Datasets:
File size: 395 Bytes
e8cce97 755719a e8cce97 755719a ec2a8bf 755719a ec2a8bf | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 | ---
pretty_name: Urdu Newspaper Benchmark
language:
- ur
task_categories:
- image-to-text
size_categories:
- n<1K
---
# Urdu Newspaper Benchmark
This dataset follows the Hugging Face `ImageFolder` format with a `metadata.csv` file.
## Split
- `train`: 829 examples
## Columns
- `HR Image`: high-resolution image
- `LR Image`: low-resolution image
- `Transcription`: ground-truth Urdu text
|