Removed kwargs

              adj_matrix: np.ndarray,

              distance_matrix: np.ndarray,

              eps: float = 1e-6,

              inf: float = 1e12,

              **kwargs):

              inf: float = 1e12):

    """Output computation for the MultiHeadedAttention layer.

    Parameters

    ----------

    ]

    x, _ = self._single_attention(query, key, value, mask, dropout_p,

                                  adj_matrix, distance_matrix, eps, inf,

                                  **kwargs)

                                  adj_matrix, distance_matrix, eps, inf)

    x = x.transpose(1, 2).contiguous().view(batch_size, -1, self.h * self.d_k)

    return self.output_linear(x)