Skip to content

GitLab

Switch branch/tag
  1. 20 Jul, 2020 1 commit
  3. 17 Jul, 2020 1 commit
  5. 16 Jul, 2020 1 commit
  7. 01 May, 2020 1 commit
  9. 24 Apr, 2020 1 commit
  11. 23 Apr, 2020 1 commit
  13. 21 Apr, 2020 1 commit
  15. 14 Apr, 2020 1 commit
  17. 09 Apr, 2020 1 commit
  19. 08 Apr, 2020 1 commit
  21. 05 Apr, 2020 1 commit
  23. 13 Feb, 2020 1 commit
  25. 17 Dec, 2019 1 commit
  27. 25 Nov, 2019 1 commit
    • Sai Ganesh Bandiatmakuri's avatar
      Inject enable_runtime_flags into benchmarks. · bcce419a
      Sai Ganesh Bandiatmakuri authored 
      This will help general debugging by enabling custom execution with  --benchmark_method_steps.

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

PiperOrigin-RevId: 282396875
      bcce419a
  29. 10 Oct, 2019 1 commit
  31. 28 Aug, 2019 1 commit
  33. 23 Aug, 2019 2 commits
  35. 22 Aug, 2019 1 commit
  37. 20 Aug, 2019 1 commit
  39. 19 Aug, 2019 1 commit
  41. 11 Jul, 2019 2 commits
  43. 08 Jul, 2019 1 commit
  45. 28 Jun, 2019 1 commit
  47. 20 Jun, 2019 2 commits
  49. 06 Jun, 2019 1 commit
  51. 05 Jun, 2019 2 commits
  53. 31 May, 2019 1 commit
  55. 29 May, 2019 2 commits
  57. 28 May, 2019 3 commits
    • Igor's avatar
      Make 'off' a string literal. · 3928d481
      Igor authored
      3928d481
    • guptapriya's avatar
      Turn dist strat off for 1 GPU benchmarks · 2be9ba5b
      guptapriya authored
      2be9ba5b
    • Igor's avatar
      Add distribute strategies to transformer. (#6883) · b9c1d1ca
      Igor authored 
      * Fixes that make transformer run.

* Remove debug print statements.

* Changed the permissions to 644.

* Fix the rest of the permissions.

* enable static batch in all benchmarks

* Restrict dist strat hack to training mode

For now we will do predict/eval without dist strat, so remove that hack in non training cases.

* Use `inputs` instead of `x` as arg name for call

Keras has different behavior based on whether the inputs are called `inputs` or not. Using `inputs` gives expected behaviors.

* Avoid extra map fn on input in dist strat case

* Update how we handle custom metrics

This new approach works with and without dist strat. The previous one didn't work with dist strat. We need to fix that but this is reasonable in meantime (b/133724664).

* Update benchmarks

* typo in metrics code

* Revert metrics change

Didn't actually work in distributed case..
      b9c1d1ca
  59. 24 May, 2019 1 commit
    • Toby Boyd's avatar
      Transformer v2 benchmark (#6860) · f2ea2f53
      Toby Boyd authored 
      * Moved common keras code to utils.

* Initial 1 gpu benchmark

- Aligned flags with resnet example
- removed code/features that are not super useful
- eval as part of train if bleu source/ref provided
- add exp_per_second hook

* Rename benchmark classes, pass batch-size and log_steps.

* fix docstring

* Predict done with checkpoints inline

- perfzero baseclass

* steps not epochs with smoother training loop.

* do not initialize history outside loop.

* 5000 between eval not 500

* estimator to keras.

* remove epochs var.

* use range not xrange.

* 200K steps for 1 gpu

* fix global step
      f2ea2f53