Commits · 20ed8cf0ce2a5bc3e2244faac5666e7c805d9e75 · ModelZoo / ResNet50_tensorflow

09 Apr, 2021 1 commit

Remove dynamic_loss_scale argument to define_performance. · 20ed8cf0

Reed Wanderman-Milne authored Apr 09, 2021

All models which support loss scaling support dynamic loss scaling, so the argument has no purpose. It used to be that some models scaled the loss manually instead of using a LossScaleOptimizer, and so did not support dynamic loss scaling.

PiperOrigin-RevId: 367719521

20ed8cf0

06 Apr, 2021 1 commit

Use tf.compat.v1 version of enable_mixed_precision_graph_rewrite. · 920defcc

Reed Wanderman-Milne authored Apr 06, 2021

The function `tf.train.experimental.enable_mixed_precision_graph_rewrite` will be removed from the TF2 namespace soon, at which point it will only be accessible under tf.compat.v1.

PiperOrigin-RevId: 367046393

920defcc

23 Mar, 2021 1 commit
- Parallelize the batching of the Shakespeare benchmark. · 0ffd6ab0
  Jay Shi authored Mar 23, 2021
```
PiperOrigin-RevId: 364541922
```
  0ffd6ab0
03 Mar, 2021 1 commit

Pass use_experimental_api=True to mixed precision functions. · 4c2f424f

Reed Wanderman-Milne authored Mar 03, 2021

The default is True, but I plan on changing it to False soon. After that, I plan on removing the argument and never using the experimental API.

PiperOrigin-RevId: 360724698

4c2f424f

13 Sep, 2020 1 commit
- [Clean up] Consolidate distribution utils. · faf7a4f5
  Hongkun Yu authored Sep 12, 2020
```
PiperOrigin-RevId: 331359058
```
  faf7a4f5
10 Sep, 2020 1 commit
- Fix syntax issue causing input shape to not be set properly for resnet. · b792005c
  Alan Chiao authored Sep 09, 2020
```
PiperOrigin-RevId: 330857499
```
  b792005c
09 Sep, 2020 1 commit
- Internal change · b47e5fe9
  Alan Chiao authored Sep 09, 2020
```
PiperOrigin-RevId: 330777498
```
  b47e5fe9
26 Aug, 2020 5 commits
- Address comments from Aug 7 · a87bb185
  Ruomei Yan authored Aug 17, 2020
  
  a87bb185
- Address review comments from mid June · 20101930
  Ruomei Yan authored Jul 28, 2020
  
  20101930
- Fix the two types of failed tests: · 7dfef01d
  Ruomei Yan authored Jun 08, 2020
```
1. the dataset_num_private_threads flags
2. clustering does not support fp16 or mixed precision training
```
  7dfef01d
- Add benchmark tests and unit tests for clustering mobilenet_v1 · 55018881
  Ruomei Yan authored Jun 04, 2020
  
  55018881
- Create an example for clustering mobilenet_v1 in resnet_imagenet_main.py · ee80adbf
  Ruomei Yan authored May 28, 2020
  
  ee80adbf
12 Aug, 2020 1 commit
- Internal change · 94561082
  Hongkun Yu authored Aug 12, 2020
```
PiperOrigin-RevId: 326286926
```
  94561082
10 Aug, 2020 1 commit
- Remove benchmark folder from the master branch. They are stale. (#9085) · c5e107ff
  Hongkun Yu authored Aug 10, 2020
  
