torch_npu.npu_normalize_batch

torch_npu.npu_normalize_batch(self, seq_len, normalize_type=0) -> Tensor

执行批量归一化。