Towards Understanding Best Practices for Quantization of Vision-Language Models
This work provides practical insights for efficient deployment of multimodal large language models, addressing memory and latency issues for practitioners, but it is incremental as it applies existing quantization techniques to a new multimodal context.
The study investigates how various quantization methods, including GPTQ and AWQ, affect the performance of multimodal vision-language models on tasks like captioning, retrieval, and question answering, finding that lower-bit quantization of the language model achieves high accuracy with reduced bits per weight.
Large language models (LLMs) deliver impressive results for a variety of tasks, but state-of-the-art systems require fast GPUs with large amounts of memory. To reduce both the memory and latency of these systems, practitioners quantize their learned parameters, typically at half precision. A growing body of research focuses on preserving the model performance with more aggressive bit widths, and some work has been done to apply these strategies to other models, like vision transformers. In our study we investigate how a variety of quantization methods, including state-of-the-art GPTQ and AWQ, can be applied effectively to multimodal pipelines comprised of vision models, language models, and their connectors. We address how performance on captioning, retrieval, and question answering can be affected by bit width, quantization method, and which portion of the pipeline the quantization is used for. Results reveal that ViT and LLM exhibit comparable importance in model performance, despite significant differences in parameter size, and that lower-bit quantization of the LLM achieves high accuracy at reduced bits per weight (bpw). These findings provide practical insights for efficient deployment of MLLMs and highlight the value of exploration for understanding component sensitivities in multimodal models. Our code is available at https://github.com/gautomdas/mmq.