Commits · 68dc253ea8273b0907cd36a45a5b78e3f6d2d024 · ModelZoo / ResNet50_tensorflow

23 Feb, 2021 2 commits
- convert Dimension to int to avoid "TypeError: unsupported operand type(s) for... · d2d32f46
  A. Unique TensorFlower authored Feb 23, 2021
```
convert Dimension to int to avoid "TypeError: unsupported operand type(s) for /: float and Dimension"

PiperOrigin-RevId: 358988747
```
  d2d32f46
- convert Dimension to int to avoid "TypeError: unsupported operand type(s) for... · cb342a74
  A. Unique TensorFlower authored Feb 23, 2021
```
convert Dimension to int to avoid "TypeError: unsupported operand type(s) for /: float and Dimension"

PiperOrigin-RevId: 358988747
```
  cb342a74
05 Jan, 2021 2 commits
- #KerasNLP Update TransformerEncoderBlock to support Q, KV as two input streams. · 97807e63
  Hongkun Yu authored Jan 05, 2021
```
PiperOrigin-RevId: 350170448
```
  97807e63
- #KerasNLP Update TransformerEncoderBlock to support Q, KV as two input streams. · 86df41f7
  Hongkun Yu authored Jan 05, 2021
```
PiperOrigin-RevId: 350170448
```
  86df41f7
21 Dec, 2020 1 commit
- [official/**.py] Rename "Arguments:" to "Args:" · 90979a21
  Samuel Marks authored Dec 21, 2020
  
  90979a21
22 Sep, 2020 2 commits
- 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
14 Sep, 2020 2 commits
- Replace keras_nlp.module with keras_nlp.layers.module · fbfa69c8
  Zhenyu Tan authored Sep 14, 2020
```
PiperOrigin-RevId: 331620370
```
  fbfa69c8
- Replace keras_nlp.module with keras_nlp.layers.module · 83c4844e
  Zhenyu Tan authored Sep 14, 2020
```
PiperOrigin-RevId: 331620370
```
  83c4844e
28 Aug, 2020 2 commits
- Define the package pattern for keras-nlp · 9ac54b65
  Hongkun Yu authored Aug 27, 2020
```
PiperOrigin-RevId: 328872353
```
  9ac54b65
- Define the package pattern for keras-nlp · 1eb0ec0e
  Hongkun Yu authored Aug 27, 2020
```
PiperOrigin-RevId: 328872353
```
  1eb0ec0e
27 Aug, 2020 2 commits
- Internal change · 37d40626
  Zhenyu Tan authored Aug 26, 2020
```
PiperOrigin-RevId: 328674302
```
  37d40626
- Internal change · ed38358f
  Zhenyu Tan authored Aug 26, 2020
```
PiperOrigin-RevId: 328674302
```
  ed38358f
19 Aug, 2020 4 commits
- Polish Seq2SeqTransformer: (1) consolidate args; (2) add tests for... · 30821184
  Hongkun Yu authored Aug 19, 2020
```
Polish Seq2SeqTransformer: (1) consolidate args; (2) add tests for distribution strategy and decoding path. (3) fix bugs

PiperOrigin-RevId: 327455733
```
  30821184
- Polish Seq2SeqTransformer: (1) consolidate args; (2) add tests for... · 4bf01a43
  Hongkun Yu authored Aug 19, 2020
```
Polish Seq2SeqTransformer: (1) consolidate args; (2) add tests for distribution strategy and decoding path. (3) fix bugs

PiperOrigin-RevId: 327455733
```
  4bf01a43
- Enforce PY3 for official/nlp/modeling · e36a5fec
  Hongkun Yu authored Aug 18, 2020
```
PiperOrigin-RevId: 327363070
```
  e36a5fec
- Enforce PY3 for official/nlp/modeling · 8a78c154
  Hongkun Yu authored Aug 18, 2020
```
PiperOrigin-RevId: 327363070
```
  8a78c154
13 Aug, 2020 2 commits
- Replace tensorflow_models MultiHeadAttention with tf.keras.MultiHeadAttention. · 2444a510
  Zhenyu Tan authored Aug 13, 2020
```
PiperOrigin-RevId: 326496940
```
  2444a510
- Replace tensorflow_models MultiHeadAttention with tf.keras.MultiHeadAttention. · 197978fe
  Zhenyu Tan authored Aug 13, 2020
```
PiperOrigin-RevId: 326496940
```
  197978fe
07 Aug, 2020 3 commits
- initializers.serialize · 601daf54
  xinliupitt authored Aug 07, 2020
  
