---
# For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/modelcard.md?plain=1
# Doc / guide: https://huggingface.co/docs/hub/model-cards
{}
---

# Official Model Card for (RPG) [Mastering Text-to-Image Diffusion: Recaptioning, Planning, and Generating with Multimodal LLMs](https://arxiv.org/abs/2401.11708)
The GitHub repo:https://github.com/YangLing0818/RPG-DiffusionMaster

RPG is a training-free paradigm that utilize MLLMs as prompt recaptioner and layout planner from diffusion models.  This paradigm can be generalized to different diffusion models, here we provide some high-quality community models in [CIVITAI](https://civitai.com/models) based on stable-diffusion v1.4/1.5 and SDXL-1.0/SDXL-Turbo for high-quality generation.
We will continue to update our model_pool combining with the ControlNet, stay tuned for our following progress.
## Stable-diffusion v1.4/1.5 based model Details
For Stable-diffusion v1.4/1.5 based models, we use

[AbsoluteReality](https://civitai.com/models/81458/absolutereality) for realistic style generation.

[AnythingV3](https://huggingface.co/Linaqruf/anything-v3.0) for anime style generation.

[Disney Pixar Cartoon](https://civitai.com/models/75650/disney-pixar-cartoon-type-b) for cartoon style generation.

## SDXL v1.0/ SDXL-Turbo based models Details
For SDXL v1.0/ SDXL-Turbo based models, we use

[AlbedoBaseXL](https://civitai.com/models/140737?modelVersionId=281176) for SDXL-baed photorealistic style generation.

[DreamShaperXL](https://civitai.com/models/112902/dreamshaper-xl) for SDXL-Turbo based photorealistic style generation.