Semantic structure of FOL is defined as follows: $\mathcal{M}$ is a structure of $\mathcal{L}$ if and only if $\mathcal{M} = \left(M, m_f, m_p\right)$ for some $M, m_f, m_p$ such that (1) $M \neq \emptyset$; (2) for each $n$-ary function symbol $f$ in $\mathcal{L}$, $m_f\left(f\right)$ is some map $F: M^{n} \to M$; (3) for each $n$-ary predicate symbol $p$, $m_p\left(p\right)$ is some set $S \subseteq M^{n}$. Specifically, $m_{p}\left(=\right) = \left\{\left(s, s\right) \vert s \in M\right\}$.
Let $P_{0}$ be an atomic predicate that takes zero argument. What is $m_{p}\left(P_{0}\right)$? Further, how to interpret $\mathcal{M} \vDash P_{0} \left[\sigma\right]$?