[Numpy-discussion] numpy Sphinx extensions, non-ascii encoding?

I like the numpy docstrings format and am exploring using Sphinx with the numpy extensions to generate documentation. So far it's really great. But I have some non-ascii (UTF-8) docstrings that break the autosummary directive code and I can't quite figure out how to fix it. Can someone give me some tips here? The problem seems to be in generating the summary table. Sorry if this is the wrong forum for this... Aric