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

Incompatibility of compute_dtype with complex-valued inputs #20278

Open
jhoydis opened this issue Sep 23, 2024 · 1 comment
Open

Incompatibility of compute_dtype with complex-valued inputs #20278

jhoydis opened this issue Sep 23, 2024 · 1 comment
Assignees
Labels
type:feature The user is asking for a new feature.

Comments

@jhoydis
Copy link

jhoydis commented Sep 23, 2024

Hi,

In #19872, you introduced the possibility for layers with complex-valued inputs.

It then seems that this statement of the API Documentation is now wrong:
image

When I feed a complex-valued input tensor into a layer (as in this unit test), it is not cast to the compute_dtype, but rather kept as it is. I would somehow expect that the compute_dtype becomes complex in this case as well.

@mehtamansi29
Copy link
Collaborator

Hi @jhoydis -

Thanks for reporting this issue. Currently compute_dtype on layers automatically cast inputs to the dtype. But for complex input is not casting the same dtype. As complex values need to handle differently for casting compute_dtype.

We will dig into this issue more and update soon. Thanks..!!

@mehtamansi29 mehtamansi29 added the type:feature The user is asking for a new feature. label Sep 25, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
type:feature The user is asking for a new feature.
Projects
None yet
Development

No branches or pull requests

2 participants