Skip to content

Conversation

@andersonic
Copy link
Contributor

Set chunks to be equal to len(balance) for pipe/transformer benchmark. Will update words per second and memory usage checks in next commit (must test on CircleCI to find appropriate values)

Before submitting

  • Was this discussed/approved via a Github issue? (no need for typos, doc improvements)
  • Did you read the contributor guideline?
  • Did you make sure to update the docs?
  • Did you write any new necessary tests?

What does this PR do?

Fixes #51 .

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃

Set chunks to be equal to len(balance) for pipe/transformer benchmark. Will update words per second and memory usage checks in next commit (must test on CircleCI to find appropriate values)
@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Aug 26, 2020
Did six runs for words-per-second, with results: 9144.40, 9163.91, 9993.01, 9082.82, 9155.09, 9000.67
Peak allocated bytes per device (which does not change between runs) were 193206272, 645632, 562688, 92688384 for devices 0, 1, 2 and 3, respectively
@codecov
Copy link

codecov bot commented Aug 26, 2020

Codecov Report

Merging #52 into master will not change coverage.
The diff coverage is n/a.

Impacted file tree graph

@@           Coverage Diff           @@
##           master      #52   +/-   ##
=======================================
  Coverage   94.28%   94.28%           
=======================================
  Files          35       35           
  Lines        2100     2100           
=======================================
  Hits         1980     1980           
  Misses        120      120           
Flag Coverage Δ
#Python 94.28% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.


Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5251a69...9c2d066. Read the comment docs.

@andersonic andersonic marked this pull request as ready for review August 26, 2020 19:36
Jun Ru Anderson added 2 commits August 27, 2020 10:20
batch size was small enough that the GPU's computing power was not the bottleneck, slowing training and specifically making more chunks slower. Increasing batch size has therefore increased training speed
ran six times, with wps 36917.44, 36797.65, 37006.03, 36872.84, 37129.31, 37003.31 and peak allocated bytes 4061909504, 4050944, 10427392, 2031824896 for devices 0,1,2 and 3 respectively.
@msbaines msbaines merged commit d1d7441 into master Aug 28, 2020
@msbaines msbaines deleted the add-chunks-to-pipe-benchmark branch August 28, 2020 21:34
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

bechmarks/transformer.py doesn't specify chunks when constructing Pipe

4 participants