Commits · 0f7580bd42acb914a7dca2e89e357f0d86d931b3 · ModelZoo / ResNet50_tensorflow

19 Nov, 2020 1 commit
- Implement XLNet Pretrainer model. · 15d1514b
  Allen Wang authored Nov 19, 2020
```
PiperOrigin-RevId: 343345144
```
  15d1514b
18 Nov, 2020 5 commits
- minor bug fix · a111ba09
  Le Hou authored Nov 18, 2020
```
PiperOrigin-RevId: 343154699
```
  a111ba09
- Internal change · 6c0250ad
  Hongkun Yu authored Nov 18, 2020
```
PiperOrigin-RevId: 343132085
```
  6c0250ad
- Support adding a layer normalization layer before the final pooling layer. · 713142d1
  Le Hou authored Nov 18, 2020
```
PiperOrigin-RevId: 343092077
```
  713142d1
- Internal change · 04f2cd31
  Hongkun Yu authored Nov 17, 2020
```
PiperOrigin-RevId: 343012072
```
  04f2cd31
- Internal change · 44a545b6
  Hongkun Yu authored Nov 17, 2020
```
PiperOrigin-RevId: 342972267
```
  44a545b6
17 Nov, 2020 1 commit
- [Cleanup] Replace tf.distribute.experimental.TPUStrategy with tf.distribute.TPUStrategy · b6ece654
  Hongkun Yu authored Nov 16, 2020
```
PiperOrigin-RevId: 342770296
```
  b6ece654
16 Nov, 2020 1 commit
- Internal change · 4d34474e
  Chen Chen authored Nov 16, 2020
```
PiperOrigin-RevId: 342726521
```
  4d34474e
12 Nov, 2020 2 commits
- Internal change · 31e4a64d
  Allen Wang authored Nov 12, 2020
```
PiperOrigin-RevId: 342129726
```
  31e4a64d
- Make two beam search utils function as public. They are helpful. · b15a86fc
  Hongkun Yu authored Nov 11, 2020
```
PiperOrigin-RevId: 341953641
```
  b15a86fc
11 Nov, 2020 1 commit
- Move mobilebert embedding and transformer layers to modeling/layers · b75a491a
  Chen Chen authored Nov 11, 2020
```
PiperOrigin-RevId: 341913299
```
  b75a491a
10 Nov, 2020 2 commits
- Allow customized masked lm in BertPretrainerV2. · 2498f88f
  Chen Chen authored Nov 10, 2020
```
PiperOrigin-RevId: 341708068
```
  2498f88f
- Internal change · e3a000ad
  Allen Wang authored Nov 10, 2020
```
PiperOrigin-RevId: 341642152
```
  e3a000ad
09 Nov, 2020 1 commit
- Internal change · 49c66406
  Poorva Potdar authored Nov 09, 2020
```
PiperOrigin-RevId: 341433786
```
  49c66406
04 Nov, 2020 1 commit
- Internal change · 1a0b9079
  Hongkun Yu authored Nov 03, 2020
```
PiperOrigin-RevId: 340580428
```
  1a0b9079
28 Oct, 2020 1 commit
- Move PackedSequenceEmbedding to model garden · bddc4930
  Chen Chen authored Oct 28, 2020
```
PiperOrigin-RevId: 339454557
```
  bddc4930
27 Oct, 2020 2 commits
- Save layers for access in EncoderScaffold. · 5ec73492
  Le Hou authored Oct 27, 2020
```
PiperOrigin-RevId: 339345258
```
  5ec73492
- Cleanup for PY3. · dfbad458
  Hongkun Yu authored Oct 26, 2020
```
PiperOrigin-RevId: 339180185
```
  dfbad458
26 Oct, 2020 3 commits
- Internal change · 764c18c5
  Will Cromar authored Oct 26, 2020
```
PiperOrigin-RevId: 339131213
```
  764c18c5
- Internal change · 801ac678
  Allen Wang authored Oct 26, 2020
```
PiperOrigin-RevId: 339095008
```
  801ac678
- Py3 cleanup for nlp/ · 89fd2bb1
  Hongkun Yu authored Oct 26, 2020
```
PiperOrigin-RevId: 339071563
```
  89fd2bb1
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 2 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