README.md

metadata

language:
  - en
multilinguality:
  - monolingual
size_categories:
  - 1M<n<10M
task_categories:
  - feature-extraction
  - sentence-similarity
pretty_name: Wikipedia Sections
tags:
  - sentence-transformers
dataset_info:
  - config_name: pair
    features:
      - name: anchor
        dtype: string
      - name: positive
        dtype: string
    splits:
      - name: train
        num_bytes: 490913561
        num_examples: 1779417
      - name: validation
        num_bytes: 60891304
        num_examples: 220400
      - name: test
        num_bytes: 61385426
        num_examples: 222957
    download_size: 295222520
    dataset_size: 613190291
  - config_name: triplet
    features:
      - name: anchor
        dtype: string
      - name: positive
        dtype: string
      - name: negative
        dtype: string
    splits:
      - name: train
        num_bytes: 733058519
        num_examples: 1779417
      - name: validation
        num_bytes: 90881953
        num_examples: 220400
      - name: test
        num_bytes: 91705993
        num_examples: 222957
    download_size: 500545462
    dataset_size: 915646465
configs:
  - config_name: pair
    data_files:
      - split: train
        path: pair/train-*
      - split: validation
        path: pair/validation-*
      - split: test
        path: pair/test-*
  - config_name: triplet
    data_files:
      - split: train
        path: triplet/train-*
      - split: validation
        path: triplet/validation-*
      - split: test
        path: triplet/test-*

Dataset Card for Wikipedia Sections

This dataset contains pairs and triplets that can be used to train and finetune Sentence Transformer embedding models. The dataset originates from Dor et al., and was downloaded from this download link. Notably, the "anchor" column contains sentences from Wikipedia, wheras the "positive" column contains other sentences from the same section. The "negative" column contains sentences from other sections.

Dataset Subsets

pair subset

• Columns: "anchor", "positive"
• Column types: str, str
• Examples:

• Collection strategy: Reading the Wikipedia Sections dataset from https://sbert.net.
• Deduplified: Yes

triplet subset

• Columns: "anchor", "positive", "negative"
• Column types: str, str, str
• Examples:

• Collection strategy: Reading the Wikipedia Sections dataset from https://sbert.net.
• Deduplified: Yes