---
language:
- ml
license: other
size_categories:
- 10K<n<100K
task_categories:
- image-to-text
---

# Malayalam-only Mozhi Dataset

This dataset is a filtered version of `darknight054/indic-mozhi-ocr`, containing only the Malayalam language edition.

## Dataset Description
- **Language:** Malayalam
- **Task:** Optical Character Recognition (OCR) / Image-to-Text
- **Source:** [darknight054/indic-mozhi-ocr](https://huggingface.co/datasets/darknight054/indic-mozhi-ocr)

## Structure
The dataset contains printed word images and their corresponding transcriptions.

## Usage
```python
from datasets import load_dataset
ds = load_dataset('trysem/Mozhi-Malayalam')
```