# Ternarization of Vision Language Models for use on edge devices

arXiv: https://arxiv.org/abs/2504.06298 · HTML: https://mlsanity.com/paper/2504.06298/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Visual token pruning and efficiency, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 14 of its 21 references that are in the map.
- First posted on arXiv 7 Apr 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Visual token pruning and efficiency](https://mlsanity.com/mllm/visual-token-pruning-and-efficiency/)

## Builds on (most cited first, 14 of 14)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone](https://mlsanity.com/paper/2404.14219/) (2024)
- [A Survey of Quantization Methods for Efficient Neural Network Inference](https://mlsanity.com/paper/2103.13630/) (2021)
- [PaliGemma: A versatile 3B VLM for transfer](https://mlsanity.com/paper/2407.07726/) (2024)
- [Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks](https://mlsanity.com/paper/2311.06242/) (2023)
- [The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits](https://mlsanity.com/paper/2402.17764/) (2024)
- [Model compression via distillation and quantization](https://mlsanity.com/paper/1802.05668/) (2018)
- [The Remarkable Robustness of LLMs: Stages of Inference?](https://mlsanity.com/paper/2406.19384/) (2024)
- [1.58-bit FLUX](https://mlsanity.com/paper/2412.18653/) (2024)
- [Simultaneously Optimizing Weight and Quantizer of Ternary Neural Network using Truncated Gaussian Approximation](https://mlsanity.com/paper/1810.01018/) (2018)
- [Optimize Deep Convolutional Neural Network with Ternarized Weights and High Accuracy](https://mlsanity.com/paper/1807.07948/) (2018)

## Abstract (arXiv)

We propose a process to compress a pre-trained Vision Language Model into a ternary version of itself instead of training a ternary model from scratch. A new initialization scheme from pre-trained weights based on the k-means algorithm is proposed to reduce the ternarization time. We implement different custom operators for executing the ternary model on the TensorFlow Lite Engine. We compare the original model with its ternary and binary versions in terms of memory consumption, inference speed and perplexity. We find that the ternary model using our custom ternary matrix multiplication operator provides a good compromise in term of memory usage and perplexity, while having the fastest token generation speed.