training icon indicating copy to clipboard operation
training copied to clipboard

Add popper workflow for single stage detector and object detection

Open JayjeetAtGithub opened this issue 5 years ago • 6 comments

In this PR, we add a Popper workflow to the single stage detector and object detection benchmark, which makes it reproducible in a container native manner.

Through this PR,

  • We cleaned up the scripts download_dataset.sh, verify_dataset.sh and run_benchmark.sh and also added few scripts like collect_system_info.sh to collect the details of the machine where the benchmark workflow is being run, extract_dataset.sh to extract the downloaded dataset only if it is verified, report_result.sh to generate a report that complies with the MLPerf training rules. All the scripts have been moved to the scripts/ directory.

  • A wf.yml file has been added to the benchmark directories that describes the entire benchmark workflow in YML format.

  • A settings.py configuration file has also been added which includes configuration for the Docker engine to use the NVIDIA runtime, access privileges of the container, etc.

  • We also modified the README.md of the benchmarks to include information about how to reproduce the workflow using Popper.

JayjeetAtGithub avatar Mar 10 '20 19:03 JayjeetAtGithub

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

:memo: Please visit https://cla.developers.google.com/ to sign.

Once you've signed (or fixed any issues), please reply here with @googlebot I signed it! and we'll verify it.


What to do if you already signed the CLA

Individual signers
Corporate signers

ℹ️ Googlers: Go here for more info.

googlebot avatar Mar 10 '20 19:03 googlebot

@googlebot I signed it!

JayjeetAtGithub avatar Mar 11 '20 04:03 JayjeetAtGithub

We found a Contributor License Agreement for you (the sender of this pull request), but were unable to find agreements for all the commit author(s) or Co-authors. If you authored these, maybe you used a different email address in the git commits than was used to sign the CLA (login here to double check)? If these were authored by someone else, then they will need to sign a CLA as well, and confirm that they're okay with these being contributed to Google. In order to pass this check, please resolve this problem and then comment @googlebot I fixed it.. If the bot doesn't comment, it means it doesn't think anything has changed.

ℹ️ Googlers: Go here for more info.

googlebot avatar Mar 11 '20 04:03 googlebot

CLAs look good, thanks!

ℹ️ Googlers: Go here for more info.

googlebot avatar Mar 11 '20 04:03 googlebot

This PR seems to refer to 2 different benchmarks: the object_detection benchmark, and the now-retired ssd-v1 benchmark. I'm not sure whether we want to incorporate the changes to the object_detection benchmark (I don't know what "popper" is or whether the committee requested this change). If we do decide to incorporate the change for the object_detection benchmark, though, we should delete the part of the PR devoted to the retired ssd-v1 benchmark.

matthew-frank avatar Dec 02 '22 17:12 matthew-frank

@JayjeetAtGithub do you believe this (embarrassingly old) PR is still relevant? If so, can you address Matt's comments above? If not, can we close?

johntran-nv avatar Mar 16 '23 18:03 johntran-nv