Skip to content

Dataset · jamesqijingsong

Illustrations, texts, and audio for Chinese dictionary

jamesqijingsong/zidian

Multimodal set of 4,307 characters with AI images, definitions, and pinyin audio for generative and language models.

What's inside

The dataset is based on the Chinese online dictionary zidian.18dao.net and contains 4,307 records in Chinese and English. It includes neural-network-generated illustrations in png format (4,307 items), text-overlay versions in jpg (4,307 items), text explanations (4,307 files), and pinyin audio in mp3 (1,702 files).

What you can do with it

  • Training multimodal models to link Chinese characters with visual imagery
  • Creating educational apps for learning Chinese language and phonetics
  • LoRA fine-tuning for generating illustrations for texts and dictionary entries
Languages
Chinese, English
Size
1K–10K rows
Downloads this month
110.8K
Likes
6

How to load it

from datasets import load_dataset

ds = load_dataset("jamesqijingsong/zidian")