GitLab

Some changes specific to prediction. Removing traces of expected results, as this is just prediction.

with serving signature that accepts JPEG image bytes instead
of a fixed size [HxWxC] image tensor.

Passing JPEG image bytes is easier for inference/serving use
cases. The model internally resizes/crops the JPEG image to
required [HxWxC] tensor before passing it on for actual model
inference.

This change aligns with Cloud TPU/ResNet-50 model that offers a
similar interface (jpeg bytes) for inferencing here:

https://github.com/tensorflow/tpu/tree/master/models/official/resnet

NOTE: This flag is set to `True` by default for ImageNet, and is
disallowed for CIFAR (as it does not apply to CIFAR).

Right now we don't have input data for prediction. So using top 10
entries of test data as input.

* bug fixes and add seed

* more random corrections

* make cleanup more robust

* return cleanup fn

* delint and address PR comments.

* delint and fix tests

* delinting is never done

* add pipeline hashing

* delint

Sometimes it takes longer than 15 seconds, and even longer than 1 minute, to spawn and create the alive file.

if-else to if only

Spaces after equal signs as well

Spaces after commas

General guidelines for Google Python Style Guide links to a deprecated style guide from google

* Fix spurious "did not start correctly" error.

The error "Generation subprocess did not start correctly" would occur if the async process started up after the main process checked for the subproc_alive file.

* Add error message

When constructing the evaluation records, data_async_generation.py would copy the records into the final directory. The main process would wait until the eval records existed. However, the main process would sometimes read the eval records before they were fully copied, causing a DataLossError.

Bypassing Export model step, if training on TPU's. As this need inference to be supported on TPU's. Remove this check once inference is supported. (#5209)

Bypassing Export model step, if training on TPU's. As this need inference to be supported on TPU's. Remove this check once inference is supported.

* Add distribution strategy to keras benchmark

* Fix comments

* Fix lints

* Fix bug on distributed training in mnist using MirroredStrategy API

* Remove unnecessary codes and chagne distribution strategy source
- Remove multi-gpu
- Remove TowerOptimizer
- Change from MirroredStrategy to distribution_utils.get_distribution_strategy

* Make ResNet robust to the case that epochs_between_evals does not divide train_epochs, and add an --eval_only option

* add some comments to make the control flow easier to follow

* address PR comments

* Add 5 epoch warmup

* get_lr with warm_up only for imagenet

* Add base_lr, remove fp16 unittest arg validation

* Remove validation check stopping v1 and FP16

* Add top_5 to to eval.

* labels shape to [?] from [?,1] matches unittest.

* Fix convergence issues for MLPerf.

Thank you to @robieta for helping me find these issues, and for providng an algorithm for the `get_hit_rate_and_ndcg_mlperf` function.

This change causes every forked process to set a new seed, so that forked processes do not generate the same set of random numbers. This improves evaluation hit rates.

Additionally, it adds a flag, --ml_perf, that makes further changes so that the evaluation hit rate can match the MLPerf reference implementation.

I ran 4 times with --ml_perf and 4 times without. Without --ml_perf, the highest hit rates achieved by each run were 0.6278, 0.6287, 0.6289, and 0.6241. With --ml_perf, the highest hit rates were 0.6353, 0.6356, 0.6367, and 0.6353.

* fix lint error

* Fix failing test

* Address @robieta's feedback

* Address more feedback

* perform a codecs check and remove unicode \ufeff if utf-8 is not present

* delint

This is done by using a higher Pickle protocol version, which the Python docs describe as being "slightly more efficient". This reduces the file write time at the beginning from 2 1/2 minutes to 5 seconds.

* Deterministic dataset order fix

In order for the order of the files to be deterministic, in `tf.data.Dataset.list_files(..., shuffle)`, shuffle needs to be True, otherwise different iterator inits will yield different file orders

* removed unnecessary shuffle of filenames

* Removed the `_FILE_SHUFFLE_BUFFER` definition

* Fix Transformer TPU crash in Python 2.X.

- Tensorflow raises an error when tf_inspect.getfullargspec is called on
a functools.partial in Python 2.X. This issue would be hit during the
eval stage of the Transformer TPU model. This change replaces the call
to functools.partial with a lambda to work around the issue.

* Remove unused import from transformer_main.

* Fix lint error.