Skip to content

Converts English words to Katana with a seq2seq model.

Notifications You must be signed in to change notification settings

wentaol/Weebifier

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Weebifier

Converts English (and not so English) words to Katana with a seq2seq model. Optionally override network predictions with a dictionary of word mappings provided as csv.

Based on this, trained with data from here

Requirements

  • numpy
  • torch

For preprocessing data:

  • sqlite3
  • pandas

For training:

  • torchtext

Usage and samples

from weebifier import Weebifier
w = Weebifier()
w.weebify("hello world")
w.weebify("some fake words:" )
w.weebify("nelinetrolls hevephiny lantifices")
w.weebify("and some harry potter spells:")
w.weebify("expelliarmus wingardium leviosa")

Output:

ヘロー ワールド
ソム フェイク ワーズ :
ネライントロールズ ヘビフィニー ランタファイシズ
アンド ソム ハリー ポター スペルズ :
エクスペリャルムス ウィンガーディアム レビオーサ

About

Converts English words to Katana with a seq2seq model.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages