GitLab

PiperOrigin-RevId: 276182092

Getting rid of experimental_allow_stateful now that we have experimental_external_state_policy instead.
that option.

We've replaced experimental_allow_stateful with experimental_external_state_policy with a default of WARN in which we'll only print a warning message and not fail. As a result, we can now safely remove all client code that explicitly set the allow_stateful_ops flag to True.

PiperOrigin-RevId: 276115313

PiperOrigin-RevId: 276103495

- Support first/last summary type
- Support bert format input processing

PiperOrigin-RevId: 276005310

PiperOrigin-RevId: 275994710

PiperOrigin-RevId: 275986990

Pass the file patterns instead of flattened file paths into tf.data.Dataset.list_file. This helps Cloud DF4x4 Bert Startup from 11min34s to 6min53s.

PiperOrigin-RevId: 275953359

PiperOrigin-RevId: 275940140

PiperOrigin-RevId: 275930249

PiperOrigin-RevId: 275917359

arparse -> argparse

275538818  by Sergio Guadarrama:

    Support grayscale input images in Slim model training

--
275355841  by Sergio Guadarrama:

    Fixed cases where tf.TensorShape was constructed with float dimensions

    This is a prerequisite for making TensorShape and Dimension more strict
    about the types of their arguments.

--
275131829  by Sergio Guadarrama:

    updates mobilenet/README.md to be github compatible adds V2+ reference to mobilenet_v1.md file and fixes invalid markdown

--

PiperOrigin-RevId: 275538818

PiperOrigin-RevId: 275879424

PiperOrigin-RevId: 275867562

PiperOrigin-RevId: 275739207

PiperOrigin-RevId: 275644913

PiperOrigin-RevId: 275610155

PiperOrigin-RevId: 275603936

PiperOrigin-RevId: 275578662

Remove train_data_size flag

PiperOrigin-RevId: 275545035

PiperOrigin-RevId: 275428252

PiperOrigin-RevId: 275417626

PiperOrigin-RevId: 275408074

PiperOrigin-RevId: 275393975

PiperOrigin-RevId: 275368483

* Merged commit includes the following changes:
275131829  by Sergio Guadarrama:

    updates mobilenet/README.md to be github compatible adds V2+ reference to mobilenet_v1.md file and fixes invalid markdown

--
274908068  by Sergio Guadarrama:

    Opensource MobilenetV3 detection models.

--
274697808  by Sergio Guadarrama:

    Fixed cases where tf.TensorShape was constructed with float dimensions

    This is a prerequisite for making TensorShape and Dimension more strict
    about the types of their arguments.

--
273577462  by Sergio Guadarrama:

    Fixing `conv_defs['defaults']` override issue.

--
272801298  by Sergio Guadarrama:

    Adds links to trained models for Moblienet V3, adds a version of minimalistic mobilenet-v3 to the definitions.

--
268928503  by Sergio Guadarrama:

    Mobilenet v2 with group normalization.

--
263492735  by Sergio Guadarrama:

    Internal change...

PiperOrigin-RevId: 275330652

PiperOrigin-RevId: 275306010

PiperOrigin-RevId: 275288636

PiperOrigin-RevId: 275192365

After Eager was moved to core Tensorflow, this notebook gives the error:
AttributeError: module 'tensorflow.contrib.eager' has no attribute 'Variable'
I just fixed it.

PiperOrigin-RevId: 275142626

PiperOrigin-RevId: 275103426

PiperOrigin-RevId: 275080469

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: 274921478