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

Add senet and results docstring #149

Merged
merged 14 commits into from
Feb 21, 2022
Merged

Conversation

Aries-chen
Copy link
Contributor

@Aries-chen Aries-chen commented Feb 16, 2022

增加了 senet 的多个子模型以及他们的文档和测试结果

  • 文档显示
    image
  • 测试结果
    image
  • 添加readme.md Yes
  • 添加changelog
    image
  • 类脑009环境下,添加单卡 se_resnet50 senet154测速, oneflow 速度比 torch 略低
    image
    image

@kaijieshi7

@Aries-chen Aries-chen added enhancement New feature or request documentation Improvements or additions to documentation labels Feb 16, 2022
@ModelCreator.register_model
def senet154(pretrained=False, progress=True, **kwargs):
"""
Constructs the senet154 model trained on ImageNet2012.
Copy link
Contributor

Choose a reason for hiding this comment

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

senet154改成SENet-154会比较好

Copy link
Contributor

Choose a reason for hiding this comment

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

下面同理~

Copy link
Contributor Author

Choose a reason for hiding this comment

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

在论文中确实用 ResNet-152、 SENet-154 来比较,但是在代码中,可以参见 models 下已有的模型

  1. 代码中的函数 senet150 类比 resnet50、resnet101 等,表示该模型自身的网络, 而非有额外的添加如 se_inception_v3
    pytorch 和其他几个大佬也是如此命名,所以我认为函数命名应该是没问题的。
  2. 论文中的 SENet-154 类比 ResNet-152,使用了大写通常用于阐述
    假如你引用的只是第300行,那我认可建议,改为 Constructs the SENet-154 model trained on ImageNet2012.

Copy link
Contributor

Choose a reason for hiding this comment

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

好,我的意思就是300行改一下23333

@ModelCreator.register_model
def se_resent50(pretrained=False, progress=True, **kwargs):
"""
Constructs the se-resent50 model trained on ImageNet2012.
Copy link
Contributor

Choose a reason for hiding this comment

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

SE-ResNet50

@rentainhe
Copy link
Contributor

还需要更新一下README里的那个模型,加一个SENet进去 @Aries-chen
然后@kaijieshi7 开杰这边帮忙double check一下,我这边没有什么问题了

@kaijieshi7
Copy link
Contributor

还需要更新一下README里的那个模型,加一个SENet进去 @Aries-chen 然后@kaijieshi7 开杰这边帮忙double check一下,我这边没有什么问题了

Changelog 和readme更新一下,还有速度?

@rentainhe
Copy link
Contributor

rentainhe commented Feb 18, 2022

还需要更新一下README里的那个模型,加一个SENet进去 @Aries-chen 然后@kaijieshi7 开杰这边帮忙double check一下,我这边没有什么问题了

Changelog 和readme更新一下,还有速度?

README更新一下,速度测试加上,ChangeLog这边还没弄好2333,之后再加也行,然后记得先合并一下主分支解决冲突

@Aries-chen Aries-chen requested a review from rentainhe February 21, 2022 10:43
@Aries-chen
Copy link
Contributor Author

已更新

@rentainhe rentainhe merged commit 0edd3a1 into main Feb 21, 2022
@rentainhe rentainhe deleted the add-senet-and-results-docstring branch February 21, 2022 11:18
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
documentation Improvements or additions to documentation enhancement New feature or request New Features Priority: 0
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants