Commits · d70eca307e2f8a949dddad6404b5ea4ef383bf4d · ModelZoo / ResNet50_tensorflow

14 Apr, 2020 1 commit
- Move benchmark_wrappers to benchmark folder. · d70eca30
  Hongkun Yu authored Apr 14, 2020
```
PiperOrigin-RevId: 306521269
```
  d70eca30
09 Apr, 2020 1 commit
- move perfzero benchmark to benchmark/ · 7351c164
  Hongkun Yu authored Apr 09, 2020
```
PiperOrigin-RevId: 305709689
```
  7351c164
06 Apr, 2020 1 commit
- Internal change · cb845189
  Hongkun Yu authored Apr 05, 2020
```
PiperOrigin-RevId: 304965863
```
  cb845189
04 Apr, 2020 1 commit
- Move shakespeare to benchmark folder. · 17eee4c4
  Hongkun Yu authored Apr 03, 2020
```
PiperOrigin-RevId: 304729158
```
  17eee4c4
05 Mar, 2020 1 commit
- Remove force_v2_in_keras_compile. experimental_run_tf_function is no-op now. · d3d7f15f
  Hongkun Yu authored Mar 05, 2020
```
PiperOrigin-RevId: 299160422
```
  d3d7f15f
28 Nov, 2019 1 commit
- Internal change · 0effd158
  Sai Ganesh Bandiatmakuri authored Nov 27, 2019
```
PiperOrigin-RevId: 282857541
```
  0effd158
27 Nov, 2019 1 commit

Inject enable_runtime_flags into benchmarks and add train_steps to shakespeare_main. · 808b2f5e

Sai Ganesh Bandiatmakuri authored Nov 27, 2019

This will help general debugging by enabling custom execution with
--benchmark_method_flags.

E.g --benchmark_method_flags=train_steps=7 will run the benchmark for only 7
steps without modifying benchmark code.

PiperOrigin-RevId: 282799341

808b2f5e

04 Sep, 2019 1 commit
- Internal change · 154e8c46
  David Chen authored Sep 03, 2019
```
PiperOrigin-RevId: 267063328
```
  154e8c46
19 Aug, 2019 1 commit
- Bump valid max threshold to .94. · bee1bbc2
  Toby Boyd authored Aug 19, 2019
  
  bee1bbc2
13 Aug, 2019 1 commit
- Remove 8xGPU test. · 0d92f34a
  Toby Boyd authored Aug 13, 2019
  
  0d92f34a
07 Aug, 2019 1 commit
- Fix 8 GPU non-CuDNN LSTM tests. (#7405) · b68a6503
  Taylor Robie authored Aug 07, 2019
```
* fix steps for 8 gpu tests
```
  b68a6503
06 Aug, 2019 1 commit

Add non-CuDNN LSTM option and PerfZero benchmarks. (#7391) · 4d93d894

Taylor Robie authored Aug 06, 2019

* add non-CuDNN LSTM

* condition eager on v2 for no cudnn benchmark

* remove implementation specification

* add comments and adjust benchmarks

* re-add implementation=2

* update docstrings

4d93d894

25 Jul, 2019 1 commit
- Additional force_v2 tests. (#7296) · 1c509f19
  Toby Boyd authored Jul 24, 2019
  
  1c509f19
23 Jul, 2019 2 commits

Single execution path tests for ResNet50, ResNet56, NCF, and Shakespeare LSTM. (#7276) · 9d8c9aa4

Toby Boyd authored Jul 23, 2019

* Add force_run_distributed tests.

* Added enable_eager

* r/force_run_distributed/force_v2_in_keras_compile

* Adding force_v2 tests and FLAGs.

* Rename method to avoid conflict.

* Add cpu force_v2 tests.

* fix lint, wrap line.

* change to force_v2_in_keras_compile

* Update method name.

* Lower mlperf target to 0.736.

9d8c9aa4

add log_steps with faster logging for 8xGPU. (#7274) · 8390b362
Toby Boyd authored Jul 23, 2019

8390b362

18 Jul, 2019 1 commit

Refactor and add benchmarks as well as accuracy tests for GPU and CPU (#7248) · e0a2b8c3

Toby Boyd authored Jul 18, 2019

* Added benchmarks and common flags.

* Add cpu tests.

* Add tracking epoch times.

* fix transformer.

* Add examples_per_second.

* fix pylint

e0a2b8c3