Supported datatypes in cuTENSOR

Recently I started using the cuTENSOR library to process tensors of the n-order, especially by contraction. I am very impressed by the capabilities of the library, but I am also interested in processing data types like int8, int16, fp16 besides FP32 and FP64 input. Is there a chance that these types will be supported in the (near) future? I wonder why these types are not supported so far, because the tensor cores cover exactly these data types.