README.md

metadata

language:
  - en
multilinguality:
  - monolingual
size_categories:
  - 100M<n<1B
task_categories:
  - feature-extraction
  - sentence-similarity
pretty_name: WikiAnswers Duplicate Questions
tags:
  - sentence-transformers
dataset_info:
  config_name: pair
  features:
    - name: anchor
      dtype: string
    - name: positive
      dtype: string
  splits:
    - name: train
      num_bytes: 78825722188
      num_examples: 761379586
  download_size: 33891162136
  dataset_size: 78825722188
configs:
  - config_name: pair
    data_files:
      - split: train
        path: pair/train-*

Dataset Card for WikiAnswers Duplicate Questions

This dataset contains duplicate questions from the WikiAnswers Corpus, formatted to be easily used with Sentence Transformers to train embedding models.

Dataset Subsets

pair subset

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

  {
    'anchor': 'How many calories is in a handful of strawberries?',
    'positive': 'How many calories are in a strawberry popsickles?',
  }
  

• Collection strategy: Reading the WikiAnswers dataset from embedding-training-data, which has lists of duplicate questions. I've considered all adjacent questions as a positive pair, plus the last and first caption. So, e.g. 5 duplicate questions results in 5 duplicate pairs.
• Deduplified: No