Rename Sequence to Segments

[NickleDave]

to match vocalpy: vocalpy/vocalpy#127

This would obviously be breaking change but I think it's important to do this, to convey that these annotation formats are the result of segmentation, and to reserve the word "sequence" to mean "a sequence of sounds produced by an animal" -- a sequence is by definition made up of segments, but not all segments are part of a sequence

We could also change the actual implementation when we do so -- I need to revisit + reread the code in detail but when I skim now I get the feeling it could be simplified, e.g. we could represent a set of line segments with a dataframe and then switch the implementation of properties like onsets_s to just return the column as values instead of the setter/getter pattern we use now.