GitLab

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

change benchmark's log verbosity to logging.INFO. it seems to me that DEBUG map to ---v=1 internally, which is way to verbose for the purpose of benchmarking.

PiperOrigin-RevId: 274040907

PiperOrigin-RevId: 273966871

PiperOrigin-RevId: 270749832

PiperOrigin-RevId: 267946336

PiperOrigin-RevId: 266242786

PiperOrigin-RevId: 266002056

PiperOrigin-RevId: 263401952

PiperOrigin-RevId: 263257133

PiperOrigin-RevId: 263204353

PiperOrigin-RevId: 263217684

PiperOrigin-RevId: 263158478

* 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.

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

* top_1 to hr_at_10_max.

* Call self._run_and_report_benchmark not Super

* 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.

* 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.

* adding a new perf test for ncf, and changing some names

* Added change to make ncf use the data from the gcp bucket, and removed the need to re-download data >1day old. Reorganized the perf-zero tests

* Add CTL benchmark

* Divide train loss by number of train steps

* increase num epochs to 10

* add benchmark for early stopping with CTL

* remove whitespace

Add early stopping logic to ncf keras when desired threshold is met. Also change the default batch size to match the tuned hyperparams

Current batch size 160000 does not converge to the desired HR. So we decrease to 99k which is known to converge. Tested locally and got to 63.5 at epoch 7. Also decreasing number of epochs as I don't see any improvement after epoch 7-8.

* Set the --clone_model_in_keras_dist_strat to None.  Remove the separate no_cloning benchmarks and add a couple of cloning ones.  Fixes the learning rate schedule to cache its ops per graph.

* Add accuracy check.

* Avoid double flag init, move data_dir to real data.

* Comment on lower accuracy target.