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

num_samples_processed of learning rate annealing #4305

Closed
yinyunfeng opened this issue Sep 21, 2017 · 2 comments
Closed

num_samples_processed of learning rate annealing #4305

yinyunfeng opened this issue Sep 21, 2017 · 2 comments
Labels
User 用于标记用户问题

Comments

@yinyunfeng
Copy link

有个疑问, 当多节点训练时,学习率退火中的num_samples_processed是一个节点的处理的样本数还是所有节点处理的样本数之和?

@lcy-seso lcy-seso added the User 用于标记用户问题 label Sep 21, 2017
@typhoonzero
Copy link
Contributor

您好,欢迎关注PaddlePaddle!

多节点训练时,paddle在参数优化时,主要有两种方式:

  • 发送梯度到pserver进行参数优化
  • trainer本地参数优化,然后将参数的diff发送到pserver

第一种方式目前应该只有momentum是这样,支持learning_rate decay的参数优化方式目前都是本地计算diff发送。所以num_samples_processed是一个节点的。

参考:

#2347

https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/trainer/RemoteParameterUpdater.cpp#L204

@Yancey1989
Copy link
Contributor

由于此issue长时间没有更新,所以先关掉了,如有问题欢迎随时打开。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
User 用于标记用户问题
Projects
None yet
Development

No branches or pull requests

4 participants