More slight changes
Treat onomatopoeia as interjection
Trim off 'no translation' and fix to again
Fix script's desire to consume to.
Fix some errors due to arcane format
Fix to deletion
Stop the to monster from gobbling up tos
Add feature to remove duplicates.
Place comments at end of entry.
Fix certain minor mistakes with sakhadic2dix.py
Add comment indicating is generate dix and remo...
Add alphabet and sdefs to the generated diction...
Shift to new dir
Fix bug where script fails when there are no tags
Add sakhadic2dix.py
Fix some typos.
Fix some typos.
Add indicators that tell the level of accuracy.
Fix some error handling stuff + rm trash file
Seperate normalization of corpus from testing c...
Force yaml to use block style; much easier to edit
Update such that tests up to 20000 lines chosen...
Rename langtest.py to lang-test.py to follow na...
Add `lang-identify` and `lang-test`.
Forgot to include partial sentence.
Bring completion rate up to 62.
Add partial translations to aid the next person...
Raise completion to 34.
created zho.json, much is left
Prevent duplicate identifyLang() and add XRegEx...
Add new auto-translate system that reduces the ...
Fix high memory usage issue.
Add auto-translate based on comparing if text =...
Fix detect_lang_interface().
Add report.tex
Fix for efficiency.
Fix some typos.
Fix no-text issue.
Log text to console for debugging;
Check that a correct language is detected befor...
Merge with Dan12's UI. May not work.
Fix typo error.
Clean up some code.
Correct typo error.
Whoops, didn't save the file.
Clean up some code and write detectLanguage tha...
Recommitting files.
Create Trainer class so that thread pool will b...
Change default no. of threads to 4 and print me...
Add ability to continue from previous point by ...
Add even more messages and better statistics + ...
Add new examples, use a threadpool instead and ...
Remove trash file.
Clean up some stuff and let the training be a c...
Update wrong parameter.
Scripts should die gracefully when Python 2 is ...
Update shebang.
Add support for multithreaded training of corpus.
Using the tokeniser is now possible by sane bei...