---
license: cc-by-4.0
task_categories:
- robotics
tags:
- sketchy
- open-x
---

# Sketchy 128x128

This dataset is a pre-processed version of the Sketchy dataset (originally from DeepMind and the Open-X Embodiment effort) used in the paper [Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling](https://huggingface.co/papers/2603.04553).

The data has been pre-processed to 128x128 pixels. Unlike the original TFRecords, this version consists of raw frames/images, facilitating use in standard PyTorch/JAX pipelines for training object-centric world models.

- **Project Page:** [https://taldatech.github.io/lpwm-web](https://taldatech.github.io/lpwm-web)
- **GitHub Repository:** [https://github.com/taldatech/lpwm](https://github.com/taldatech/lpwm)

## Usage

To train the Latent Particle World Model (LPWM) on the Sketchy dataset, you can use the following command from the [official repository](https://github.com/taldatech/lpwm):

```bash
python train_lpwm.py --dataset sketchy
```

## Citation

```bibtex
@inproceedings{
daniel2026latent,
title={Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling},
author={Tal Daniel and Carl Qi and Dan Haramati and Amir Zadeh and Chuan Li and Aviv Tamar and Deepak Pathak and David Held},
booktitle={The Fourteenth International Conference on Learning Representations},
year={2026},
url={https://openreview.net/forum?id=lTaPtGiUUc}
}
```