Scaling Vision Transformers to 22 Billion Parameters
Abstract
The scaling of Transformers has driven breakthrough capabilities for language models. At present, the largest large language models (LLMs) contain upwards of 100B parameters. Vision Transformers (ViT) have introduced the same architecture to image and video modelling, but these have not yet been successfully scaled to nearly the same degree; the largest dense ViT contains 4B parameters (Chen et al., 2022). We present a recipe for highly efficient and stable training of a 22B-parameter ViT (ViT-22B) and perform a wide variety of experiments on the resulting model. When evaluated on downstream tasks (often with a lightweight linear model on frozen features), ViT-22B demonstrates increasing performance with scale. We further observe other interesting benefits of scale, including an improved tradeoff between fairness and performance, state-of-the-art alignment to human visual perception in terms of shape/texture bias, and improved robustness. ViT-22B demonstrates the potential for "LLM-like" scaling in vision, and provides key steps towards getting there.
Community
Scaling Vision Transformers to New Heights: ViT-22B Explored
Links π:
π Subscribe: https://www.youtube.com/@Arxflix
π Twitter: https://x.com/arxflix
π LMNT (Partner): https://lmnt.com/
Models citing this paper 8
Browse 8 models citing this paperDatasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 33
Collections including this paper 0
No Collection including this paper