Multi IF

Dataset Summary We introduce Multi-IF, a new benchmark designed to assess LLMs' proficiency in following multi-turn and multilingual instructions. Multi-IF, which utilizes a hybrid framework combining LLM and human annotators, expands upon the IFEval by incorporating multi-turn sequences and translating the English prompts into another 7 languages, resulting in a dataset of 4501 multilingual conversations, where each has three turns. Our evaluation of 14 state-of-the-art LLMs on… Карточка открытого датасета. Файлы на площадку не копируются. Только некоммерческое использование. Продажа на площадке запрещена. Автор: facebook. Лицензия: cc-by-nc-2.0.