skylenage-ai/DeepVision-103K
Image Text To TextEN
Created by skylenage-ai at 2026, the skylenage-ai/DeepVision-103K is a image text to text dataset in EN in Parquet format.
About skylenage-ai/DeepVision-103K
🔭 DeepVision-103K
A Visually Diverse, Broad-Coverage, and Verifiable Mathematical Dataset for Multimodal Reasoning
Training on DeepVision-103K yields top performance on both multimodal mathematical reasoning and general multimoda...
Details
- Task
- Image Text To Text
- Language
- EN
- Format
- Parquet
- Rows / instances
- N/A
- Creator
- skylenage-ai
- Year
- 2026