Instructions to use hchen657Naza/PWM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use hchen657Naza/PWM with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("hchen657Naza/PWM", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
File size: 3,300 Bytes
dd253d4 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 | # ๐ RISE
<div id="top" align="left">
<a href="https://opendrivelab.com/rise/"><img src="https://img.shields.io/badge/Proj_Page-blue" alt="Project Page"></a>
<a href="https://arxiv.org/abs/2602.11075"><img src="https://img.shields.io/badge/arXiv-2602.11075-b31b1b" alt="arXiv"></a>
<a href="https://x.com/jiazhi_yang2024/status/2022247675060797636"><img src="https://img.shields.io/badge/summary-000000?logo=x&logoColor=white" alt="X"></a>
</div>
<img src="rise_assets/teaser.gif" alt="RISE demo" width="100%">
</img>
## ๐ฅ Highlights
<!-- RISE is a self-improving robot policy framework that turns world models into a practical learning environment for real-world manipulation. In short, we make the following three key contributions: -->
- **A compositional world model.**
A principled design that combines a controllable multi-view dynamics model with a progress value model, yielding informative advantages for robust policy improvement.
- **RL in imagination.**
A scalable self-improving framework that bootstraps robot policies through imaginary rollouts, avoiding the hardware cost and laborious reset of real-world interactions.
- **Real-world manipulation gains.**
Non-trivial performance improvements on challenging dexterous tasks, including +35% on dynamic brick sorting, +45% on backpack packing, and +35% on box closing.
## ๐บ๏ธ Overview
RISE repository is structured with three major parts:
- `policy_and_value/policy_offline_and_value`: OpenPI-based offline policy and value training. They are put together since they share similar model architecture and training pipeline.
- `dynamics/dynamics_model`: Action-conditioned dynamics model.
- `policy_and_value/policy_online`: Online RL for policy improvement.
## ๐งโ๐ป Getting Started
- [๐ป Installation](docs/installation.md)
- [๐ Offline Policy Training & Value Model](docs/offline_learning.md)
- [๐ฎ Dynamics Model](docs/dynamics_model.md)
- [๐ฆพ Online Policy Training](docs/online_training.md)
- [๐ ๏ธ Deployment on Piper](docs/deploy.md)
## โค๏ธ Acknowledgement
We thank the following projects for their open-source contributions: [OpenPI Pi0 / Pi05](https://github.com/openpi/openpi), [Genie Envisioner](https://github.com/AgibotTech/Genie-Envisioner), [RLinf](https://github.com/RLinf/RLinf), [LTX-Video](https://github.com/Lightricks/LTX-Video).
## ๐ข News
- [2026/04/22] Training code and pre-trained dynamics model are released.
- [2026/02/11] Paper released on [arXiv](https://arxiv.org/abs/2602.11075).
## ๐ License and Citation
All assets and code in this repository are under the Apache 2.0 license unless specified otherwise. The data and checkpoint are under CC BY-NC-SA 4.0. Other modules inherit their own distribution licenses.
```bibtex
@article{rise2026,
title={RISE: Self-Improving Robot Policy with Compositional World Model},
author={Yang, Jiazhi and Lin, Kunyang and Li, Jinwei and Zhang, Wencong and Lin, Tianwei and Wu, Longyan and Su, Zhizhong and Zhao, Hao and Zhang, Ya-Qin and Chen, Li and Luo, Ping and Yue, Xiangyu and Li, Hongyang},
journal={arXiv preprint arXiv:2602.11075},
year={2026}
}
```
[](https://star-history.com/#OpenDriveLab/RISE&Date)
|