no code implementations • 5 Dec 2023 • Mikhail Tikhomirov, Daniil Chernyshev
Latest instruction-tuned large language models (LLM) show great results on various tasks, however, they often face performance degradation for non-English input.
