README.md

---
language:
- en
multilinguality:
- monolingual
size_categories:
- 1M<n<10M
task_categories:
- feature-extraction
- sentence-similarity
pretty_name: altlex
tags:
- sentence-transformers
dataset_info:
  config_name: pair
  features:
  - name: text
    dtype: string
  - name: simplified
    dtype: string
  splits:
  - name: train
    num_bytes: 29717244
    num_examples: 112696
  download_size: 19447229
  dataset_size: 29717244
configs:
- config_name: pair
  data_files:
  - split: train
    path: pair/train-*
---

# Dataset Card for altlex

This dataset is a collection of pairs of English Wikipedia entries and their simplified variants. See [altlex](https://github.com/chridey/altlex/) for additional information.
This dataset can be used directly with Sentence Transformers to train embedding models.

## Dataset Subsets

### `pair` subset

* Columns: "text", "simplified"
* Column types: `str`, `str`
* Examples:
    ```python
    {
      'text': "Valentin Berlinsky retired from the Borodin Quartet in September 2007 , and was succeeded by his protégé , Vladimir Balshin , but he still remained the group 's mentor .",
      'simplified': 'Valentin Berlinsky retired from the Borodin Quartet in September 2007 .',
    }
    ```
* Collection strategy: Reading the SimpleWiki dataset from [embedding-training-data](https://huggingface.co/datasets/sentence-transformers/embedding-training-data).
* Deduplified: No