Hugging Face's logo

Datasets:
mvp-lab
/
LLaVA-OneVision-2-Data

Tasks:
Video-Text-to-Text
Visual Question Answering
Image-Text-to-Text
Modalities:
Image
Text
Video
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
llava
multimodal
video
spatial-reasoning
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
LLaVA-OneVision-2-Data / mid_training_video /caption_v0
27.7 GB
  • 1 contributor
 History: 1 commit
xiangan's picture
xiangan
Add caption_v0 split jsonl files
b7365ef verified