Datasetpublic

Sankar-2910/genz-to-english

GenZ-to-English Translation Dataset A high-quality text-to-text dataset for translating Gen Z slang into clear, standard English. The dataset is designed for training and evaluating language models that convert modern internet slang into natural, readable English while preserving the original meaning. Overview This dataset contains 300k++ curated translation pairs covering a wide range of contemporary internet slang. It includes expressions commonly found across… See the full description on the dataset page: https://huggingface.co/datasets/Sankar-2910/genz-to-english.

sourceHugging Faceapache-2.0updated 17d agoView on Hugging Face

1likes153downloads

Dataset Card

GenZ-to-English Translation Dataset

A high-quality text-to-text dataset for translating Gen Z slang into clear, standard English.

The dataset is designed for training and evaluating language models that convert modern internet slang into natural, readable English while preserving the original meaning.

Overview

This dataset contains 300k++ curated translation pairs covering a wide range of contemporary internet slang.

It includes expressions commonly found across TikTok, Discord, Instagram, gaming communities, memes, and everyday online conversations.

Example Translations

Gen Z	Standard English
`bro got cooked in chem`	He performed very poorly in chemistry.
`that's lowkey fire ngl`	That's honestly really good.
`she ghosted after one date`	She stopped responding after one date without explanation.
`unc got ratio'd for saying drip is dead`	The older person received overwhelming criticism for saying stylish clothing is outdated.

Dataset Structure

The dataset is provided as a CSV with three columns.

Column	Description
`instruction`	Translation instruction
`input`	Gen Z slang sentence
`output`	Standard English translation

Example:

csv

instruction,input,output
Translate the following Gen Z slang sentence into clear, standard English.,bro got cooked in chem,He performed very poorly in chemistry.

Coverage

The dataset includes slang from multiple categories.

TikTok & Viral Slang

—rizz
—gyat
—aura
—glow up

Gaming

—NPC
—cooked
—locked in
—sus

Relationships

—green flag
—beige flag
—ick
—ghosted

Meme Culture

—canon event
—brainrot
—Fanum tax
—mogging
—looksmaxxing

Text Messaging

—fr
—ngl
—bet
—lowkey
—highkey
—no cap

Fashion

—drip
—aesthetic
—slayed
—ate

Everyday Conversations

—school
—workplace
—friend groups
—family interactions

The dataset also includes compound slang phrases where multiple expressions appear within the same sentence.

How the Dataset Was Created

The dataset was built through a combination of:

—manually written examples,
—synthetic examples generated using multiple publicly available language models,
—filtering,
—deduplication,
—manual quality review.

Translations prioritize preserving meaning and context rather than producing literal word-for-word conversions.

Intended Uses

This dataset is suitable for:

—training translation-style language models,
—benchmarking Gen Z slang understanding,
—educational applications,
—language normalization research,
—evaluating informal-to-formal text conversion.

Limitations

—Internet slang evolves rapidly, so newer expressions may not be represented.
—Some phrases are intentionally humorous or context-dependent.
—The dataset primarily focuses on English-language internet slang.

License

This dataset is released under the Apache License 2.0.

You are free to:

—use it commercially,
—modify it,
—redistribute it,
—train models using it,

provided that the Apache-2.0 license and accompanying notices are preserved.

Citation

If you use this dataset in research or public projects, please cite it.

bibtex

@dataset{genz_to_english,
  title={Genz-to-english Translation Dataset},
  author={sankar-2910},
  year={2026},
  publisher={Hugging Face}
}

Acknowledgements

This dataset was created using open-source NLP tools and a combination of manually authored and synthetic examples, followed by filtering and quality review.