1) If we do not care on the order of summands (it looks to be the case).
For each ordered pair $(e,f)$ take a sum $1+x^{ef}+x^{2ef}+\dots$, generating function is a product of these guys over all $(e,f)$. This may be further rewritten in different ways. At first, we may fix $e$, then we have $\prod_f (1-x^{ef})^{-1}=P(x^{e})$, where $P(t)=\prod (1-t^f)^{-1}=\sum p(n)t^n$ is the Euler generating function for partitions. So, the answer is $\prod_e P(x^e)$. It is appropriate to get asymptotics. We may instead fix $ef=n$, then we get $\prod_n (1-x^n)^{-\tau(n)}$, where $\tau(n)$ is a number of divisors function.
Of course, both functions are written on the OEIS page from the comment by Brendan McKay.
2) If we care on the order of summands.
For $r=1$ the number of representations is number of divisors of $n$, thus generating function equals $h(x)=\sum_{k\geqslant 1} x^k/(1-x^k)$. For arbitrary $r$ generating function if $h^r$, thus if $r$ varies, we finally get
$$
h+h^2+\dots=\frac{h}{1-h}.
$$