  c5e107ff
11 Jun, 2020 1 commit
- Internal change · e5641ef5
  A. Unique TensorFlower authored Jun 10, 2020
```
PiperOrigin-RevId: 315819192
```
  e5641ef5
19 May, 2020 1 commit
- [Clean up] Remove enable_eager in the session config: Model garden is TF2 only now. · c2666cea
  Hongkun Yu authored May 19, 2020
```
Remove is_v2_0

PiperOrigin-RevId: 312336907
```
  c2666cea
14 May, 2020 1 commit
- Remove profiler callback which is calling tensorflow internal library. · ec3a4616
  Hongkun Yu authored May 13, 2020
```
Profiler has been integrated with TensorBoard callback and released in TF 2.2

PiperOrigin-RevId: 311475711
```
  ec3a4616
26 Apr, 2020 1 commit
- Remove utils/logs usage for official models. · 2c6f12e3
  Hongkun Yu authored Apr 25, 2020
```
PiperOrigin-RevId: 308451074
```
  2c6f12e3
20 Apr, 2020 1 commit
- Move more resnet_imagenet files to benchmark. · 0b0ca66b
  Allen Wang authored Apr 20, 2020
```
PiperOrigin-RevId: 307421021
```
  0b0ca66b
17 Apr, 2020 2 commits
- Move resnet_imagenet_benchmark.py to the benchmarks/models folder. · a174bf5b
  Allen Wang authored Apr 17, 2020
```
PiperOrigin-RevId: 307137297
```
  a174bf5b
- Remove hooks. · 2ddd716e
  Hongkun Yu authored Apr 16, 2020
```
PiperOrigin-RevId: 306994199
```
  2ddd716e
10 Apr, 2020 1 commit
- Internal change · a8b5cb7a
  Hongkun Yu authored Apr 10, 2020
```
PiperOrigin-RevId: 305897677
```
  a8b5cb7a
04 Apr, 2020 1 commit
- Move shakespeare to benchmark folder. · 17eee4c4
  Hongkun Yu authored Apr 03, 2020
```
PiperOrigin-RevId: 304729158
```
  17eee4c4
25 Mar, 2020 1 commit

Check if there are GPUs instead of if TF is built with CUDA support. · ad09cf49

Reed Wanderman-Milne authored Mar 25, 2020

The TF pip packages are always built with CUDA support, so tf.test.is_built_with_cuda() would return True even if the user had no GPU.

PiperOrigin-RevId: 302928378

ad09cf49

17 Mar, 2020 1 commit
- tf.compat.v1.logging implemented with absl · 3043566d
  ayushmankumar7 authored Mar 18, 2020
  
  3043566d
16 Mar, 2020 2 commits
- Internal changes. · 01748b24
  Allen Wang authored Mar 16, 2020
```
PiperOrigin-RevId: 301237793
```
  01748b24
- Move ResNet related files to the ResNet folder (e.g. utils, CTL, CFit, etc.) · ad710aa1
  Allen Wang authored Mar 16, 2020
```
PiperOrigin-RevId: 301185629
```
  ad710aa1
12 Mar, 2020 1 commit
- Internal change · e8afeaee
  Jose Baiocchi authored Mar 11, 2020
```
PiperOrigin-RevId: 300461275
```
  e8afeaee
11 Mar, 2020 1 commit
- Internal change · 7708847e
  A. Unique TensorFlower authored Mar 10, 2020
```
PiperOrigin-RevId: 300247220
```
  7708847e
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
26 Feb, 2020 1 commit
- Fix issue #8196 · 9a4b0371
  Hongkun Yu authored Feb 26, 2020
```
PiperOrigin-RevId: 297298712
```
  9a4b0371
24 Feb, 2020 1 commit
- Use learning schedule op by default in Resnet50. Remove learning rate callback code. · 02af9bb5
  Zongwei Zhou authored Feb 24, 2020
```
PiperOrigin-RevId: 296988935
```
  02af9bb5
28 Jan, 2020 1 commit
- Move resnet cifar to benchmark models. · dbdd17c4
  Hongkun Yu authored Jan 28, 2020
```
PiperOrigin-RevId: 292007895
```
  dbdd17c4
29 Aug, 2019 1 commit

Use new mixed_float16 policy for resnet. · dcd0e7ad

Reed Wanderman-Milne authored Aug 29, 2019

The old infer_float32_policies policy will be removed from TensorFlow soon.

To test convergence, I ran the Resnet50KerasAccuracy.benchmark_8_gpu_fp16 benchmark. I got an accuracy of 0.76037 and an exp_per_second of 6908.

PiperOrigin-RevId: 266191126

dcd0e7ad

22 Aug, 2019 1 commit
- Internal change · 9af989ce
  Hongkun Yu authored Aug 22, 2019
```
PiperOrigin-RevId: 264895439
```
  9af989ce