- 05 Jul, 2021 1 commit
-
-
Frederick Liu authored
PiperOrigin-RevId: 383066915
-
- 04 Jul, 2021 1 commit
-
-
Frederick Liu authored
PiperOrigin-RevId: 382960239
-
- 03 Jul, 2021 1 commit
-
-
Frederick Liu authored
PiperOrigin-RevId: 382846192
-
- 02 Jul, 2021 2 commits
-
-
Dan Kondratyuk authored
PiperOrigin-RevId: 382833629
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 382655843
-
- 01 Jul, 2021 7 commits
-
-
Jaehong Kim authored
Added use_keras_upsampling_2d flags to use keras upsampling layer instead of optimized custom spatial_transform_ops. PiperOrigin-RevId: 382626602
-
Dan Kondratyuk authored
PiperOrigin-RevId: 382606601
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 382571340
-
Abdullah Rashwan authored
PiperOrigin-RevId: 382561014
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 382547358
-
Abdullah Rashwan authored
PiperOrigin-RevId: 382462213
-
Abdullah Rashwan authored
PiperOrigin-RevId: 382422745
-
- 30 Jun, 2021 4 commits
-
-
Xianzhi Du authored
PiperOrigin-RevId: 382408293
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 382396878
-
Xianzhi Du authored
PiperOrigin-RevId: 382382393
-
Abdullah Rashwan authored
PiperOrigin-RevId: 382240268
-
- 29 Jun, 2021 2 commits
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 382051540
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 382051212
-
- 28 Jun, 2021 1 commit
-
-
Dan Kondratyuk authored
PiperOrigin-RevId: 381908843
-
- 25 Jun, 2021 5 commits
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 381540013
-
Fan Yang authored
PiperOrigin-RevId: 381516130
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 381396116
-
Fan Yang authored
PiperOrigin-RevId: 381368435
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 381367878
-
- 24 Jun, 2021 8 commits
-
-
A. Unique TensorFlower authored
Adds a feature to process a dictionary for `input_paths` in `DataConfig`, to allow combining multiple datasets using a user defined combine_fn. PiperOrigin-RevId: 381363688
-
Atze00 authored
-
A. Unique TensorFlower authored
Add model export config to RetinaNet, which optionally cast model outputs to floats, and normalize output box coordinates. PiperOrigin-RevId: 381334851
-
Atze00 authored
-
Jaehong Kim authored
PiperOrigin-RevId: 381320453
-
Reed Wanderman-Milne authored
PiperOrigin-RevId: 381302776
-
Yeqing Li authored
PiperOrigin-RevId: 381301573
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 381152422
-
- 23 Jun, 2021 3 commits
-
-
Vincent Dumoulin authored
PiperOrigin-RevId: 381089283
-
Reed Wanderman-Milne authored
In nlp/train.py and vision/beta/train.py, certain flags are marked as required. Additionally, in certain functions, error messages are improved if a necessary flag is not specified, which is a fallback in case a file calling define_flags() does not mark the necessary flags are required. Previously if any of these flags were not specified, it would crash with a cryptic error message, making it hard to tell what went wrong. In a subsequent change, I will mark flags as required in more files which call define_flags(). PiperOrigin-RevId: 381066985
-
Reed Wanderman-Milne authored
Before, it returned None. But almost every use of get_distribution_strategy() assumes an actual strategy is returned and crashes when None is returned. Returning the default strategy fixes these issues and is equivalent to using no strategy, as the default strategy is always in effect when no other strategy is used. PiperOrigin-RevId: 380951055
-
- 22 Jun, 2021 3 commits
-
-
Yeqing Li authored
Adds a video classification export module for example. The module is subject to change in the near future. PiperOrigin-RevId: 380907598
-
Yeqing Li authored
PiperOrigin-RevId: 380858205
-
Reed Wanderman-Milne authored
Some parts of the forward pass would previously overflow in float16. Such parts are now done in float32. PiperOrigin-RevId: 380857663
-
- 21 Jun, 2021 2 commits
-
-
Xianzhi Du authored
PiperOrigin-RevId: 380687139
-
Vincent Dumoulin authored
PiperOrigin-RevId: 380650647
-