A token classifier gives you per-token tags. Nobody wants per-token tags: they want entities.
Given tokens and tags in BIO format, return the entities as a list of (text, label, start, end) tuples:
B-X begins an entity of type XI-X continues the entity of type X immediately before itO is outside any entitytext is the tokens joined by a single spacestart and end are token indices, end exclusiveYou have read the opening. The rest is free with a Bezen account, and no card is needed.
Behind the sign-in