The key is to combine the uniform continuity of $f'$ with the mean-value theorem in order to deduce what is called "uniform differentiability." For any $h > 0$ we can use the MVT to find $0 < k < h$ such that
$$
\frac{f(x+h) - f(x)}{h} - f'(x) = f'(x+k) - f'(x).
$$
For $\epsilon >0$ we pick $\delta >0$ such that $z,y \in \mathbb{R}$ and $|z-y| < \delta$ implies that $|f'(z)-f'(y) | < \epsilon$. Then for $0 < h < \delta$ we use the above to estimate
$$
\left\vert \frac{f(x+h) - f(x)}{h} - f'(x) \right\vert= |f'(x+k) - f'(x)| < \epsilon
$$
since $0 < k < h < \delta$. Since $x$ is arbitrary we find that
$$
0 < h < \delta \Rightarrow \sup_{x } \left\vert \frac{f(x+h) - f(x)}{h} - f'(x) \right\vert < \epsilon.
$$
This is actually a stronger result than what you need. I'll leave it to you to fit the two together.