Datasets:

Dharma-AI
/

DharmaOCR-Benchmark

Tasks:

Image-Text-to-Text

Modalities:

Formats:

Languages:

Size:

ArXiv:

Tags:

document-understanding

brazilian-portuguese

text-recognition

handwriting-recognition

Libraries:

License:

Dataset card Data Studio Files Files and versions

DharmaOCR-Benchmark / data

2.53 GB

2 contributors

History: 57 commits

Paulo-DharmaAI's picture

Remove benchmark shard: data/benchmark-00005-of-00006.parquet

03d7a86 verified 10 days ago