- 24 Jul, 2020 6 commits
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 323098007
-
Allen Wang authored
PiperOrigin-RevId: 323016002
-
Chen Chen authored
PiperOrigin-RevId: 323013252
-
Hongkun Yu authored
PiperOrigin-RevId: 322960082
-
Ruoxin Sang authored
PiperOrigin-RevId: 322937598
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 322923416
-
- 23 Jul, 2020 4 commits
-
-
Allen Wang authored
PiperOrigin-RevId: 322884473
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 322844988
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 322713374
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 322707701
-
- 22 Jul, 2020 3 commits
-
-
Allen Wang authored
PiperOrigin-RevId: 322657835
-
Hongkun Yu authored
PiperOrigin-RevId: 322594580
-
A. Unique TensorFlower authored
Explicit set the dtype in embedding layer. Otherwise, using the layer in TF1.x will create an int32 tensor. PiperOrigin-RevId: 322473236
-
- 21 Jul, 2020 4 commits
-
-
Hongkun Yu authored
(1) call() consume kwargs and implement _build_from_signature (layers added inside init_scope) (2) make build/call_attention as public. PiperOrigin-RevId: 322454619
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 322428923
-
Bruce Fontaine authored
PiperOrigin-RevId: 322415899
-
Hongkun Yu authored
PiperOrigin-RevId: 322387911
-
- 20 Jul, 2020 6 commits
-
-
Allen Wang authored
PiperOrigin-RevId: 322244056
-
Allen Wang authored
PiperOrigin-RevId: 322234532
-
Tomer Kaftan authored
Make hack in official/vision/detection models that enters the backend keras graph stop happening once we enable the Functional API KerasTensors refactoring: As a workaround hack for the tf op layer conversion being fragile, the detection models have to explicitly enter the Keras backend graph. When we enable the KerasTensors refactoring of the Functional API internals, the op layer conversion will be much more reliable and this hack will not be necessary. In addition, the hack actually causes the models to break when we enable the refactoring (because it causes tensors to leak out of a graph). So, this CL changes the existing hack to stop applying once we've enabled the KerasTensors refactoring. PiperOrigin-RevId: 322229802
-
Allen Wang authored
PiperOrigin-RevId: 322216928
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 322197751
-
Hongkun Yu authored
Modify the unit test of encoder scaffold to demonstrate that a subclass embedding network can be used. PiperOrigin-RevId: 322083775
-
- 18 Jul, 2020 2 commits
-
-
Yuefeng Zhou authored
PiperOrigin-RevId: 321890251
-
Hongkun Yu authored
PiperOrigin-RevId: 321883870
-
- 17 Jul, 2020 5 commits
-
-
Mark Daoust authored
PiperOrigin-RevId: 321856261
-
Allen Wang authored
PiperOrigin-RevId: 321855387
-
Hongkun Yu authored
PiperOrigin-RevId: 321817352
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 321720155
-
Allen Wang authored
PiperOrigin-RevId: 321676606
-
- 16 Jul, 2020 3 commits
-
-
Allen Wang authored
PiperOrigin-RevId: 321652782
-
Hongkun Yu authored
PiperOrigin-RevId: 321596968
-
Hongkun Yu authored
PiperOrigin-RevId: 321591514
-
- 15 Jul, 2020 1 commit
-
-
Jin Young Sohn authored
PiperOrigin-RevId: 321398549
-
- 14 Jul, 2020 6 commits
-
-
Allen Wang authored
PiperOrigin-RevId: 321261790
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 321238481
-
Jeremiah Harmsen authored
PiperOrigin-RevId: 321119040
-
Chen Chen authored
PiperOrigin-RevId: 321112219
-
Ruoxin Sang authored
PiperOrigin-RevId: 321105003
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 321100219
-