arXiv reaDer
LLaVA-KD: A Framework of Distilling Multimodal Large Language Models
The success of Large Language Models (LLM) has led researchers to explore Multimodal Large Language Models (MLLM) for unified visual and linguistic understanding. However, the increasing model size and computational complexity of MLLM limit their use in resource-constrained environments. Small-scale MLLM (s-MLLM) aims to retain the capabilities of the large-scale model (l-MLLM) while reducing computational demands, but resulting in a significant decline in performance. To address the aforementioned issues, we propose a novel LLaVA-KD framework to transfer knowledge from l-MLLM to s-MLLM. Specifically, we introduce Multimodal Distillation (MDist) to minimize the divergence between the visual-textual output distributions of l-MLLM and s-MLLM, and Relation Distillation (RDist) to transfer l-MLLM's ability to model correlations between visual features. Additionally, we propose a three-stage training scheme to fully exploit the potential of s-MLLM: 1) Distilled Pre-Training to align visual-textual representations, 2) Supervised Fine-Tuning to equip the model with multimodal understanding, and 3) Distilled Fine-Tuning to further transfer l-MLLM capabilities. Our approach significantly improves performance without altering the small model's architecture. Extensive experiments and ablation studies validate the effectiveness of each proposed component. Code will be available at https://github.com/Fantasyele/LLaVA-KD.
updated: Fri Oct 25 2024 06:19:13 GMT+0000 (UTC)
published: Mon Oct 21 2024 17:41:28 GMT+0000 (UTC)
参考文献 (このサイトで利用可能なもの) / References (only if available on this site)
被参照文献 (このサイトで利用可能なものを新しい順に) / Citations (only if available on this site, in order of most recent)
Amazon.co.jpアソシエイト