Mordal: Automated Pretrained Model Selection for Vision Language Models

TL;DR

Mordal is an automated framework that efficiently searches for the optimal vision-language model for specific tasks, significantly reducing computational costs and discovering new high-performing models.

Contribution

Introducing Mordal, the first automated multimodal model search framework that optimizes VLM selection for specific tasks with reduced computational effort.

Findings

Mordal reduces search GPU hours by up to 11.6 times.

It successfully finds the best VLM for various tasks.

Discovered new VLMs outperforming existing models.

Abstract

Incorporating multiple modalities into large language models (LLMs) is a powerful way to enhance their understanding of non-textual data, enabling them to perform multimodal tasks. Vision language models (VLMs) form the fastest growing category of multimodal models because of their many practical use cases, including in healthcare, robotics, and accessibility. Unfortunately, even though different VLMs in the literature demonstrate impressive visual capabilities in different benchmarks, they are handcrafted by human experts; there is no automated framework to create task-specific multimodal models. We introduce Mordal, an automated multimodal model search framework that efficiently finds the best VLM for a user-defined task without manual intervention. Mordal achieves this both by reducing the number of candidates to consider during the search process and by minimizing the time required to evaluate each remaining candidate. Our evaluation shows that Mordal can find the best VLM for a given problem using up to $8.9\times$--$11.6\times$ lower GPU hours than grid search. In the process of our evaluation, we have also discovered new VLMs that outperform their state-of-the-art counterparts.