Skip to content
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.

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

Enhance distributed train performance #7608

Merged

Conversation

typhoonzero
Copy link
Contributor

Fix #7607

@helinwang helinwang self-requested a review January 18, 2018 04:39
Copy link
Contributor

@gongweibao gongweibao left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

可以先merge,然后我再继续整理优化代码。

// Get from multiple trainers, we don't care about the order in which
// the gradients arrives, just add suffix 0~n and merge the gradient.
rpc_service_->SetCond(0);
for (size_t i = 0; i < barrier_size; ++i) {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  • 这一块代码太长了, 导致main loop比较复杂,容易出错。建议改成单独的一个函数。

gongweibao
gongweibao previously approved these changes Jan 18, 2018
Copy link
Contributor

@gongweibao gongweibao left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

Copy link
Contributor

@gongweibao gongweibao left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM++

@typhoonzero typhoonzero merged commit 58be41f into PaddlePaddle:develop Jan 19, 2018
@typhoonzero typhoonzero deleted the distributed_split_selectedrows branch January 19, 2018 03:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants