Datasets:

---

google

/

WaxalNLP

like 219

Follow

Google 51.2k

Hi,

Thanks for checking out the dataset and pointing out this issue. Our team will work on correcting the misalignment and alert you once the new version is pushed out.

Best,

Perry

Same for Bambara.

Thank you @nairaxo and @Tonic we're checking it out and will update you as soon as possible.

thank you google team for being so available for this opensourced and community-powered project - i'll take you up on your offer to be notified as i've been wanting something like this for approx. half a decade , take your time , a little more waiting wont actually hurt :-)

Thank @Adiack and the team for this great work! It's just impressive!

fingers crossed we're not forgotten here :-)

the thirst continue for wol shards that are correctly parsed 🤗🤗

Hello everyone,

I've just fixed the alignment issues for the Wolof language (cc @Tonic ) and published the corrected dataset in the GalsenAI repository. I've also shared the process and the notebook I used so that people like @nairaxo can fix the same issue for their languages. I also identified some corrupted files and removed them, which reduced the dataset's original size.

I hope this will be useful to you while you wait for the official fixes to be released by @perrynelson and @Adiack 's team.

Huge thank you to @derguene and the GalsenAI community! The team really appreciates the effort and contribution; it will surely be useful to the community!