My issue and questions isn't so much languages but character sets, expanding say an ascii dataset to include all the non-ascii language glyphs or other unicode

My issue and questions isn't so much languages but character sets, expanding say an ascii dataset to include all the non-ascii language glyphs or other unicode characters has got to balloon the parameter count