ja-ginza

Japanese multi-task CNN trained on UD-Japanese BCCWJ r2.8 + GSK2014-A(2019). Assigns word2vec token vectors. Components: tok2vec, parser, ner, morphologizer, atteribute_ruler, compound_splitter, buns…

Installation

In a virtualenv (see these instructions if you need to create one):

pip3 install ja-ginza

Dependencies

Releases

Version Released Bookworm
Python 3.11
Trixie
Python 3.13
Files
5.2.0 2024-03-30    
5.1.3 2023-09-25    
5.1.2 2022-08-09
5.1.0 2021-12-04
5.0.0 2021-08-25
4.0.0 2020-08-16
3.1.0 2020-01-15
3.0.0 2020-01-14

Issues with this package?

Page last updated 2026-09-03 08:47:40 UTC