Commits · d945e99f666a096ecaf57fd6e2d556444de1dd8a · ModelZoo / ResNet50_tensorflow

"tests/gptj/test_modeling_tf_gptj.py" did not exist on "9c83b96e627427aeeada7a84bae026cafc17ccd2"

24 Dec, 2018 2 commits
- fix some lint errors. · d945e99f
  Toby Boyd authored Dec 23, 2018
  
  d945e99f
- fixed lint issues. · 863a44ce
  Toby Boyd authored Dec 23, 2018
  
  863a44ce
23 Dec, 2018 2 commits

Add return stats after run for cifar · 61ec2907

Toby Boyd authored Dec 23, 2018

- kears_cifar_main returns dict of stats after run.
- unit tests added for keras_common.
- lint issues fixed in all 4 files.

61ec2907

decay changed 2e-4 and batch_norm_decay .997 · 8d5c1684
Toby Boyd authored Dec 23, 2018
```
 - Average of 5 runs single gpu is 93%+ range 92.5 to 93.4 top_1
```
8d5c1684

22 Dec, 2018 2 commits
- Fix a comment · b83f1e88
  Shining Sun authored Dec 22, 2018
  
  b83f1e88
- Merged commits from Toby · 59afe5d8
  Shining Sun authored Dec 22, 2018
  
  59afe5d8
21 Dec, 2018 4 commits
- bug fixes · c923a420
  Shining Sun authored Dec 21, 2018
  
  c923a420
- In two runs 92% to 92.5%. Added decay on everything. · 59a20a31
  Toby Boyd authored Dec 21, 2018
  
  59a20a31
- Change test to use run() to match API change. · ba39a3db
  Toby Boyd authored Dec 21, 2018
  
  ba39a3db
- Synth data for cifar · 80dcd27c
  Shining Sun authored Dec 21, 2018
  
  80dcd27c
20 Dec, 2018 5 commits
- Synth data works · 2f8481da
  Shining Sun authored Dec 20, 2018
  
  2f8481da
- Added some alternative code. About to do dataset · db80d57a
  Shining Sun authored Dec 20, 2018
  
  db80d57a
- Add skip_eval flag and change to new optimizer · 51fc02ae
  Shining Sun authored Dec 20, 2018
  
  51fc02ae
- bug fixes and clean ups · 6f881f77
  Shining Sun authored Dec 20, 2018
  
  6f881f77
- Before all the data related change · 424c2045
  Shining Sun authored Dec 19, 2018
  
  424c2045
19 Dec, 2018 2 commits
- Commit before pull · 26a2c3b9
  Shining Sun authored Dec 19, 2018
  
  26a2c3b9
- Clean up keras_resnet_model and rename it to resnet56 · 42420ce8
  Shining Sun authored Dec 19, 2018
  
  42420ce8
18 Dec, 2018 2 commits
- remove debug code · cd034c8c
  Shining Sun authored Dec 18, 2018
  
  cd034c8c
- set learning rate for eager mode · 84266860
  Shining Sun authored Dec 18, 2018
  
  84266860
15 Dec, 2018 1 commit
- bug fix · b9e30b11
  Shining Sun authored Dec 15, 2018
  
  b9e30b11
14 Dec, 2018 5 commits
- bug fix · 6b673d03
  Shining Sun authored Dec 14, 2018
  
  6b673d03
- bug fixes · cf723306
  Shining Sun authored Dec 14, 2018
  
  cf723306
- first attempt to extract common parts of two main files · 3fd9c7fe
  Shining Sun authored Dec 14, 2018
  
  3fd9c7fe
- fix the dataset flag · eaf2bd1b
  Shining Sun authored Dec 14, 2018
  
  eaf2bd1b
- first attempt to merge the two main files · 84d3c62c
  Shining Sun authored Dec 14, 2018
  
  84d3c62c
10 Dec, 2018 1 commit
- Modify to work with perfzero framework. · 2f2a04c7
  Toby Boyd authored Dec 10, 2018
  
  2f2a04c7
08 Dec, 2018 2 commits
- Add option to turn on eager in imagenet · c254938d
  Priya Gupta authored Dec 08, 2018
  
  c254938d
- Add a copy of Jing's ResNet50 model and use that in imagenet_main · aa1d6176
  Priya Gupta authored Dec 08, 2018
  
  aa1d6176
07 Dec, 2018 4 commits
- Add reporting of stats and turn off dist_strat. · 1b3c9ba6
  Toby Boyd authored Dec 07, 2018
  
  1b3c9ba6
- Initial benchmark script idea. · 834cff94
  Toby Boyd authored Dec 07, 2018
  
  834cff94
- Add validation every epoch. Remove random seed setting. Change logging frequency · 3060b534
  Priya Gupta authored Dec 07, 2018
  
  3060b534
- Add validation per epoch · e17e64bb
  Priya Gupta authored Dec 07, 2018
  
  e17e64bb
05 Dec, 2018 1 commit
- Add option to run eager; make 1 GPU case run without DS · 59788849
  Priya Gupta authored Dec 05, 2018
  
  59788849
03 Dec, 2018 2 commits
- use v2 optimizer and dlr · b9d2b1bb
  Priya Gupta authored Dec 03, 2018
  
  b9d2b1bb
- Change LR schedule to adjust according to batch size · e932712b
  Priya Gupta authored Dec 03, 2018
  
  e932712b
30 Nov, 2018 4 commits
- Setting the initial learning rate. · 746a927c
  Amit Patankar authored Nov 30, 2018
  
  746a927c
- Setting the CIFAR parameters for the appropriate learning rate and num gpus. · 1cd1999e
  Amit Patankar authored Nov 30, 2018
  
  1cd1999e
- Updating the base learning rate to 3.2 · 3570932e
  Amit Patankar authored Nov 30, 2018
  
  3570932e
- Experimenting with cifar keras. · b2af31a9
  Amit Patankar authored Nov 30, 2018
  
  b2af31a9
27 Nov, 2018 1 commit
- Fix steps per epoch and tensorboard logging · 9f34ac99
  Priya Gupta authored Nov 27, 2018
  
  9f34ac99