Unified Reinforcement and Imitation Learning for Vision-Language Models
This addresses the challenge of deploying large VLMs in resource-limited settings, offering a novel training approach that is incremental in improving efficiency and performance.
The paper tackles the problem of making Vision-Language Models (VLMs) more efficient for resource-constrained environments by introducing Unified Reinforcement and Imitation Learning (RIL), which combines reinforcement and imitation learning to train lightweight student VLMs that achieve significant performance gains, narrowing the gap with state-of-the-art models and surpassing them in some cases.
Vision-Language Models (VLMs) have achieved remarkable progress, yet their large scale often renders them impractical for resource-constrained environments. This paper introduces Unified Reinforcement and Imitation Learning (RIL), a novel and efficient training algorithm designed to create powerful, lightweight VLMs. RIL distinctively combines the strengths of reinforcement learning with adversarial imitation learning. This enables smaller student VLMs not only to mimic the sophisticated text generation of large teacher models but also to systematically improve their generative capabilities through reinforcement signals. Key to our imitation framework is an LLM-based discriminator that adeptly distinguishes between student and teacher outputs, complemented by guidance from multiple large teacher VLMs to ensure diverse learning. This unified learning strategy, leveraging both reinforcement and imitation, empowers student models to achieve significant performance gains, making them competitive with leading closed-source VLMs. Extensive experiments on diverse vision-language benchmarks demonstrate that RIL significantly narrows the performance gap with state-of-the-art open- and closed-source VLMs and, in several instances, surpasses them.