Skip to content

Place histogram bucketing logic on CPU explicitly when using TPUStrategy #2883

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
nfelt merged 3 commits into from
Nov 5, 2019
Merged

Place histogram bucketing logic on CPU explicitly when using TPUStrategy #2883

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6a4d47e
Update summary_v2.py
hongjunChoi Nov 4, 2019
5b2b4a1
Update summary_v2.py
hongjunChoi Nov 4, 2019
e308949
Update summary_v2.py
hongjunChoi Nov 4, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 17 additions & 5 deletions tensorboard/plugins/histogram/summary_v2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,11 +72,23 @@ def histogram(name, data, step=None, buckets=None, description=None):
summary_scope = (
getattr(tf.summary.experimental, 'summary_scope', None) or
tf.summary.summary_scope)
with summary_scope(
name, 'histogram_summary', values=[data, buckets, step]) as (tag, _):
tensor = _buckets(data, bucket_count=buckets)
return tf.summary.write(
tag=tag, tensor=tensor, step=step, metadata=summary_metadata)

def histogram_summary(data, buckets, histogram_metadata, step):
with summary_scope(
name, 'histogram_summary', values=[data, buckets, step]) as (tag, _):
tensor = _buckets(data, bucket_count=buckets)
return tf.summary.write(
tag=tag, tensor=tensor, step=step, metadata=histogram_metadata)

# `_buckets()` has dynamic output shapes which is not supported on TPU's. As so, place
# the bucketing ops on outside compilation cluster so that the function in executed on CPU.
# TODO(https://github.com/tensorflow/tensorboard/issues/2885): Remove this special
# handling once dynamic shapes are supported on TPU's.
if isinstance(tf.distribute.get_strategy(),
tf.distribute.experimental.TPUStrategy):
return tf.compat.v1.tpu.outside_compilation(
histogram_summary, data, buckets, summary_metadata, step)
return histogram_summary(data, buckets, summary_metadata, step)


def _buckets(data, bucket_count=None):
Expand Down