I'm trying to construct a Fitch-style proof for $(P \to Q) \lor (Q \to R)$ using reductio ad absurdum and the introduction and elimination rules for conjunction, disjunction, and implication. I'm not allowed to use ex falso quodlibet.
I've started by assuming the negation $\lnot((P \to Q) \lor (Q \to R))$ and got to $\lnot(P \to Q)$ and $\lnot(Q \to R)$, but have no idea what to do anymore.
This is all that I've got:
1. ¬((P → Q) v (Q → R)) [Assumption]
2. P → Q [Assumption]
3. (P → Q) V (Q → R) [2, V introduction]
4. ((P → Q) v (Q → R)) & ¬((P → Q) v (Q → R)) [1,3, & I]
5. ¬(P → Q) [2-4, - I]
6. Q → R [Assumption]
7. (P → Q) v (Q → R) [6, V I]
8. ((P → Q) v (Q → R)) & ¬((P → Q) v (Q → R)) [1,7, & I]
9. ¬(Q → R) [6-8, - I]
I've tried to proceed by assuming $Q$ and then $P$, but didn't manage to get anything that allowed to negate the initial assumption.

Hint: Assume $\neg((P\to Q)\lor(Q\to R))$ aiming to derive a contradiction. To do that assume $Q$ aiming to derive $R$.
You will, try again. $$\def\fitch#1#2{~~\begin{array}{|l} #1 \\ \hline #2\end{array}}\fitch{}{\fitch{\neg((P\to Q)\lor(Q\to R))}{\fitch{Q}{\fitch{P}{\vdots}\\\vdots\\\bot\\R}\\Q\to R\\(P\to Q)\lor(Q\to R)\\\bot}\\\neg\neg((P\to Q)\lor(Q\to R))\\(P\to Q)\vee(Q\to R)}$$
Ex falso quodlibet is derived from reductio ad absurdum . If you can derive a contradiction, you can use reductio ad absurdum on the assumption of any negation.$$\fitch{\cdots}{~~\vdots\\A\\~~\vdots\\\neg A\\\bot\\\fitch{\neg B}{\bot}\\B}$$