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

【Hackathon 5th No.57】Neural networks for topology optimization #559

Closed
wants to merge 4 commits into from

Conversation

NKNaN
Copy link
Contributor

@NKNaN NKNaN commented Oct 5, 2023

PR types

Others

PR changes

Others

Describe

PaddlePaddle/Paddle#57262

未完全使用PaddleScience API,主要问题有:

  1. dataset的transform不支持随机翻转和随机旋转
  2. dataset的transform只能做用在input上,随机翻转和随机旋转需要对input和label同时作用
  3. 源代码中的dataloader在每一个batch中对stop iteration time做采样一次(记为k),截取input中channel为k和k-1的两个通道作为batch_input,目前手写实现

结果

论文源代码:
papers

复现:
ours

与论文中的指标误差:
(论文Table 1 & Table 2)
errors

@paddle-bot
Copy link

paddle-bot bot commented Oct 5, 2023

Thanks for your contribution!

@luotao1
Copy link
Collaborator

luotao1 commented Oct 9, 2023

开发之前请撰写 RFC 文档:PaddleSciece RFC 文档

@lijialin03
Copy link
Contributor

lijialin03 commented Oct 12, 2023

  1. dataset的transform不支持随机翻转和随机旋转

目前 PaddleScience 存在一个允许用户进行数据处理自定义的类FunctionalTransform,可以参考PaddleScience 官方 API 文档

  1. dataset的transform只能做用在input上,随机翻转和随机旋转需要对input和label同时作用

已进行修改,PR正在合入中 #565

  1. 源代码中的dataloader在每一个batch中对stop iteration time做采样一次(记为k),截取input中channel为k和k-1的两个通道作
    为batch_input,目前手写实现

可以为 PaddleScience 新增该功能,或按照1的方法使用,或用其他方法实现皆可

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants