A Cheaper and Better Diffusion Language Model with Soft-Masked Noise

TL;DR

This paper introduces Masked-Diffuse LM, a novel diffusion model for language that uses soft-masking and categorical prediction to improve efficiency and generation quality over existing diffusion models for text.

Contribution

The paper proposes a linguistically-informed diffusion process with soft-masking and categorical prediction, addressing limitations of Gaussian noise in discrete language modeling.

Findings

Outperforms state-of-the-art diffusion models in text generation quality

Achieves lower training costs compared to existing diffusion-based language models

Demonstrates effectiveness across five controlled generation tasks

Abstract

Diffusion models that are based on iterative denoising have been recently proposed and leveraged in various generation tasks like image generation. Whereas, as a way inherently built for continuous data, existing diffusion models still have some limitations in modeling discrete data, e.g., languages. For example, the generally used Gaussian noise can not handle the discrete corruption well, and the objectives in continuous spaces fail to be stable for textual data in the diffusion process especially when the dimension is high. To alleviate these issues, we introduce a novel diffusion model for language modeling, Masked-Diffuse LM, with lower training cost and better performances, inspired by linguistic features in languages. Specifically, we design a linguistic-informed forward process which adds corruptions to the text through strategically soft-masking to better noise the textual data. Also, we directly predict the categorical distribution with cross-entropy loss function in every diffusion step to connect the continuous space and discrete space in a more efficient and straightforward way. Through experiments on 5 controlled generation tasks, we demonstrate that our Masked-Diffuse LM can achieve better generation quality than the state-of-the-art diffusion models with better efficiency.