8000 Random Seed Issue by HarborYuan · Pull Request #6354 · open-mmlab/mmdetection · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Random Seed Issue #6354

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

Jump to bottom
Closed
wants to merge 2 commits into from
Closed

Random Seed Issue #6354

wants to merge 2 commits into from

Conversation

HarborYuan
Copy link

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

#6339

Modification

#6339

BC-breaking (Optional)

Does the modification introduce changes that break the backward-compatibility of the downstream repos?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.

Checklist

  1. Pre-commit or other linting tools are used to fix the potential lint issues.
  2. The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
  3. If the modification has potential influence on downstream projects, this PR should be tested with downstream projects, like MMDet or MMCls.
  4. The documentation has been modified accordingly, like docstring or example tutorials.

@HarborYuan
Copy link
Author
HarborYuan commented Oct 31, 2021 •
edited
Loading

@hhaAndroid @RangiLyu
Thanks for your work on mmdetection.
Is there any suggestion about this PR?

@RangiLyu
Copy link
Member
RangiLyu commented Nov 1, 2021

Thanks for your contribution! We are now assessing the impact of this change.
Here are some brief suggestions:

  • When the seed is not set in the args, it will randomly generate a seed and broadcast to all ranks. How about integrating these codes into the set_random_seed function in mmdet/apis/train.py?
  • When args.seed is None, the seed in cfg will set to None. The hyperparam cfg.seed is used in some other places such as the data loader sampler. I think the args.seed should also be set to the random_num in this PR.

@HarborYuan
Copy link
Author

Thanks for your reponse.

You can make changes on the PR based on your understanding. By the way, I think there is another thing needs to be considered:
The current codes only support gpu, do we need to support multi process cpu tasks?

@RangiLyu
Copy link
Member
RangiLyu commented Nov 1, 2021

You can make changes on the PR based on your understanding.

Okay, we will take over this PR and do some modifications.

By the way, I think there is another thing needs to be considered: The current codes only support gpu, do we need to
support multi process cpu tasks?

CPU DDP or DP training support is not in our plan.

@RangiLyu RangiLyu mentioned this pull request Nov 1, 2021
Closed
24 tasks
@hhaAndroid
Copy link
Collaborator

@HarborYuan @RangiLyu I will follow this PR to continue to modify.

@hhaAndroid hhaAndroid mentioned this pull request Nov 8, 2021
Merged
@hhaAndroid hhaAndroid closed this Nov 8, 2021
@Junjun2016 Junjun2016 mentioned this pull request Nov 14, 2021
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@hhaAndroid hhaAndroid Awaiting requested review from hhaAndroid

@RangiLyu RangiLyu Awaiting requested review from RangiLyu

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@HarborYuan @RangiLyu @hhaAndroid
0