- 30 Jan, 2021 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 354647274
-
- 29 Jan, 2021 1 commit
-
-
Yeqing Li authored
PiperOrigin-RevId: 354434423
-
- 22 Jan, 2021 1 commit
-
-
Le Hou authored
PiperOrigin-RevId: 353327814
-
- 20 Jan, 2021 2 commits
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 352730170
-
Hongkun Yu authored
PiperOrigin-RevId: 352728348
-
- 13 Jan, 2021 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 351524712
-
- 08 Jan, 2021 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 350687522
-
- 24 Dec, 2020 1 commit
-
-
Chen Chen authored
when the number of files is less than the number of input pipelines, fall back to "_read_single_file" which will send all files to every workers. PiperOrigin-RevId: 348865649
-
- 21 Dec, 2020 1 commit
-
-
Samuel Marks authored
-
- 19 Dec, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 348291651
-
- 14 Dec, 2020 2 commits
-
-
Chen Chen authored
PiperOrigin-RevId: 347480509
-
Abdullah Rashwan authored
PiperOrigin-RevId: 347439073
-
- 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
-
- 08 Dec, 2020 1 commit
-
-
Chen Chen authored
PiperOrigin-RevId: 346397839
-
- 01 Dec, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 345054031
-
- 30 Nov, 2020 1 commit
-
-
Chen Chen authored
PiperOrigin-RevId: 344836089
-
- 17 Nov, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 342910130
-
- 13 Nov, 2020 1 commit
-
-
Le Hou authored
PiperOrigin-RevId: 342179055
-
- 12 Nov, 2020 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 342098346
-
- 11 Nov, 2020 1 commit
-
-
Hongkun Yu authored
When loss is NaN, the weights should also be NaN. PiperOrigin-RevId: 341886095
-
- 30 Oct, 2020 1 commit
-
-
Pankaj Kanwar authored
PiperOrigin-RevId: 339927293
-
- 28 Oct, 2020 2 commits
-
-
Yeqing Li authored
PiperOrigin-RevId: 339504167
-
Hongkun Yu authored
PiperOrigin-RevId: 339488247
-
- 26 Oct, 2020 1 commit
-
-
Will Cromar authored
PiperOrigin-RevId: 339131213
-
- 19 Oct, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 337927475
-
- 15 Oct, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 337242735
-
- 13 Oct, 2020 3 commits
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 336973004
-
Hongkun Yu authored
PiperOrigin-RevId: 336960641
-
Hongkun Yu authored
PiperOrigin-RevId: 336795303
-
- 09 Oct, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 336339472
-
- 08 Oct, 2020 2 commits
-
-
Hongkun Yu authored
Factor out optimizer creation logic from the trainer, so as to avoid access runtime config inside Trainer. Add validation method on the ExperimentConfig as clients may use other containers instead of the model garden config dataclass. PiperOrigin-RevId: 336132223
-
Chen Chen authored
PiperOrigin-RevId: 336042908
-
- 02 Oct, 2020 1 commit
-
-
Hongkun Yu authored
Remove type annotation for params. Make the trainer consume both model and task. PiperOrigin-RevId: 334964198
-
- 16 Sep, 2020 1 commit
-
-
Chen Chen authored
Support DataConfig.input_path to be a list of files, and remove the `shards` argument from InputReader() PiperOrigin-RevId: 331902956
-
- 15 Sep, 2020 2 commits
-
-
Hongkun Yu authored
PiperOrigin-RevId: 331861386
-
Hongkun Yu authored
PiperOrigin-RevId: 331857334
-
- 09 Sep, 2020 1 commit
-
-
Rajagopal Ananthanarayanan authored
PiperOrigin-RevId: 330660757
-
- 08 Sep, 2020 1 commit
-
-
Hongkun Yu authored
PiperOrigin-RevId: 330583627
-
- 27 Aug, 2020 1 commit
-
-
Ruoxin Sang authored
PiperOrigin-RevId: 328789639
-
- 19 Aug, 2020 1 commit
-
-
A. Unique TensorFlower authored
PiperOrigin-RevId: 327459481
-