metadata
license: apache-2.0
language:
- en
size_categories:
- n<1K
tags:
- taxonomy
- label-set
- text-classification
- few-shot
- synthetic
Categories-100-en
A curated set of 100 human-readable English category names, generated via self-instruct prompting with DeepSeek V4 Flash 0731.
Dataset Description
- Total items: 100
- Format: JSONL (one JSON object per line)
- Field:
"category"– human-readable category name (string) - Language: English
- License: Apache 2.0
- Source: LLM-generated via DeepSeek V4 Flash 0731 (self-instruct). Not affiliated with DeepSeek.
Use cases
- As a label set for zero-shot or few-shot text classification demos
- For testing long-list selection / tag extraction in LLMs
- As a quick vocabulary of common subject areas for educational tools
Limitations
- Categories are flat (no hierarchy) and some overlap (e.g., "Physics" vs."Astronomy", "Art" vs. "Painting").
- Generated by an LLM; expect some inconsistencies in granularity.
Quickstart
from datasets import load_dataset
ds = load_dataset("RowRed/categories-100-en", split="train")
categories = [row["category"] for row in ds]
print(len(categories)) # 100
License
This dataset is released under the Apache License 2.0.
Attribution:
Generated via self-instruct prompting using DeepSeek V4 Flash 0731 (MIT-licensed). Not affiliated with DeepSeek.
Citation
@misc{categories100en,
author = {RowRed},
title = {Categories-100-en},
year = {2026},
url = {https://huggingface.co/datasets/RowRed/categories-100-en}
}