Skip to content

Dataset · b3x0m

Corpus of 18+ Chinese novels

b3x0m/Chinese-H-Novels

A large corpus of 18+ Chinese novels for language modeling and classification tasks.

What's inside

A text corpus of adult Chinese novels; the dataset card notes that it should be used at your own risk. Metadata size is 100M<n<1B, and the language is Chinese. Converted to parquet for convenient downloading.

What you can do with it

  • Pre-training or fine-tuning language models on Chinese fiction
  • Training LoRAs for novel stylization or generation
  • Testing NSFW text classifiers or filters
Tasks
Text classification, Summarization, Token classification, Question answering, Text generation, Fill mask
Languages
Chinese
Size
100M–1B rows
Downloads this month
2.5K
Likes
257

How to load it

from datasets import load_dataset

ds = load_dataset("b3x0m/Chinese-H-Novels")