Retrieval-Augmented Multimodal Language Modeling

TL;DR

This paper introduces RA-CM3, a retrieval-augmented multimodal model that fetches relevant external information to improve text and image generation, outperforming prior models like DALL-E with less training compute.

Contribution

The paper presents RA-CM3, the first multimodal model integrating retrieval of external data for both text and image generation, enhancing scalability and performance.

Findings

RA-CM3 outperforms DALL-E and CM3 on image and caption generation tasks.

RA-CM3 requires less training compute (<30% of DALL-E).

RA-CM3 demonstrates novel capabilities like faithful image generation and in-context learning.

Abstract

Recent multimodal models such as DALL-E and CM3 have achieved remarkable progress in text-to-image and image-to-text generation. However, these models store all learned knowledge (e.g., the appearance of the Eiffel Tower) in the model parameters, requiring increasingly larger models and training data to capture more knowledge. To integrate knowledge in a more scalable and modular way, we propose a retrieval-augmented multimodal model, which enables a base multimodal model (generator) to refer to relevant text and images fetched by a retriever from external memory (e.g., documents on the web). Specifically, for the retriever, we use a pretrained CLIP, and for the generator, we train a CM3 Transformer on the LAION dataset. Our resulting model, named Retrieval-Augmented CM3 (RA-CM3), is the first multimodal model that can retrieve and generate both text and images. We show that RA-CM3 significantly outperforms baseline multimodal models such as DALL-E and CM3 on both image and caption generation tasks (12 FID and 17 CIDEr improvements on MS-COCO), while requiring much less compute for training (<30% of DALL-E). Moreover, we show that RA-CM3 exhibits novel capabilities, such as faithful image generation and multimodal in-context learning (e.g., image generation from demonstrations).