Skip to content

Feature/support swedish language#2

Open
MariaSL wants to merge 20 commits intorloganiv:masterfrom
MariaSL:feature/support-swedish-language
Open

Feature/support swedish language#2
MariaSL wants to merge 20 commits intorloganiv:masterfrom
MariaSL:feature/support-swedish-language

Conversation

@MariaSL
Copy link
Copy Markdown

@MariaSL MariaSL commented Sep 9, 2019

Description

This PR enables generating data for both English & Swedish languages (and potentially other languages).

Major changes:

  • Introduced a language parameter to scripts thats parse wikitext, build Wiki DB, process html, build Relation DB, build Alias DB, annotate;
  • Added scripts for coreference resolution and nel;
  • Moved several functions (extract_redirect, generate_instances) to utils;
  • Updated requirements;
  • Updated README with step-by-step explanation of data generation process.

All tests are passing

@rloganiv
Copy link
Copy Markdown
Owner

Hi @MariaSL,

This looks fantastic! Thank you so much for pushing these changes - in addition to the support for multiple languages, it also is much cleaner!

In order to merge this work into the master branch, I need to ensure that the English Linked WikiText-2 dataset can still be reproduced. I notice some constants have been replaced/introduced which only work for Swedish (e.g. the list of months in render.py). Can you enable "add edits from maintainers" so I can address these issues / add some more tests?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants