tomaarsen's picture
tomaarsen HF staff
Update README.md
4a0972d verified
|
raw
history blame
1.19 kB
metadata
language:
  - en
multilinguality:
  - monolingual
size_categories:
  - 1M<n<10M
task_categories:
  - feature-extraction
  - sentence-similarity
pretty_name: Wikipedia Sentences
tags:
  - sentence-transformers
dataset_info:
  features:
    - name: sentence
      dtype: string
  splits:
    - name: train
      num_bytes: 958187457
      num_examples: 7871825
  download_size: 683051610
  dataset_size: 958187457
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Dataset Card for Wikipedia Sentences (English)

This dataset contains 7.87 million English sentences and can be used in knowledge distillation of embedding models.

Dataset Details

  • Columns: "sentence"
  • Column types: str
  • Examples:
    {
      'sentence': "After the deal was approved and NONG's stock rose to $13, Farris purchased 10,000 shares at the $2.50 price, sold 2,500 shares at the new price to reimburse the company, and gave the remaining 7,500 shares to Landreville at no cost to him."
    }
    
  • Collection strategy: Downloaded from https://sbert.net/datasets/wikipedia-en-sentences.txt.gz and uploaded without further modifications
  • Deduplified: No