```
  17d26cb7
04 Nov, 2019 1 commit
- Internal change · 28c5f526
  Will Cromar authored Nov 04, 2019
```
PiperOrigin-RevId: 278394352
```
  28c5f526
02 Nov, 2019 1 commit
- Internal change · b3e74d15
  Hongkun Yu authored Nov 01, 2019
```
PiperOrigin-RevId: 278028560
```
  b3e74d15
01 Nov, 2019 2 commits

Fix mnist_test.py · 0734276a
Will Cromar authored Nov 01, 2019
```
PiperOrigin-RevId: 278024052
```
0734276a

Support fp16 using tf.keras.mixed_precision in CTL resnet. · ffa522ea

Reed Wanderman-Milne authored Nov 01, 2019

To test, I ran the following command:

python resnet_ctl_imagenet_main.py --batch_size=2048 --data_dir ~/imagenet --datasets_num_private_threads=14 --epochs_between_evals=10 --model_dir ~/tmp_model_dir --clean --num_gpus=8 --train_epochs=90 --dtype=fp16

I got 76.15% final evaluation accuracy.

PiperOrigin-RevId: 278010061

ffa522ea