---
pipeline_tag: image-text-to-text
---
This model is the stage 2 checkpoint of one of the thirteen settings, Stable Diffusion 1.5, used in the [Law of Vision Representation in MLLMs](https://arxiv.org/abs/2408.16357).