Commits · b63c955f988a84012d54ba8f7ea9cb46f498dd23 · ModelZoo / ResNet50_tensorflow

03 Nov, 2020 1 commit
- Internal change · b63c955f
  Pankaj Kanwar authored Nov 03, 2020
```
PiperOrigin-RevId: 340527173
```
  b63c955f
30 Oct, 2020 2 commits
- Internal change · 08d3c799
  Le Hou authored Oct 30, 2020
```
PiperOrigin-RevId: 339901099
```
  08d3c799
- Internal change · 162f3741
  Hongkun Yu authored Oct 29, 2020
```
PiperOrigin-RevId: 339804825
```
  162f3741
28 Oct, 2020 1 commit
- Move PackedSequenceEmbedding to model garden · 290cd97b
  Chen Chen authored Oct 28, 2020
```
PiperOrigin-RevId: 339454557
```
  290cd97b
27 Oct, 2020 3 commits
- Save layers for access in EncoderScaffold. · 9f9b4307
  Le Hou authored Oct 27, 2020
```
PiperOrigin-RevId: 339345258
```
  9f9b4307
- Add in a comment about legacy input masks for XLNet. · 05a95e16
  Allen Wang authored Oct 27, 2020
```
PiperOrigin-RevId: 339267061
```
  05a95e16
- Cleanup for PY3. · b1809d94
  Hongkun Yu authored Oct 26, 2020
```
PiperOrigin-RevId: 339180185
```
  b1809d94
26 Oct, 2020 7 commits
- Add in IMDB dataset processor to TF-NLP. · d3265962
  Allen Wang authored Oct 26, 2020
```
PiperOrigin-RevId: 339134320
```
  d3265962
- Internal change · 4dc945c0
  Will Cromar authored Oct 26, 2020
```
PiperOrigin-RevId: 339131213
```
  4dc945c0
- Internal change · 3421f8c6
  Hongkun Yu authored Oct 26, 2020
```
PiperOrigin-RevId: 339113572
```
  3421f8c6
- Internal change · b95fa6e1
  Allen Wang authored Oct 26, 2020
```
PiperOrigin-RevId: 339095008
```
  b95fa6e1
- Make 'embedding_layer' an optional config field for backwards compatibility. · 53f76b23
  A. Unique TensorFlower authored Oct 26, 2020
```
PiperOrigin-RevId: 339090333
```
  53f76b23
- Py3 cleanup for nlp/ · cd5e1204
  Hongkun Yu authored Oct 26, 2020
```
PiperOrigin-RevId: 339071563
```
  cd5e1204
- Add support for SQuAD BERT export · 282adff5
  Elizabeth Kemp authored Oct 26, 2020
```
PiperOrigin-RevId: 339018438
```
  282adff5
24 Oct, 2020 1 commit
- Do not create embedding projection attr if it is None, for backwards compatibility. · 463dc2d0
  A. Unique TensorFlower authored Oct 23, 2020
```
PiperOrigin-RevId: 338786173
```
  463dc2d0
23 Oct, 2020 1 commit

Shift all attribute assignments in functional subclass models below the... · 589f0269

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

589f0269

22 Oct, 2020 2 commits
- Internal change · 414b7b74
  Poorva Potdar authored Oct 22, 2020
```
PiperOrigin-RevId: 338521847
```
  414b7b74
- Workaround a known issue with control dependency on external tensors · aca137c1
  Ran Chen authored Oct 22, 2020
```
This fix tensorflow_models/official/nlp/bert/run_pretraining failure with
explicit allreduce.

PiperOrigin-RevId: 338521478
```
  aca137c1
21 Oct, 2020 2 commits
- Change Seq2SeqTransformer inputs to dictionary. · f3641f23
  Hongkun Yu authored Oct 21, 2020
```
PiperOrigin-RevId: 338309524
```
  f3641f23
- Avoid loading hub model twice · 09b9dad7
  Chen Chen authored Oct 21, 2020
```
PiperOrigin-RevId: 338249187
```
  09b9dad7
20 Oct, 2020 1 commit
- internal change · 6e63dfee
  Chen Chen authored Oct 20, 2020
```
PiperOrigin-RevId: 338095907
```
  6e63dfee
19 Oct, 2020 1 commit
- Internal change · cb2528eb
  Hongkun Yu authored Oct 19, 2020
```
PiperOrigin-RevId: 337927475
```
  cb2528eb
17 Oct, 2020 2 commits

Do not access self.embeddings.dtype. · 126ce652

Reed Wanderman-Milne authored Oct 17, 2020

Soon, AutoCastVariable.dtype will refer to the variable dtype, not the compute
dtype. This change stops using AutoCastVariable.dtype.

