Skip to content

【Hackathon 6th Fundable Projects 4 No.2】remove build_strategy useless attribution #6782

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

Merged
merged 1 commit into from
Jul 23, 2024
Merged

【Hackathon 6th Fundable Projects 4 No.2】remove build_strategy useless attribution #6782

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 0 additions & 18 deletions docs/api/paddle/static/BuildStrategy_cn.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,15 +64,6 @@ bool 类型。表明是否融合(fuse) relu 和 depthwise_conv2d，节省 GPU

COPY-FROM: paddle.static.BuildStrategy.fuse_relu_depthwise_conv

gradient_scale_strategy
'''''''''

``paddle.static.BuildStrategy.GradientScaleStrategy`` 类型。在 ``ParallelExecutor`` 中，存在三种定义 loss 对应梯度( *loss@grad* )的方式，分别为 ``CoeffNumDevice``, ``One`` 与 ``Customized``。默认情况下，``ParallelExecutor`` 根据设备数目来设置 *loss@grad*。如果用户需要自定义 *loss@grad*，可以选择 ``Customized`` 方法。默认为 ``CoeffNumDevice`` 。

**代码示例**

COPY-FROM: paddle.static.BuildStrategy.gradient_scale_strategy

memory_optimize
'''''''''

Expand All @@ -88,15 +79,6 @@ reduce_strategy

COPY-FROM: paddle.static.BuildStrategy.reduce_strategy

remove_unnecessary_lock
'''''''''

bool 类型。设置 True 会去除 GPU 操作中的一些锁操作，``ParallelExecutor`` 将运行得更快，默认为 True。

**代码示例**

COPY-FROM: paddle.static.BuildStrategy.remove_unnecessary_lock

sync_batch_norm
'''''''''

Expand Down