Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -7,4 +7,25 @@ This repository contains the data for [Open-Qwen2VL: Compute-Efficient Pre-Train
|
|
7 |
|
8 |
Project page: https://victorwz.github.io/Open-Qwen2VL
|
9 |
|
10 |
-
Code: https://github.com/Victorwz/Open-Qwen2VL
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
|
8 |
Project page: https://victorwz.github.io/Open-Qwen2VL
|
9 |
|
10 |
+
Code: https://github.com/Victorwz/Open-Qwen2VL
|
11 |
+
|
12 |
+
## Dataset
|
13 |
+
- ccs_ebdataset: CC3M-CC12M-SBU filtered by CLIP by BLIP-1
|
14 |
+
- datacomp_medium_dfn_webdataset: DataComp-Medium-128M filtered by DFN, we just select this subset based the uids released by DFN
|
15 |
+
- datacomp_medium_mlm_filter_su_85_union_dfn_webdataset: DataComp-Medium-128M filtered by DFN union DataComp-Medium-128M filtered by MLM-Filter based on the semantic understanding metric with threshold 85
|
16 |
+
|
17 |
+
|
18 |
+
|
19 |
+
## Acknowledgement
|
20 |
+
This work was partially supported by the BioPACIFIC Materials Innovation Platform of the National Science Foundation under Award No. DMR-1933487
|
21 |
+
|
22 |
+
|
23 |
+
## Citation
|
24 |
+
```bibtex
|
25 |
+
@article{Open-Qwen2VL,
|
26 |
+
title={Open-Qwen2VL: Compute-Efficient Pre-Training of Fully-Open Multimodal LLMs on Academic Resources},
|
27 |
+
author={Wang, Weizhi and Tian, Yu and Yang, Linjie and Wang, Heng and Yan, Xifeng},
|
28 |
+
journal={arXiv preprint arXiv:2504.00595},
|
29 |
+
year={2025}
|
30 |
+
}
|
31 |
+
```
|