PiperOrigin-RevId: 337670544

126ce652

Internal change · 19113a57
Ruoxin Sang authored Oct 16, 2020
```
PiperOrigin-RevId: 337609198
```
19113a57

16 Oct, 2020 1 commit
- Log transformer_scaffold configs. · 992a5bfa
  Le Hou authored Oct 16, 2020
```
PiperOrigin-RevId: 337540237
```
  992a5bfa
15 Oct, 2020 1 commit
- Fix command in XLNet README. · 14a6c92e
  Allen Wang authored Oct 15, 2020
```
PiperOrigin-RevId: 337366497
```
  14a6c92e
14 Oct, 2020 2 commits
- Implement XLNet QA model. · 7a6a8741
  Allen Wang authored Oct 14, 2020
```
PiperOrigin-RevId: 337134894
```
  7a6a8741
- Internal change · 87e4768e
  Hongkun Yu authored Oct 13, 2020
```
PiperOrigin-RevId: 336991213
```
  87e4768e
13 Oct, 2020 3 commits
- Slice Transformer outputs with a TF Op layer instead of a Lambda layer. · cd39fd58
  A. Unique TensorFlower authored Oct 13, 2020
```
Lambda layers are fundamentally non-portable since they serialize
Python bytecode.

PiperOrigin-RevId: 336917172
```
  cd39fd58
- Slice BERT output with a TF Op layer instead of a Lambda layer. · cf084d8b
  A. Unique TensorFlower authored Oct 13, 2020
```
Lambda layers are fundamentally non-portable since they serialize
Python bytecode.

PiperOrigin-RevId: 336910013
```
  cf084d8b
- Move core configs to core/ folder. Leave the configs for legacy models. · 33c61588
  Hongkun Yu authored Oct 12, 2020
```
PiperOrigin-RevId: 336795303
```
  33c61588
12 Oct, 2020 1 commit
- Implement SpanLabeler for XLNet. · fcb43c38
  Allen Wang authored Oct 12, 2020
```
PiperOrigin-RevId: 336709640
```
  fcb43c38
08 Oct, 2020 1 commit
- internal change · 92752da2
  Chen Chen authored Oct 08, 2020
```
PiperOrigin-RevId: 336043643
```
  92752da2
07 Oct, 2020 2 commits
- Fix Typo in XLNet readme · 0ba8a3b0
  Allen Wang authored Oct 07, 2020
```
PiperOrigin-RevId: 335973329
```
  0ba8a3b0
- Minor bug fixes · d4557be2
  Le Hou authored Oct 07, 2020
```
PiperOrigin-RevId: 335944731
```
  d4557be2
06 Oct, 2020 2 commits

Fix https://github.com/tensorflow/models/issues/9336 · 92e115a1

Chen Chen authored Oct 06, 2020

to support both model and encoder as key in ckpt when exporting to tfhub

PiperOrigin-RevId: 335668823

92e115a1

Clean up internal seq2seqTransformer usage. · 662a40be

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

662a40be

05 Oct, 2020 1 commit

PR #42625: Add training call argument for MultiHeadAttention · 229f1dd5

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

229f1dd5

03 Oct, 2020 1 commit
- Support to convert albert tf1 ckpt to tf2's BertPretrainerV2 · 3a9ed6bd
  Chen Chen authored Oct 03, 2020
```
PiperOrigin-RevId: 335226408
```
  3a9ed6bd
02 Oct, 2020 1 commit
- Internal change · 5b6be76b
  Hongkun Yu authored Oct 02, 2020
```
PiperOrigin-RevId: 335106919
```
  5b6be76b