-
Notifications
You must be signed in to change notification settings - Fork 5.6k
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
fix the bug in sigmoid_cross_entropy_with_logits_grad_kernel #64253
Conversation
你的PR提交成功,感谢你对开源项目的贡献! |
T up = -e_x * abs_grad * pos_weight; | ||
T term3 = up / down; | ||
|
||
T diff = term1 - label + term3; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
建议调用其他kernel实现
补充相关单侧 |
Done |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
T diff = simoid_x * pos_weight_idx - label; | ||
T term1 = (x > 0) ? static_cast<T>(1) : static_cast<T>(0); | ||
|
||
T e_x = std::exp(-std::abs(x)); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这里为什么要加 std::abs(x) ? e_x 变量不是指的e^(-x) 么?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
因为前向中的e^x是e^-abs(x),所以反向也需要保持e^-abs(x)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
…addle#64253) * fix the bug in sigmoid_cross_entropy_with_logits_grad_kernel * add the test for sigmoid_cross_entropy_with_logits_grad * modify the test * modify the grad kernel * modify the kernel
PR Category
Others
PR Types
Bug fixes
Description
sigmoid_cross_entropy_with_logits
op的反向kernel计算实现,与自动微分求导计算出的反向结果不一致。现提交PR进行修复,具体分析以及反向kernel计算方式错误仍能通过单测的分析见下面的Issue: