Open Datasets - a Norm Collection

Norm 's Collections

Image / Video Gen

Multimodal Language Model

Fundamental Research

Computer Vision

Open Datasets

updated 7 days ago

Thank you for sharing your dataset. I’ve fed them to my model, and they are benefit to it.

vikhyatk/lnqa

Viewer • Updated Aug 18 • 303k • 231 • 79

Note VQA
lmms-lab/LLaVA-OneVision-Data

Viewer • Updated Sep 10 • 3.72M • 8.21k • 119
naver-clova-ix/synthdog-en

Viewer • Updated Jan 31 • 66k • 1.55k • 16

Note OCR
Mutonix/Vript_Chinese

Viewer • Updated Jun 15 • 294k • 5 • 7

Note Caption, Video
ShareGPT4Video/ShareGPT4Video

Viewer • Updated Jul 8 • 40.2k • 487 • 175

Note Caption, Video
FreedomIntelligence/ALLaVA-4V

Viewer • Updated Sep 2 • 143k • 1.26k • 64

Note VQA, Caption
BAAI/DenseFusion-1M

Viewer • Updated Aug 21 • 1.18M • 24 • 26
tomg-group-umd/pixelprose

Viewer • Updated Jun 23 • 15.6M • 83 • 122
priyank-m/MJSynth_text_recognition

Viewer • Updated Jul 4, 2023 • 8.92M • 20 • 3

Note OCR
Rich Human Feedback for Text-to-Image Generation

Paper • 2312.10240 • Published Dec 15, 2023 • 19

Note Aesthetic Benchmark
HuggingFaceFV/finevideo

Viewer • Updated 18 days ago • 43.8k • 484 • 250
lmms-lab/LLaVA-Video-178K

Viewer • Updated about 9 hours ago • 1.63M • 517 • 45

Note Video Caption; Video QA Level-1: Every 10 seconds; Level-2: Every 30 seconds; Level-3: encapsulate the entire video.
lmms-lab/llava-critic-113k

Viewer • Updated 6 days ago • 113k • 392 • 14

Note DPO; Reflection