searx.utils.detect_language : add model from fastlangid

Browse source

The project fastlangid has a FastText model to achieve higher accuracy
for Japanese, Korean, Chinese languages.

This commits includes the model from this project,
and change searx.utils.detect_language to rely on that model
as a second stage.

The model is lazy loaded when the base model detect Japanese, Korean or Chinese.

https://github.com/currentslab/fastlangid

...

This commit is contained in:

Alexandre Flament 2023-09-17 09:31:47 +00:00

parent 7dfcc3386e

commit 931e1f6503

2 changed files with 51 additions and 13 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

BIN

searx/data/model_s.ftz Normal file

View file

Binary file not shown.