  601daf54
- initializers.get · 9575381a
  xinliupitt authored Aug 07, 2020
  
  9575381a
- attention_initializer · 2deb09ae
  xinliupitt authored Aug 07, 2020
  
  2deb09ae
03 Aug, 2020 1 commit
- intermediate dropout · dc03c043
  xinliupitt authored Aug 03, 2020
  
  dc03c043
30 Jul, 2020 4 commits
- transformer test case · 72efa854
  xinliupitt authored Jul 30, 2020
  
  72efa854
- get_config and doc · 1c5dca7f
  xinliupitt authored Jul 30, 2020
  
  1c5dca7f
- Remove backward compatibility hack. · 52504780
  Hongkun Yu authored Jul 29, 2020
```
PiperOrigin-RevId: 323927843
```
  52504780
- Remove backward compatibility hack. · de8c6ace
  Hongkun Yu authored Jul 29, 2020
```
PiperOrigin-RevId: 323927843
```
  de8c6ace
29 Jul, 2020 2 commits
- docstrings · d39321b1
  xinliupitt authored Jul 29, 2020
  
  d39321b1
- transformer, attention layers · 23804bc5
  xinliupitt authored Jul 29, 2020
  
  23804bc5
21 Jul, 2020 2 commits

[MultiheadAttention] Apply suggestions from RFC: · 36101ab4

Hongkun Yu authored Jul 21, 2020

(1) call() consume kwargs and implement _build_from_signature (layers added inside init_scope)
(2) make build/call_attention as public.

PiperOrigin-RevId: 322454619

36101ab4

[MultiheadAttention] Apply suggestions from RFC: · 570d9a2b

Hongkun Yu authored Jul 21, 2020

(1) call() consume kwargs and implement _build_from_signature (layers added inside init_scope)
(2) make build/call_attention as public.

PiperOrigin-RevId: 322454619

570d9a2b

08 Jul, 2020 2 commits
- Migrate all DenseEinsum to tf.keras.experimental.EinsumDense · 9a9f0b65
  Hongkun Yu authored Jul 08, 2020
```
PiperOrigin-RevId: 320240466
```
  9a9f0b65
- Migrate all DenseEinsum to tf.keras.experimental.EinsumDense · 43f5340f
  Hongkun Yu authored Jul 08, 2020
```
PiperOrigin-RevId: 320240466
```
  43f5340f
22 Jun, 2020 1 commit
- Internal change · 8aa44501
  Hongkun Yu authored Jun 21, 2020
```
PiperOrigin-RevId: 317596394
```
  8aa44501
19 Jun, 2020 1 commit
- Move TransformerDecoderLayer to modeling/ · 58e805e0
  Hongkun Yu authored Jun 19, 2020
```
PiperOrigin-RevId: 317330705
```
  58e805e0
29 May, 2020 1 commit

Proposes the full functionality of MultiHeadAttention layer. This change first... · 8a13ca4e

Hongkun Yu authored May 29, 2020

Proposes the full functionality of MultiHeadAttention layer. This change first goes to model garden NLP library.

PiperOrigin-RevId: 313847485

8a13ca4e

28 May, 2020 1 commit

Use float32 activation in Transformer. · 94b1efc1

Reed Wanderman-Milne authored May 28, 2020

Float32 is used if the model uses mixed precision with bfloat16. Float16 activation are unchanged.

The motivation is that BERT with the LAMB optimizer with a gelu activation has an unstable loss when gelu is in bfloat16. Unfortunately, it is not easy to check if the LAMB optimizer and gelu is used, and perhaps there are other cases that work better with float32 activations instead of bfloat16 activations, so we always do the activation in float32 instead of bfloat16.

PiperOrigin-RevId: 313618322

94b1efc1

12 May, 2020 2 commits
- Make Transformer Layer back compatible. · ac5fff19
  Hongkun Yu authored May 12, 2020
```
PiperOrigin-RevId: 311165658
```
  ac5fff19
- Internal Change · 2878da2e
  Chen Chen authored May 12, 2020
```
PiperOrigin-RevId: 311072125
```
  2878da2e
10 May, 2020 1 commit
- Internal change · 09c5ae2f
  Hongkun Yu authored May 09, 2020
```
PiperOrigin-RevId: 310767440
```
  09c5ae2f