[CUDA][WebGPU] Add packed sparse attention indexer for continuous batching - #32618

[CUDA][WebGPU] Add packed sparse attention indexer for continuous batching#32618

kunal-vaishnavi with Copilot wants to merge 20 commits into

copilot/implement-sparse-attention-indexer-againmicrosoft/onnxruntime:copilot/implement-sparse-attention-indexer-againfrom

copilot/copilotimplement-packed-sparse-attention-indexermicrosoft/onnxruntime:copilot/copilotimplement-packed-sparse-attention-indexerCopy head branch name to clipboard