Handle empty tensors in CUDA InstanceNormalization

[apsonawane]

This pull request improves the handling of empty input tensors in the CUDA implementation of the InstanceNormalization operator and adds corresponding unit tests to ensure correct behavior. The main focus is to ensure compliance with the ONNX specification, which allows empty inputs and expects empty outputs.

CUDA Implementation Improvements:

• Updated InstanceNorm<T>::ComputeInternal and InstanceNorm<MLFloat16>::ComputeInternal in instance_norm.cc to check for empty input tensors and return early, ensuring that empty inputs produce empty outputs as required by ONNX. This prevents unnecessary computation and potential errors when the input size is zero. [1] [2]

Unit Test Additions:

• Added two CUDA-specific tests in instance_norm_op_test.cc:
  • InstanceNormEmptyChannel_Cuda verifies correct handling when the channel dimension is zero.
  • InstanceNormEmptyBatch_Cuda verifies correct handling when the batch dimension is zero.
    These tests confirm that the operator returns empty outputs without error for these edge cases.

[Copilot]

Pull request overview

This PR updates the CUDA InstanceNormalization kernel to correctly handle valid ONNX cases where the input tensor has zero elements (any dimension is 0), ensuring it returns an empty output without invoking cuDNN paths that would otherwise hit divide-by-zero/invalid descriptor scenarios. It also adds CUDA-targeted unit tests to cover zero-channel and zero-batch edge cases.

Changes:

• Add early-return guards in CUDA InstanceNorm<T>::ComputeInternal (including MLFloat16 specialization) for x_shape.Size() == 0 after input validation.
• Add CUDA-only tests to verify empty outputs are produced (and no error occurs) when N==0 or C==0.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
onnxruntime/core/providers/cuda/nn/instance_norm.cc	Early-return on empty input tensors (after ValidateInputs) to avoid invalid cuDNN/statistics paths and produce empty outputs.
onnxruntime/test/providers/cpu/nn/instance_norm_op_test.cc	Adds CUDA-scoped tests for empty channel and empty batch cases to validate correct empty-output behavior.