| title | Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs |
|---|---|
| authors | Luka Ribar, Jeevan Bhoot, Douglas Orr |
| date | 2026-08-21 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2608.21134 |
| upvotes | 9 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Deploying vision-language models (VLMs) on mobile devices is challenging due to their significant memory and compute requirements. We present a framework for quantizing VLMs for efficient inference on resource-constrained hardware. Our approach combines a quantization pipeline that uses the model itself to generate training data and does not require access to the training setup, with a novel 2.7-bit-per-parameter format supporting efficient execution on Arm CPUs. We validate our approach by compressing the Llama 3.2 11B Vision Instruct model to 3.7 GB with 8-bit activations, preserving strong performance on a set of standard visual question answering tasks. |
cite
@misc{ribar2026llama,
title={Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs},
author={Luka Ribar and Jeevan Bhoot and Douglas Orr},
year={2026},
eprint={2608.21134},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2608.21134}
}