Commits · f5a1343f97bf315b94a4f8a00663bd0f106155cf · ModelZoo / ResNet50_tensorflow

23 Oct, 2020 1 commit

Shift all attribute assignments in functional subclass models below the... · f5a1343f

A. Unique TensorFlower authored Oct 23, 2020

Shift all attribute assignments in functional subclass models below the super().__init__ call, and enable setattr tracking on all functional subclasses.

PiperOrigin-RevId: 338759531

f5a1343f

22 Oct, 2020 1 commit
- Internal change · cdc4cad7
  Poorva Potdar authored Oct 22, 2020
```
PiperOrigin-RevId: 338521847
```
  cdc4cad7
21 Oct, 2020 1 commit
- Change Seq2SeqTransformer inputs to dictionary. · b8c62df7
  Hongkun Yu authored Oct 21, 2020
```
PiperOrigin-RevId: 338309524
```
  b8c62df7
16 Oct, 2020 1 commit
- Log transformer_scaffold configs. · 3bc37a6a
  Le Hou authored Oct 16, 2020
```
PiperOrigin-RevId: 337540237
```
  3bc37a6a
14 Oct, 2020 1 commit
- Implement XLNet QA model. · 44fc54a1
  Allen Wang authored Oct 14, 2020
```
PiperOrigin-RevId: 337134894
```
  44fc54a1
13 Oct, 2020 1 commit
- Slice Transformer outputs with a TF Op layer instead of a Lambda layer. · 5879ca04
  A. Unique TensorFlower authored Oct 13, 2020
```
Lambda layers are fundamentally non-portable since they serialize
Python bytecode.

PiperOrigin-RevId: 336917172
```
  5879ca04
12 Oct, 2020 1 commit
- Implement SpanLabeler for XLNet. · 3e74d7e1
  Allen Wang authored Oct 12, 2020
```
PiperOrigin-RevId: 336709640
```
  3e74d7e1
07 Oct, 2020 1 commit
- Minor bug fixes · fba8ea2f
  Le Hou authored Oct 07, 2020
```
PiperOrigin-RevId: 335944731
```
  fba8ea2f
06 Oct, 2020 1 commit

Clean up internal seq2seqTransformer usage. · 00e598f0

Hongkun Yu authored Oct 05, 2020

Add it to export. TODO: move the encoder and decoder layer out to layers/ or decoders/.

PiperOrigin-RevId: 335560860

00e598f0

05 Oct, 2020 1 commit

PR #42625: Add training call argument for MultiHeadAttention · 75c6e3bc

Tzu-Wei Sung authored Oct 05, 2020

Imported from GitHub PR https://github.com/tensorflow/tensorflow/pull/42625

Miss `training` call argument for dropout layer. /cc @tanzhenyu for visibility.
Copybara import of the project:

--
ba2198fc735a2deca08c68a3c6266d02e01dfe3b by Tzu-Wei Sung <windqaq@gmail.com>:

