Fix model card layout: inline badges and side-by-side demo grids
Browse files
README.md
CHANGED
|
@@ -18,9 +18,7 @@ license: apache-2.0
|
|
| 18 |
|
| 19 |
**MiniWorld: Democratizing the Training of Video World Models from Scratch**
|
| 20 |
|
| 21 |
-
<a href="https://zhao-yian.github.io/MiniWorld/"><img src="https://img.shields.io/badge/Project-Page-1f6feb?style=for-the-badge&logo=googlechrome&logoColor=white" alt="Project Page"></a>
|
| 22 |
-
<img src="https://img.shields.io/badge/arXiv-Coming%20Soon-b31b1b?style=for-the-badge&logo=arxiv&logoColor=white" alt="arXiv">
|
| 23 |
-
<a href="https://github.com/zhao-yian/MiniWorld"><img src="https://img.shields.io/badge/GitHub-Code-181717?style=for-the-badge&logo=github&logoColor=white" alt="GitHub"></a>
|
| 24 |
|
| 25 |
MiniWorld is a minimal and reproducible framework for training streaming video
|
| 26 |
world models from scratch. Instead of adapting a pretrained bidirectional video
|
|
@@ -40,13 +38,9 @@ training scripts, and evaluation utilities live in the GitHub repository.
|
|
| 40 |
Each tile is a 253-frame streaming rollout from the `1B` checkpoint, generated
|
| 41 |
from a single observed frame plus the control signal.
|
| 42 |
|
| 43 |
-
<p align="center">
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
</p>
|
| 47 |
-
<p align="center">
|
| 48 |
-
<sub><b>DROID</b> action-conditioned rollouts (left) and <b>RealEstate10K</b> camera-conditioned rollouts (right), shown at 2× speed. Click either grid for the full-resolution video, or see the <a href="https://zhao-yian.github.io/MiniWorld/">project page</a> for all 100 rollouts.</sub>
|
| 49 |
-
</p>
|
| 50 |
|
| 51 |
## Model Summary
|
| 52 |
|
|
|
|
| 18 |
|
| 19 |
**MiniWorld: Democratizing the Training of Video World Models from Scratch**
|
| 20 |
|
| 21 |
+
<a href="https://zhao-yian.github.io/MiniWorld/"><img src="https://img.shields.io/badge/Project-Page-1f6feb?style=for-the-badge&logo=googlechrome&logoColor=white" alt="Project Page"></a> <img src="https://img.shields.io/badge/arXiv-Coming%20Soon-b31b1b?style=for-the-badge&logo=arxiv&logoColor=white" alt="arXiv"> <a href="https://github.com/zhao-yian/MiniWorld"><img src="https://img.shields.io/badge/GitHub-Code-181717?style=for-the-badge&logo=github&logoColor=white" alt="GitHub"></a>
|
|
|
|
|
|
|
| 22 |
|
| 23 |
MiniWorld is a minimal and reproducible framework for training streaming video
|
| 24 |
world models from scratch. Instead of adapting a pretrained bidirectional video
|
|
|
|
| 38 |
Each tile is a 253-frame streaming rollout from the `1B` checkpoint, generated
|
| 39 |
from a single observed frame plus the control signal.
|
| 40 |
|
| 41 |
+
<p align="center"><a href="https://huggingface.co/zhaoyian01/MiniWorld/resolve/main/assets/demo_droid.mp4"><img src="assets/demo_droid.webp" width="420" alt="DROID action-conditioned rollouts"></a> <a href="https://huggingface.co/zhaoyian01/MiniWorld/resolve/main/assets/demo_re10k.mp4"><img src="assets/demo_re10k.webp" width="420" alt="RealEstate10K camera-conditioned rollouts"></a></p>
|
| 42 |
+
|
| 43 |
+
<p align="center"><sub><b>DROID</b> action-conditioned rollouts (left) and <b>RealEstate10K</b> camera-conditioned rollouts (right), shown at 2× speed. Click either grid for the full-resolution video, or see the <a href="https://zhao-yian.github.io/MiniWorld/">project page</a> for all 100 rollouts.</sub></p>
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
|
| 45 |
## Model Summary
|
| 46 |
|