- 09 Aug, 2022 1 commit
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 466412169
-
- 13 Jul, 2022 1 commit
-
-
Stephan Lee authored
PiperOrigin-RevId: 460731093
-
- 16 Jun, 2022 1 commit
-
-
Ron Shapiro authored
PiperOrigin-RevId: 455390116
-
- 15 Jun, 2022 1 commit
-
-
Ron Shapiro authored
PiperOrigin-RevId: 455136234
-
- 14 Jun, 2022 1 commit
-
-
Ron Shapiro authored
PiperOrigin-RevId: 454861214
-
- 18 Jan, 2022 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 422612138
-
- 03 Dec, 2021 1 commit
-
-
Dinghua Li authored
PiperOrigin-RevId: 413973623
-
- 13 Nov, 2021 1 commit
-
-
Ruoxin Sang authored
Add more documentation to recommend users to set `drop_remainder=False` when exhausting the eval dataset in Orbit. PiperOrigin-RevId: 409539451
-
- 25 Oct, 2021 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 405317879
-
- 20 Oct, 2021 2 commits
-
-
Dinghua Li authored
PiperOrigin-RevId: 404623863
-
Dinghua Li authored
Allow passing a `tf.distribute.InputOptions` to `orbit.make_distributed_dataset` through the `input_options` key word argument. PiperOrigin-RevId: 404597006
-
- 07 Sep, 2021 1 commit
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 395317528
-
- 26 Jul, 2021 1 commit
-
-
Ruoxin Sang authored
Don't early return if the global step is not updated by `num_steps` as expected. This still allows summary and metric reporting. PiperOrigin-RevId: 386967587
-
- 17 Jun, 2021 1 commit
-
-
Dan Holtmann-Rice authored
PiperOrigin-RevId: 379875442
-
- 02 Jun, 2021 1 commit
-
-
Dan Holtmann-Rice authored
PiperOrigin-RevId: 377131555
-
- 28 May, 2021 1 commit
-
-
Dan Holtmann-Rice authored
PiperOrigin-RevId: 376280713
-
- 27 May, 2021 1 commit
-
-
Rebecca Chen authored
PiperOrigin-RevId: 376239468
-
- 10 Apr, 2021 1 commit
-
-
Hongkun Yu authored
Use sparse_categorical_crossentropy for test as the loss object default does not work on tpustrategy + the single task trainer already handles the reduction. PiperOrigin-RevId: 367757677
-
- 09 Apr, 2021 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 367564187
-
- 30 Mar, 2021 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 365828860
-
- 25 Mar, 2021 1 commit
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 365133146
-
- 22 Mar, 2021 1 commit
-
-
A. Unique TensorFlower authored
In current implementation of StandardEvaluator, the iterator of evaluate dataset will go back (actually create a new iterator) to the beginning when evaluate() is called every time. In some case, the iterator need to go ahead forever. So add the new option "recreate_iterator_for_each_eval" to control this behavior. In some case, the iterator creation is time consuming. For example, careting an iterator of distributed dataset for multi worker in eager mode. So it's better to make a repeat dataset instead of recreating an iterator. If "recreate_iterator_for_each_eval" is True, then the iterator will go back; Otherwise, keep going ahead. PiperOrigin-RevId: 364356169
-
- 18 Mar, 2021 1 commit
-
-
Dana Movshovitz-Attias authored
PiperOrigin-RevId: 363673488
-
- 13 Mar, 2021 1 commit
-
-
Yeqing Li authored
PiperOrigin-RevId: 362729857
-
- 10 Mar, 2021 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 362151859
-
- 05 Feb, 2021 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 355925766
-
- 14 Jan, 2021 1 commit
-
-
Ruoxin Sang authored
don't get "while/" as name prefix. PiperOrigin-RevId: 351689649
-
- 13 Jan, 2021 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 351512206
-
- 05 Jan, 2021 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 350170448
-
- 29 Dec, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 349347368
-
- 24 Dec, 2020 1 commit
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 348868176
-
- 11 Dec, 2020 1 commit
-
-
Hongkun Yu authored
We use kwargs for all arguments and enforce them should be kwargs. Reorder args. PiperOrigin-RevId: 346908658
-
- 19 Nov, 2020 1 commit
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 343220781
-
- 17 Nov, 2020 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 342803522
-
- 13 Nov, 2020 3 commits
-
-
anivegesana authored
-
anivegesana authored
-
Hongkun Yu authored
PiperOrigin-RevId: 342276315
-
- 10 Nov, 2020 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 341713387
-
- 01 Nov, 2020 2 commits
-
-
vishnubanna authored
-
Vishnu Banna authored
-