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.

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

An example to show how to run image classification based on v2 API on cluster #2070

Closed
qingqing01 opened this issue May 9, 2017 · 2 comments
Closed

An example to show how to run image classification based on v2 API on cluster #2070

qingqing01 opened this issue May 9, 2017 · 2 comments
Assignees
@wanghaoshuang

Comments

@qingqing01
Copy link
Contributor

qingqing01 commented May 9, 2017 •
edited
Loading

这个例子是希望告诉大家如何在集群上训练大规模图像分类，当前基于MPI集群，数据基于ImageNet，集群上的数据打batch适合一些，所以建议对数据打batch。通过集群的例子，重点希望优化的是v2 API图像分类的性能，需要评估下，可能需要优化的是数据IO那块。

另外，希望通过这个例子和PaddlePaddle/models里图像分类的例子对图像预处理这块的代码形成一个比较规范的模块，当前python/paddle/utils/image_util.py 和 python/paddle/utils/image_multiproc.py 是基于DataProvider的，一些操作也不是很犯规，希望通过这两个例子，把图像预处理一些通用的操作形成规范的代码，方便用户使用。
@wanghaoshuang wanghaoshuang self-assigned this May 9, 2017
@lcy-seso
Copy link
Contributor

lcy-seso commented May 9, 2017

我们是不是也需要对 V2 API 做一些性能回归测试呢？已经有一些关于训练/测试速度变慢的问题出现。

速度一直是PaddlePaddle 的优势，性能下降，对一些需要将已经训练得非常稳定的 V1 任务切换到 V2 的用户，会带来影响。

@qingqing01 qingqing01 mentioned this issue May 10, 2017
Merged
@lcy-seso
Copy link
Contributor

lcy-seso commented Aug 27, 2017 •
edited
Loading

This has been done. So I close this issue.

heavengate pushed a commit to heavengate/Paddle that referenced this issue Aug 16, 2021
@nemonameless
[Dygraph] add SSD/SSDLite mbv1v3 (PaddlePaddle#2070)
d4b30ff 
* fix ssd ssdlite scheduler, add cosdecay

* fix mbv1v3 BatchNorm mean variance

* update ssd mbv1 voc modelzoo

* fix ssd mbv1 warmup, update modelzoo

* fix cosdecay
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@wanghaoshuang wanghaoshuang

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
4 participants
@lcy-seso @luotao1 @wanghaoshuang @qingqing01