onnxruntime

[js/webgpu] Fix shader compilation errors in cumsum (#18779)

### Description
This PR fixes below shader compilation errors:
```
Tint WGSL reader failure: :39:31 error: no matching overload for operator + (f32, i32)

5 candidate operators:
  operator + (T, T) -> T  where: T is abstract-float, abstract-int, f32, i32, u32 or f16
  operator + (vecN<T>, T) -> vecN<T>  where: T is abstract-float, abstract-int, f32, i32, u32 or f16
  operator + (T, vecN<T>) -> vecN<T>  where: T is abstract-float, abstract-int, f32, i32, u32 or f16
  operator + (vecN<T>, vecN<T>) -> vecN<T>  where: T is abstract-float, abstract-int, f32, i32, u32 or f16
  operator + (matNxM<T>, matNxM<T>) -> matNxM<T>  where: T is abstract-float, f32 or f16

                    sum = sum + get_inputByIndices(inputIndices);
                              ^


 - While validating [ShaderModuleDescriptor "CumSum"]
 - While calling [Device].CreateShaderModule([ShaderModuleDescriptor "CumSum"]).