For random variables $X$ and $Y$, the conditional entropy of $Y$ given $X$is defined as $$H(Y|X) = - \sum_{x, y} p(x, y) \log p(y|x) = \sum_x p(x) H(Y|X=x)$$ where $$H(Y|X=x) = - \sum_y p(y|x) \log p(y|x)$$
note: $p(x, y) = p(x) \times p(y|x)$ can be used to prove the above equality.
Similarly, for $H(Y|X, Z)$, we write $$H(Y|X, Z) = \sum_z p(z) H(Y|X, Z=z)$$ where $$H(Y|X, Z=z) = - \sum_{x, y} p(x, y|z) \log p(y|x, z)$$
This is Definition 2.15 of chapter II from Information Theory and Network Coding book.
I get confused on how to prove $H(Y|X, Z)$.
This looks simple, if you look at the first proof:
$$H(Y|X) = - \sum_{x, y} p(x, y) \log p(y|x) = - \sum_{x, y} p(x) p(y|x) \log p(y|x).$$
Performing the sum over $y$, we get:
$$H(Y|X) = \sum_x p(x) H(Y|X=x)$$
Similarly, for $H(Y|X,Z)$:
$$H(Y|X, Z) = \sum_{y,x,z} p(y,x,z) \log p(y|x,z) = \sum_{y,x,z} p(z) p(y,x|z) \log p(y|x,z)$$
performing sum over $y,x$:
$$H(Y|X, Z) = \sum_z p(z) H(Y|X, Z=z)$$
Hope this cleared your doubts. If not, let's discuss in this thread.