onnxruntime

58521fb8 - Make training CUDA kernels to adhere established code structure patterns (#10735)

Make training CUDA kernels to adhere established code structure patterns (#10735) Current training optimizer kernels include CPU headers that affects changes that we can make in the CPU code with C++14 compiler and other refactoring efforts. Rearrange the kernel according to the established patterns and do not include headers that are not needed.