Update README.md
Browse files
README.md
CHANGED
|
@@ -7,15 +7,23 @@ library_name: transformers
|
|
| 7 |
|
| 8 |
# VBVR: A Very Big Video Reasoning Suite
|
| 9 |
|
| 10 |
-
<a href="" target="_blank">
|
|
|
|
|
|
|
|
|
|
| 11 |
<img alt="Code" src="https://img.shields.io/badge/VBVR-Code-100000?style=flat-square&logo=github&logoColor=white" height="20" />
|
| 12 |
</a>
|
| 13 |
<a href="" target="_blank">
|
| 14 |
<img alt="arXiv" src="https://img.shields.io/badge/arXiv-VBVR-red?logo=arxiv" height="20" />
|
| 15 |
</a>
|
| 16 |
-
|
| 17 |
-
<
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
</a>
|
| 20 |
|
| 21 |
|
|
@@ -130,9 +138,10 @@ to unseen reasoning tasks. **Together, VBVR lays a foundation for the next stage
|
|
| 130 |
VBVR-Wan2.2 is trained from Wan2.2-I2V-A14B without architectural modifications, as the goal of VBVR-Wan2.2 is to *investigate data scaling behavior* and provide a *strong baseline model* for the video reasoning research community. Leveraging the VBVR-Dataset, which to our knowledge constitutes one of the largest video reasoning datasets to date, VBVR-Wan2.2 achieved highest score on VBVR-Bench.
|
| 131 |
|
| 132 |
In this release, we present
|
| 133 |
-
[**VBVR-Wan2.2**](https://huggingface.co/Video-Reason/VBVR-Wan2.2)
|
| 134 |
-
[**VBVR-
|
| 135 |
-
[**VBVR-
|
|
|
|
| 136 |
|
| 137 |
|
| 138 |
## 🛠️ QuickStart
|
|
|
|
| 7 |
|
| 8 |
# VBVR: A Very Big Video Reasoning Suite
|
| 9 |
|
| 10 |
+
<a href="https://video-reason.com" target="_blank">
|
| 11 |
+
<img alt="Code" src="https://img.shields.io/badge/Project%20-%20Homepage-4285F4" height="20" />
|
| 12 |
+
</a>
|
| 13 |
+
<a href="https://github.com/orgs/Video-Reason/repositories" target="_blank">
|
| 14 |
<img alt="Code" src="https://img.shields.io/badge/VBVR-Code-100000?style=flat-square&logo=github&logoColor=white" height="20" />
|
| 15 |
</a>
|
| 16 |
<a href="" target="_blank">
|
| 17 |
<img alt="arXiv" src="https://img.shields.io/badge/arXiv-VBVR-red?logo=arxiv" height="20" />
|
| 18 |
</a>
|
| 19 |
+
<a href="https://huggingface.co/Video-Reason/VBVR-Dataset" target="_blank">
|
| 20 |
+
<img alt="Leaderboard" src="https://img.shields.io/badge/%F0%9F%A4%97%20_VBVR_Dataset-Data-ffc107?color=ffc107&logoColor=white" height="20" />
|
| 21 |
+
</a>
|
| 22 |
+
<a href="https://huggingface.co/Video-Reason/VBVR-Bench-Data" target="_blank">
|
| 23 |
+
<img alt="Leaderboard" src="https://img.shields.io/badge/%F0%9F%A4%97%20_VBVR_Bench-Data-ffc107?color=ffc107&logoColor=white" height="20" />
|
| 24 |
+
</a>
|
| 25 |
+
<a href="https://huggingface.co/Video-Reason/VBVR-Bench-Leaderboard" target="_blank">
|
| 26 |
+
<img alt="Leaderboard" src="https://img.shields.io/badge/%F0%9F%A4%97%20_VBVR_Bench-Leaderboard-ffc107?color=ffc107&logoColor=white" height="20" />
|
| 27 |
</a>
|
| 28 |
|
| 29 |
|
|
|
|
| 138 |
VBVR-Wan2.2 is trained from Wan2.2-I2V-A14B without architectural modifications, as the goal of VBVR-Wan2.2 is to *investigate data scaling behavior* and provide a *strong baseline model* for the video reasoning research community. Leveraging the VBVR-Dataset, which to our knowledge constitutes one of the largest video reasoning datasets to date, VBVR-Wan2.2 achieved highest score on VBVR-Bench.
|
| 139 |
|
| 140 |
In this release, we present
|
| 141 |
+
[**VBVR-Wan2.2**](https://huggingface.co/Video-Reason/VBVR-Wan2.2),
|
| 142 |
+
[**VBVR-Dataset**](https://huggingface.co/Video-Reason/VBVR-Dataset),
|
| 143 |
+
[**VBVR-Bench-Data**](https://huggingface.co/Video-Reason/VBVR-Bench-Data) and
|
| 144 |
+
[**VBVR-Bench-Leaderboard**](https://huggingface.co/Video-Reason/VBVR-Bench-Leaderboard).
|
| 145 |
|
| 146 |
|
| 147 |
## 🛠️ QuickStart
|