From Tokens to Numbers: Continuous Number Modeling for SVG Generation

TL;DR

This paper introduces Continuous Number Modeling (CNM), a novel approach for SVG generation that directly models numerical parameters as continuous values, improving training efficiency and visual quality over token-based methods.

Contribution

The paper proposes CNM, a continuous modeling framework for SVG parameters, and demonstrates its effectiveness through a multimodal transformer trained on 2 million samples with reinforcement learning.

Findings

Training speed increased by over 30%

Higher perceptual fidelity compared to alternatives

Effective for high-quality vector graphic generation

Abstract

For certain image generation tasks, vector graphics such as Scalable Vector Graphics (SVGs) offer clear benefits such as increased flexibility, size efficiency, and editing ease, but remain less explored than raster-based approaches. A core challenge is that the numerical, geometric parameters, which make up a large proportion of SVGs, are inefficiently encoded as long sequences of tokens. This slows training, reduces accuracy, and hurts generalization. To address these problems, we propose Continuous Number Modeling (CNM), an approach that directly models numbers as first-class, continuous values rather than discrete tokens. This formulation restores the mathematical elegance of the representation by aligning the model's inputs with the data's continuous nature, removing discretization artifacts introduced by token-based encoding. We then train a multimodal transformer on 2 million raster-to-SVG samples, followed by fine-tuning via reinforcement learning using perceptual feedback to further improve visual quality. Our approach improves training speed by over 30% while maintaining higher perceptual fidelity compared to alternative approaches. This work establishes CNM as a practical and efficient approach for high-quality vector generation, with potential for broader applications. We make our code available http://github.com/mikeogezi/CNM.