Closed SamuelCahyawijaya closed 1 year ago
This is amazing ✨
Because you have traced the currently implemented dataloader and fixed them, I do think this should be merged before other dataloader is merged, and the new ones could follow the new templates you give too.
Thank you @SamuelCahyawijaya !
Will the github-action bot tests made by @christianwbsn be affected by this PR by the way?
Will the github-action bot tests made by @christianwbsn be affected by this PR by the way?
I think as long as we update the requirements.txt
and test_nusantara.py
, it shouldn't be a problem. But might need to do adjustment for source only kind of test
/test dataset=emot
Check test log here: https://github.com/IndoNLP/nusa-crowd/actions/runs/3103308786
/test dataset=emot
Check test log here: https://github.com/IndoNLP/nusa-crowd/actions/runs/3103716716
/test dataset=emot
Check test log here: https://github.com/IndoNLP/nusa-crowd/actions/runs/3103770784
Will the github-action bot tests made by @christianwbsn be affected by this PR by the way?
Good point @bryanwilie! There is an error, somehow the runner find nusantara/nusa_datasets/*
instead of nusacrowd/nusa_datasets/*
. I am currently asking @christianwbsn on Slack regarding this issue, since I am not really familiar with this github runner.
/test dataset=emot
Check test log here: https://github.com/IndoNLP/nusa-crowd/actions/runs/3103904153
This is amazing sparkles
Because you have traced the currently implemented dataloader and fixed them, I do think this should be merged before other dataloader is merged, and the new ones could follow the new templates you give too.
Thank you @SamuelCahyawijaya !
Okay, so... I didn't see this comment before... So I already merged a few PRs without knowing, sorry about that. I guess I can merge this and change the import paths for the newly accepted dataloaders.
Installing
nusacrowd
package:This is to make running experiment more convenient. For later release phase, if agreed, we can also push this to pypi and/or add the installation guidelines to the readme