GitLab

PiperOrigin-RevId: 304908157

PiperOrigin-RevId: 301277545

PiperOrigin-RevId: 299160422

PiperOrigin-RevId: 283449673

PiperOrigin-RevId: 282650416

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

PiperOrigin-RevId: 270926016

Move movielens to recommendation

PiperOrigin-RevId: 269680664

PiperOrigin-RevId: 267486140

PiperOrigin-RevId: 267063166

PiperOrigin-RevId: 263401952

PiperOrigin-RevId: 263257133

PiperOrigin-RevId: 263204353

262988559  by A. Unique TensorFlower<gardener@tensorflow.org>:

    Enable NCF TF 2.0 model to run on TPUStrategy.

--
262971756  by A. Unique TensorFlower<gardener@tensorflow.org>:

    Internal change

262967691  by hongkuny<hongkuny@google.com>:

    Internal

--

PiperOrigin-RevId: 262988559

* Add run_eagerly for ctl.

* fix test name and do not set "default".

* force_v2_in_keras_compile FLAG default to None and added seperate temp path.

* switch to force testing 1v path not force v2 path.

* Rename function force_v1_path.

* Merged commit includes the following changes:
260800447  by A. Unique TensorFlower<gardener@tensorflow.org>:

    Fix tf function signature errors for NCF model.

--
260767064  by psv<psv@google.com>:

    Renames the `run_distributed` flag to `experimental_run_tf_function` and modifies the tests generated by the keras_all_modes decorator to better reflect that.

--

PiperOrigin-RevId: 260800447

* Update ncf_keras_benchmark.py

* Update ncf_keras_benchmark.py

* Update ncf_keras_benchmark.py

* Update ncf_keras_benchmark.py

260622966  by A. Unique TensorFlower<gardener@tensorflow.org>:

    Add absl_app import.

--

PiperOrigin-RevId: 260622966

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

* 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.

This combination does not yet work. Fail early with an explicit message instead of throwing error later on.

* Ncf perf changes 1)exclude metric layer from CTL train step 2)dataset optimization to fix size of the sample_weights, preventing a costly broadcast during loss calculation for multi-gpu case

* reduce iterations from 20 to 12.

* add fp16 dynamic batch accuracy check.

* fix existing lint issue.

* borrowing a tf1.x optimization which converts gradients from sparse to dense for better perf

* cleanup after code review

* 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.