Spaces:
Running
Running
Update README.md
Browse files
README.md
CHANGED
|
@@ -7,4 +7,72 @@ sdk: static
|
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
+
# VMEvalKit 🎥🧠
|
| 11 |
+
|
| 12 |
+
|
| 13 |
+
<div align="center">
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
[](https://grow-ai-like-a-child.com/video-reason/)
|
| 17 |
+
[](https://github.com/hokindeng/VMEvalKit/paper/video-models-start-to-solve/Video_Model_Start_to_Solve.pdf)
|
| 18 |
+
[](https://huggingface.co/VideoReason)
|
| 19 |
+
[](https://github.com/hokindeng/VMEvalKit/issues/132)
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
</div>
|
| 23 |
+
|
| 24 |
+
A framework to evaluate reasoning capabilities in video generation models at scale.
|
| 25 |
+
|
| 26 |
+
<p align="center">
|
| 27 |
+
|
| 28 |
+
</p>
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
## 🎬 Supported Models
|
| 34 |
+
|
| 35 |
+
VMEvalKit provides unified access to **40 video generation models** across **11 provider families**:
|
| 36 |
+
|
| 37 |
+
For commercial APIs, we support Luma Dream Machine, Google Veo, Google Veo 3.1, WaveSpeed WAN 2.1, WaveSpeed WAN 2.2, Runway ML, OpenAI Sora. For open-source models, we support HunyuanVideo, VideoCrafter, DynamiCrafter, Stable Video Diffusion, Morphic, LTX-Video, and so on. See [here](docs/models/README.md) for details.
|
| 38 |
+
|
| 39 |
+
## Invitation to Collaborate 🤝
|
| 40 |
+
|
| 41 |
+
VMEvalKit is meant to be a permissively open-source **shared playground** for everyone. If you’re interested in machine cognition, video models, evaluation, or anything anything 🦄✨, we’d love to build with you:
|
| 42 |
+
|
| 43 |
+
* 🧪 Add new reasoning tasks (planning, causality, social, physical, etc.)
|
| 44 |
+
* 🎥 Plug in new video models (APIs or open-source)
|
| 45 |
+
* 📊 Experiment with better evaluation metrics and protocols
|
| 46 |
+
* 🧱 Improve infrastructure, logging, and the web dashboard
|
| 47 |
+
* 📚 Use VMEvalKit in your own research and share back configs/scripts
|
| 48 |
+
* 🌟🎉 Or Anything anything 🦄✨
|
| 49 |
+
|
| 50 |
+
💬 **Join us on Slack** to ask questions, propose ideas, or start a collab:
|
| 51 |
+
[Slack Invite](https://join.slack.com/t/growingailikeachild/shared_invite/zt-309yqd0sl-W8xzOkdBPha1Jh5rnee78A) 🚀
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
## Research
|
| 55 |
+
|
| 56 |
+
Here we keep track of papers spinned off from this code infrastructure and some works in progress.
|
| 57 |
+
|
| 58 |
+
- [**"Video Models Start to Solve Chess, Maze, Sudoku, Mental Rotation, and Raven's Matrices"**](paper/video-models-start-to-solve/Video_Model_Start_to_Solve.pdf)
|
| 59 |
+
|
| 60 |
+
This paper implements our experimental framework and demonstrates that leading video generation models (Sora-2 etc) can perform visual reasoning tasks with >60% success rates. See [**results**](https://grow-ai-like-a-child.com/video-reason/).
|
| 61 |
+
|
| 62 |
+
## License
|
| 63 |
+
|
| 64 |
+
Apache 2.0
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
## Citation
|
| 68 |
+
|
| 69 |
+
If you find VMEvalKit useful in your research, please cite:
|
| 70 |
+
|
| 71 |
+
```bibtex
|
| 72 |
+
@misc{VMEvalKit,
|
| 73 |
+
author = {VMEvalKit Team},
|
| 74 |
+
title = {VMEvalKit: A framework for evaluating reasoning abilities in foundational video models},
|
| 75 |
+
year = {2025},
|
| 76 |
+
howpublished = {\url{https://github.com/Video-Reason/VMEvalKit}}
|
| 77 |
+
}
|
| 78 |
+
```
|