stable diffusion with dynamic width, height

[Suprhimp]

Feature request

If optimum-neuron can run with dynamic width, height in stable diffusion, there will be no reason to use NVIDIA GPU I think..

But I know it is hard job(or impossible job) to do. I want to read and modify code to achieve this.

Is there any hint or is there any possibility to use stable diffusion with dynamic pixel ratio, or just array of fixed pixel ratio?

I'm very looking forward this..

Motivation

Again, If this is possible, there is no reason to use stable diffusion in normal GPU.

Your contribution

I can make an test code, and I'd love to contribute if dynamic pixel ratio compilation is possible.

[HuggingFaceDocBuilderDev]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

[HuggingFaceDocBuilderDev]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

[HuggingFaceDocBuilderDev]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread. Thank you!

[github-actions]

This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.

[github-actions]

This issue was closed because it has been stalled for 5 days with no activity.