Add training call argument for MultiHeadAttention

Remove trailing space

PiperOrigin-RevId: 335507127

75c6e3bc

02 Oct, 2020 1 commit
- Implement XLNet Classifier model. · 32c79ea6
  Allen Wang authored Oct 02, 2020
```
PiperOrigin-RevId: 335069348
```
  32c79ea6
25 Sep, 2020 4 commits
- Internal change · e4f89602
  Chen Chen authored Sep 25, 2020
```
PiperOrigin-RevId: 333784610
```
  e4f89602
- Internal change · 2a5c349d
  Chen Chen authored Sep 25, 2020
```
PiperOrigin-RevId: 333784610
```
  2a5c349d
- Internal change · 68d3263d
  Chen Chen authored Sep 25, 2020
```
PiperOrigin-RevId: 333772508
```
  68d3263d
- Internal change · 07a07f6a
  Chen Chen authored Sep 25, 2020
```
PiperOrigin-RevId: 333772508
```
  07a07f6a
24 Sep, 2020 2 commits
- Add norm_first to Transformer Scaffold; Add an option in gated_feedword to... · 7332c1ca
  A. Unique TensorFlower authored Sep 24, 2020
```
Add norm_first to Transformer Scaffold; Add an option in gated_feedword to disable the output layer_norm.

PiperOrigin-RevId: 333591020
```
  7332c1ca
- Add norm_first to Transformer Scaffold; Add an option in gated_feedword to... · eaff981c
  A. Unique TensorFlower authored Sep 24, 2020
```
Add norm_first to Transformer Scaffold; Add an option in gated_feedword to disable the output layer_norm.

PiperOrigin-RevId: 333591020
```
  eaff981c
22 Sep, 2020 6 commits
- Implement XLNetBase. This is the underlying network shared between XLNet models. · 0864d66e
  Allen Wang authored Sep 22, 2020
```
PiperOrigin-RevId: 333183470
```
  0864d66e
- Implement XLNetBase. This is the underlying network shared between XLNet models. · 5a2edc7b
  Allen Wang authored Sep 22, 2020
```
PiperOrigin-RevId: 333183470
```
  5a2edc7b
- Internal Change · 1ac53637
  Chen Chen authored Sep 22, 2020
```
PiperOrigin-RevId: 333163906
```
  1ac53637
- Internal Change · 4c226604
  Chen Chen authored Sep 22, 2020
```
PiperOrigin-RevId: 333163906
```
  4c226604
- Internal change · b5f5bbdd
  Hongkun Yu authored Sep 21, 2020
```
PiperOrigin-RevId: 332992539
```
  b5f5bbdd
- Internal change · c3d53565
  Hongkun Yu authored Sep 21, 2020
```
PiperOrigin-RevId: 332992539
```
  c3d53565
21 Sep, 2020 8 commits
- Internal change · 5b9feb6d
  A. Unique TensorFlower authored Sep 21, 2020
```
PiperOrigin-RevId: 332953126
```
  5b9feb6d
- Internal change · 83704fde
  A. Unique TensorFlower authored Sep 21, 2020
```
PiperOrigin-RevId: 332953126
```
  83704fde
- Internal change · 976ba237
  Hongkun Yu authored Sep 21, 2020
```
PiperOrigin-RevId: 332891157
```
  976ba237
- Internal change · 6f832c92
  Hongkun Yu authored Sep 21, 2020
```
PiperOrigin-RevId: 332891157
```
  6f832c92
- Internal change · 52b16a1a
  Chen Chen authored Sep 21, 2020
```
PiperOrigin-RevId: 332806032
```
  52b16a1a
- Internal change · 56bbc17e
  Chen Chen authored Sep 21, 2020
```
PiperOrigin-RevId: 332806032
```
  56bbc17e
- Clean keras_nlp.BertEncoder: · 2e007a84
  Chen Chen authored Sep 21, 2020
```
Remove return_all_encoder_outputs argument from keras_nlp.BertEncoder as it is unused and we will always return all encoder outputs in the returned dict.

PiperOrigin-RevId: 332802761
```
  2e007a84
- Clean keras_nlp.BertEncoder: · 71a2fc91
  Chen Chen authored Sep 21, 2020
```
Remove return_all_encoder_outputs argument from keras_nlp.BertEncoder as it is unused and we will always return all encoder outputs in the returned dict.

PiperOrigin-RevId: 332802761
```
  71a2fc91
20 Sep, 2020 2 commits

Migrate all TF-NLP tasks to consume models with dictionary outputs and... · 13ed261f

Hongkun Yu authored Sep 20, 2020

Migrate all TF-NLP tasks to consume models with dictionary outputs and encoders with dictionary outputs.

PiperOrigin-RevId: 332740616

13ed261f

Migrate all TF-NLP tasks to consume models with dictionary outputs and... · 3ce3a5c5

Hongkun Yu authored Sep 20, 2020

Migrate all TF-NLP tasks to consume models with dictionary outputs and encoders with dictionary outputs.

PiperOrigin-RevId: 332740616

3ce3a5c5

18 Sep, 2020 7 commits
- Use self.get_temp_dir and simplify the test · 674506c4
  Hongkun Yu authored Sep 18, 2020
```
PiperOrigin-RevId: 332512380
```
  674506c4
- Use self.get_temp_dir and simplify the test · 84c5fcd3
  Hongkun Yu authored Sep 18, 2020
```
PiperOrigin-RevId: 332512380
```
  84c5fcd3
- Add MaskedLM for keras_nlp · 96ca6e05
  Zhenyu Tan authored Sep 18, 2020
```
PiperOrigin-RevId: 332506544
```
  96ca6e05
- Add MaskedLM for keras_nlp · 02275d26
  Zhenyu Tan authored Sep 18, 2020
```
PiperOrigin-RevId: 332506544
```
  02275d26
- Implement Transformer XL · c0b2da2d
  Allen Wang authored Sep 18, 2020
```
PiperOrigin-RevId: 332486600
```
  c0b2da2d
- Implement Transformer XL · 2c98b4b0
  Allen Wang authored Sep 18, 2020
```
PiperOrigin-RevId: 332486600
```
  2c98b4b0
- Internal change · c7e31961
  A. Unique TensorFlower authored Sep 18, 2020
```
PiperOrigin-RevId: 332476566
```
  c7e31961