Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model
We introduce Transfusion, a recipe for training a multi-modal model over discrete and continuous data. Transfusion combines the language modeling loss function (next token prediction) with diffusion to train a single transformer over mixed-modality sequences. We pretrain multiple Transfusion models...
Saved in:
Main Authors | , , , , , , , , , |
---|---|
Format | Journal Article |
Language | English |
Published |
20.08.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | We introduce Transfusion, a recipe for training a multi-modal model over
discrete and continuous data. Transfusion combines the language modeling loss
function (next token prediction) with diffusion to train a single transformer
over mixed-modality sequences. We pretrain multiple Transfusion models up to 7B
parameters from scratch on a mixture of text and image data, establishing
scaling laws with respect to a variety of uni- and cross-modal benchmarks. Our
experiments show that Transfusion scales significantly better than quantizing
images and training a language model over discrete image tokens. By introducing
modality-specific encoding and decoding layers, we can further improve the
performance of Transfusion models, and even compress each image to just 16
patches. We further demonstrate that scaling our Transfusion recipe to 7B
parameters and 2T multi-modal tokens produces a model that can generate images
and text on a par with similar scale diffusion models and language models,
reaping the benefits of both worlds. |
---|---|
DOI: | 10.48550/arxiv.2408.11039 |