GenRecal: Generation after Recalibration from Large to Small Vision-Language Models
This addresses the problem of computational inefficiency for deploying VLMs in real-world scenarios, offering a general-purpose solution for model distillation across heterogeneous architectures, though it is incremental in advancing existing distillation methods.
The paper tackles the challenge of deploying large vision-language models (VLMs) on resource-constrained devices by proposing GenRecal, a distillation framework that transfers knowledge from large to small VLMs across diverse architectures, achieving significant performance improvements over baseline and large-scale models in experiments.
Recent advancements in vision-language models (VLMs) have leveraged large language models (LLMs) to achieve performance on par with closed-source systems like GPT-4V. However, deploying these models in real-world scenarios, particularly on resource-constrained devices, remains challenging due to their substantial computational demands. This has spurred interest in distilling knowledge from large VLMs into smaller, more efficient counterparts. A key challenge arises here from the diversity of VLM architectures, which are built on different LLMs and employ varying token types-differing in vocabulary size, token splits, and token index ordering. To address this challenge of limitation to a specific VLM type, we present Generation after Recalibration (GenRecal), a general-purpose distillation framework for VLMs. GenRecal incorporates a Recalibrator that aligns and adapts feature representations between heterogeneous VLMs, enabling effective knowledge transfer across different types of VLMs. Through extensive experiments on multiple challenging benchmarks, we demonstrate that GenRecal significantly improves baseline performances, eventually outperforming large-scale open- and closed-source VLMs.