I am familiar with the Generalized singular value decomposition (GSVD). In GSVD, given matrices $A_1 \in \mathbb{C}^{m_1 \times n}$ and $A_2 \in \mathbb{C}^{m_2 \times n}$ it is possible to decompose the two matrices such that \begin{align*} A_1 &= U_1 \Sigma_1 Q^H, \\ A_2 &= U_2 \Sigma_2 Q^H. \end{align*} Is it possible to generalize further with $M$ matrices? i.e. given $A_i$, $i \in 1, 2, \cdots, M$, decompose such that: \begin{align*} A_1 &= U_1 \Sigma_1 Q^H, \\ A_2 &= U_2 \Sigma_2 Q^H, \\ & \vdots \\ A_M &= U_M \Sigma_M Q^H. \end{align*}
I have done a few web and Google Scholar searches, but I may be using the wrong keywords.
Bonus points if you can recommend a way to do this using Python. :)