Skip to content

[Op] Implement GPU version of KvResourceSparseApplyAdam. #535

New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

fuhailin
Copy link
Collaborator

@fuhailin fuhailin commented Nov 9, 2022

No description provided.


const Device& device = ctx->eigen_device<Device>();
OP_REQUIRES_OK(ctx,
functor::KvSparseApplyAdamAsync<Device, T, Tindex, Tstep>()(
Copy link
Member

Choose a reason for hiding this comment

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

这里直接调用AdamAsync的functor吗?

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

是的,我注意到AdamAsync的functor中逻辑代码与Adam相同,只是AdamAsync多了apply_sparse_rmsprop参数,于是考虑复用functor逻辑代码,不同在于将Adam 中调用的functor 中apply_sparse_rmsprop 置为了false

Copy link
Member

Choose a reason for hiding this comment

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

嗯,我建议新写一个functor::KvSparseApplyAdam,而不是复用这个,避免别人改AdamAsync的functor而影响到这里的逻辑正确性

# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants