README.md

metadata

language:
  - en
license: odc-by
task_categories:
  - text-classification

c4-en-lowercase

A 125 000-row subset of the C4 dataset's English configuration, divided into:

• train: 100 000 rows
• validation: 25 000 rows

Each row contains:

• text: Original text
• lowercased: Lowercase version of text

Data Instance Example

{
  "text": "Graduate from Cambridge University, currently doing my masters in International Politics and Economics at LSE.",
  "lowercased": "graduate from cambridge university, currently doing my masters in international politics and economics at lse."
}

Supported Tasks

• Studying capitalization patterns
• Text normalization
• Case-insensitive text analysis

Limitations

• May contain web-crawled text biases
• English language content
• Small sample size
• Not thoroughly shuffled