About the dataset - Githubissues

AIDC-AI / Ovis

A novel Multimodal Large Language Model (MLLM) architecture, designed to structurally align visual and textual embeddings.

Apache License 2.0

506 stars 29 forks source link

Closed lucasjinreal closed 2 months ago

lucasjinreal commented 2 months ago

Does the pixelprose have overlapped with ovis-cc12m? since pixelprose already contains cc12m

runninglsy commented 2 months ago

While ovis-cc12m and pixelprose may share some overlapping images, their respective captions are distinct.