CoIN: A Benchmark of Continual Instruction Tuning for Multimodal Large Language Models

Open in new window