By harnessing the capabilities of large language models (LLMs), recent large multimodal models (LMMs) have shown remarkable versatility in open-world multimodal understanding. Nevertheless, they are usually parameter-heavy and computation-intensive, thus hindering their applicability in resource-constrained scenarios. To this end, several lightweight LMMs have been proposed successively to maximize the capabilities under constrained scale (e.g., 3B). Despite the encouraging results achieved by these methods, most of them only focus on one or two aspects of the design space, and the key design choices that influence model capability have not yet been thoroughly investigated. In this paper, we conduct a systematic study for lightweight LMMs from the aspects of model architecture, training strategy, and training data. Based on our findings, we obtain Imp -- a family of highly capable LMMs at the 2B-4B scales. Notably, our Imp-3B model steadily outperforms all the existing lightweight LMMs of similar size, and even surpasses the state-of-the-art LMMs at the 13B scale. With low-bit quantization and resolution reduction techniques, our Imp model can be deployed on a Qualcomm Snapdragon 8Gen3 mobile chip with a high inference speed of about 13 tokens/s.

翻译：通过利用大型语言模型（LLM）的能力，近期的大型多模态模型（LMM）在开放世界的多模态理解中展现出非凡的通用性。然而，它们通常参数庞大且计算密集，从而阻碍了其在资源受限场景中的适用性。为此，研究者们相继提出了一些轻量级LMM，以在受限规模（如3B参数）下最大化其能力。尽管这些方法取得了令人鼓舞的结果，但其中大多数仅关注设计空间的一个或两个方面，且影响模型能力的关键设计选择尚未得到充分研究。本文从模型架构、训练策略和训练数据三个方面对轻量级LMM进行了系统性研究。基于我们的发现，我们得到了Imp——一个在2B-4B参数规模上具有高能力的LMM系列。值得注意的是，我们的Imp-3B模型稳定优于所有现有同规模轻量级LMM，甚至超越了13B参数规模的先进LMM。结合低位量化与分辨率降低技术，我们的Imp模型可在高通骁龙8Gen3移动芯片上部署，推理速度高达约13 tokens/s。