taishi-i/toiro

A tool for comparing tokenizers

52
/ 100
Established

When working with Japanese text for natural language processing, this tool helps you evaluate and select the best 'tokenizer' for your specific needs. It takes raw Japanese text and various tokenization methods as input, then shows you how each one breaks the text into words, their processing speed, and their performance on tasks like text classification. This is invaluable for data scientists or NLP engineers focusing on Japanese language processing.

121 stars. Available on PyPI.

Use this if you need to choose the most suitable Japanese tokenizer for a new project and want to compare options based on speed, segmentation accuracy, or downstream application performance.

Not ideal if your work exclusively involves languages other than Japanese, as its primary focus and features are tailored for Japanese text processing.

Japanese-NLP text-tokenization linguistic-analysis model-benchmarking text-segmentation
Maintenance 6 / 25
Adoption 10 / 25
Maturity 25 / 25
Community 11 / 25

How are scores calculated?

Stars

121

Forks

9

Language

Python

License

Apache-2.0

Last pushed

Nov 09, 2025

Commits (30d)

0

Dependencies

6

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/nlp/taishi-i/toiro"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.