Consider calling sns.set(font_scale=1.4)
before plotting your data. This will scale all fonts in your legend and on the axes.
My plot went from this,
To this,
Of course, adjust the scaling to whatever you feel is a good setting.
Code:
sns.set(font_scale=1.4)
cmap = sns.diverging_palette(h_neg=210, h_pos=350, s=90, l=30, as_cmap=True)
sns.clustermap(data=corr, annot=True, fmt="d", cmap="Blues", annot_kws={"size": 16})