Created by: wojtuss

Please, enable INT8 quantization of the fusion_gru op using oneDNN-based INT8 kernel.