Skip to content

GitLab

Switch branch/tag
  1. 14 Apr, 2020 1 commit
  3. 05 Apr, 2020 1 commit
  5. 17 Mar, 2020 1 commit
  7. 05 Mar, 2020 1 commit
  9. 17 Dec, 2019 1 commit
  11. 03 Dec, 2019 1 commit
  13. 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
  15. 16 Oct, 2019 1 commit
    • Reed Wanderman-Milne's avatar
      Add support for the tf.keras.mixed_precision API in NCF · cb913691
      Reed Wanderman-Milne authored 
      To test, I did 50 fp32 runs and 50 fp16 runs. I used the following command:

python ncf_keras_main.py --dataset=ml-20m --num_gpus=1 --train_epochs=10 --clean --batch_size=99000 --learning_rate=0.00382059 --beta1=0.783529 --beta2=0.909003 --epsilon=1.45439e-7 --layers=256,256,128,64 --num_factors=64 --hr_threshold=0.635 --ml_perf --nouse_synthetic_data --data_dir ~/ncf_data_dir_python3 --model_dir ~/tmp_model_dir --keras_use_ctl

For the fp16 runs, I added --dtype=fp16. The average hit-rate for both fp16 and fp32 was 0.6365. I also did 50 runs with the mixed precision graph rewrite, and the average hit-rate was 0.6363. The difference is likely due to noise.

PiperOrigin-RevId: 275059871
      cb913691
  17. 10 Oct, 2019 2 commits
  19. 23 Sep, 2019 1 commit
  21. 10 Sep, 2019 1 commit
  23. 09 Sep, 2019 1 commit
  25. 29 Aug, 2019 1 commit
  27. 28 Aug, 2019 1 commit
  29. 19 Aug, 2019 1 commit
  31. 14 Aug, 2019 4 commits
  33. 13 Aug, 2019 3 commits
  35. 09 Aug, 2019 4 commits
  37. 07 Aug, 2019 1 commit
  39. 06 Aug, 2019 1 commit
  41. 01 Aug, 2019 1 commit
  43. 30 Jul, 2019 1 commit
  45. 29 Jul, 2019 1 commit
    • Hongjun Choi's avatar
      Merged commit includes the following changes: (#7322) · 803f833c
      Hongjun Choi authored 
      260228553  by priyag<priyag@google.com>:

    Enable transformer and NCF official model tests. Also fix some minor issues so that all tests pass with TF 1 + enable_v2_behavior.

--
260043210  by A. Unique TensorFlower<gardener@tensorflow.org>:

    Add logic to train NCF model using offline generated data.

--
259778607  by priyag<priyag@google.com>:

    Internal change

259656389  by hongkuny<hongkuny@google.com>:

    Internal change

PiperOrigin-RevId: 260228553
      803f833c
  47. 25 Jul, 2019 1 commit
  49. 24 Jul, 2019 2 commits
  51. 23 Jul, 2019 1 commit
  53. 08 Jul, 2019 1 commit
  55. 24 Jun, 2019 1 commit
  57. 21 Jun, 2019 1 commit
    • Toby Boyd's avatar
      NCF XLA and Eager tests with a refactor of resnet flags to make this cleaner. (#7067) · a68f65f8
      Toby Boyd authored 
      * XLA FP32 and first test

* More XLA benchmarks FP32.

* Add eager to NCF and refactor resnet.

* fix v2_0 calls and more flag refactor.

* Remove extra flag args.

* 90 epoch default

* add return

* remove xla not used by estimator.

* Remove duplicate run_eagerly.

* fix flag defaults.

* Remove fp16_implementation flag option.

* Remove stop early on mlperf test.

* remove unneeded args.

* load flags from keras mains.
      a68f65f8
  59. 18 Jun, 2019 1 commit