You are given a non-empty dictionary articles keyed by article id (string). Each value is a record {'tags': list[str]} where tags is a non-empty list of tag strings (a tag may repeat within one article, and each such occurrence counts).
Return the single tag that occurs the most times across all articles combined. If two or more tags share the highest count, return the one that is smallest alphabetically.
NOTE: This is a function-type question — you only write the function. Do not read input or print anything; a hidden checker calls your function and checks its return value.
Arguments
{'a1': {'tags': ['py', 'ml']}, 'a2': {'tags': ['py', 'web']}, 'a3': {'tags': ['ml', 'py']}}
Returns
'py'