Did ancient Greek scholars ever adapt Roman numerals?

Greeks did not adopt Roman numerals, like, ever. (“Roman Numerals? We taught those beef eaters everything they know!”) Where the West uses Roman numerals, Greek continues to use Greek numerals; see examples in Nick Nicholas’ answer to Is it possible to shorten the ordinal numbers in modern Greek? I’m honestly not aware of any tradition of using Roman numerals in Greek, outside of as a third level of numbering (like how English goes from 1, 2, 3 to a, b, c to i, ii, iii: Greek numerals are already equivalent to a, b, c, so Greek needs an alternative.)

Milesian numerals continued in use right through Byzantium; the transition was made direct from Milesian numerals to Arabic numerals, in the Renaissance. Late Byzantine scholars (e.g. Theodore Meliteniotes) were aware of Arabic numerals, and described them, but went on using Milesian numerals. Which made their astronomical texts rather painful to read…

Manuel Glyzonios’ Arithmetic (Λογαριαστική), published in 1568 in Venice, certainly used Arabic numbers, and the Glitzouni was how Greeks learned arithmetic for the next few centuries. (A PDF used to be online at Ελληνομνήμων, a repository of texts at Athens Uni about the history of STEM in Modern Greece. Long disappeared of course.) O Εμμανουήλ Γλυζώνιος και η Λογαριαστική του,πιο γνωστή ως Γλυτζούνι cites it as saying:

Γίγνωσκε ότι η λεγόμενη νούλα, ήντινα βάζομεν έμπροσθεν εις το 1 και λογίζεται 10, αυτό το λέγουσιν ελληνικά ουδέν και δια τούτο πληροί μόνον την θέσιν ψηφίου…..

You should know that the so-called nulla, which we put after 1 and count it as 10, is called in Hellenic [= Ancient Greek] nil (ouden), and for that reason it only takes up the position of a number.

Greek now calls zero mēden, which is a synonym of ouden.