pub fn apply_binary_elementwise<F>( a: &Tensor, b: &Tensor, operation: F, _context: &str, ) -> Result<Tensor>where F: Fn(f32, f32) -> f32,
Common pattern for element-wise tensor operations with broadcasting