Towards Robust Instruction Tuning on Multimodal Large Language Models
This work addresses the bottleneck of manual data creation for multimodal instruction tuning, offering an incremental improvement in efficiency for researchers and practitioners in AI.
The paper tackles the problem of reducing human labor in generating high-quality instruction-following data for multimodal large language models by introducing INSTRAUG, an automatic instruction augmentation method that expands datasets by 30 times and significantly improves model alignment across 12 tasks, equivalent to scaling up training data multiple times.
Fine-tuning large language models (LLMs) on multi-task instruction-following data has been proven to be a powerful learning paradigm for improving their zero-shot capabilities on new tasks. Recent works about high-quality instruction-following data generation and selection require amounts of human labor to conceive model-understandable instructions for the given tasks and carefully filter the LLM-generated data. In this work, we introduce an automatic instruction augmentation method named INSTRAUG in multimodal tasks. It starts from a handful of basic and straightforward meta instructions but can expand an instruction-following dataset by 30 times. Results on two popular multimodal instructionfollowing benchmarks MULTIINSTRUCT and InstructBLIP show that INSTRAUG can significantly improve the alignment of multimodal large language models (MLLMs) across 12 multimodal tasks, which is even equivalent to the benefits of scaling up training data multiple times.