Performance Improvement: Offloading IOAdapter Resize and FlowFormer++ Compute Weight to CUDA

[coca-huang]

Summary

This pull request introduces a significant performance improvement by offloading the resize operation in IOAdapter and the compute_weight operation in FlowFormer++ to CUDA, whenever CUDA is available. This change aims to reduce CPU usage substantially and leverage GPU acceleration for enhanced processing speed.

Changes

• Modified the resize method in IOAdapter to detect CUDA availability and execute on GPU when possible.
• Updated FlowFormer++'s compute_weight function to perform computations on CUDA instead of the CPU.

[hmorimitsu]

Thank you for the PR. Could you also update the compute_weight function in flowformer and matchflow models?

[hmorimitsu]

Can you also add the argument indexing="ij" in torch.meshgrid to avoid warnings?