Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis

TL;DR

This paper thoroughly explores the design space of combining large language models with diffusion transformers for text-to-image synthesis, providing empirical comparisons, analysis, and practical training guidelines.

Contribution

It offers a detailed empirical study and reproducible training recipes for deep fusion of LLMs and DiTs in multi-modal generation, addressing gaps in previous research.

Findings

Controlled comparisons with baseline methods

Analysis of key design choices

Reproducible training recipes

Abstract

This paper does not describe a new method; instead, it provides a thorough exploration of an important yet understudied design space related to recent advances in text-to-image synthesis -- specifically, the deep fusion of large language models (LLMs) and diffusion transformers (DiTs) for multi-modal generation. Previous studies mainly focused on overall system performance rather than detailed comparisons with alternative methods, and key design details and training recipes were often left undisclosed. These gaps create uncertainty about the real potential of this approach. To fill these gaps, we conduct an empirical study on text-to-image generation, performing controlled comparisons with established baselines, analyzing important design choices, and providing a clear, reproducible recipe for training at scale. We hope this work offers meaningful data points and practical guidelines for future research in multi-modal generation.