---
language:
- he
license: mit
task_categories:
- text-generation
tags:
- hebrew
- literature
- project-ben-yehuda
- public-domain
pretty_name: Hebrew Project Ben-Yehuda (parquet)
dataset_info:
  features:
    - name: id
      dtype: int64
    - name: title
      dtype: string
    - name: authors
      dtype: string
    - name: genre
      dtype: string
    - name: text
      dtype: string
  splits:
    - name: train
---

# Hebrew Project Ben-Yehuda

Public domain Hebrew literature from [Project Ben-Yehuda](https://benyehuda.org), built from the [public domain dump](https://github.com/projectbenyehuda/public_domain_dump) (release 2022-12). Plaintext without nikkud by default.

## Schema

`id`, `title`, `authors`, `genre`, `text`

## Usage

```python
from datasets import load_dataset
ds = load_dataset("parquet", data_files="path/to/data/train.parquet", split="train")
```