IndoNLP / nusa-crowd

A collaborative project to collect datasets in Indonesian languages.

Apache License 2.0

260 stars 61 forks source link

Closed SamuelCahyawijaya closed 11 months ago

SamuelCahyawijaya commented 1 year ago

Dataset	nusa_translation_mt
Description	NusaTranslation is a sentence-level datasets which covers 11 local languages in Indonesia. The dataset is human-translated from a part of IndoLEM Sentiment and EmoT dataseets where a native-speaker annotator are requested to translate to the target language given an Indonesian sentence. The data cover ~72k sentence pairs of translation data.
License	CC-BY-NC-SA 4.0

catlaughing commented 1 year ago

self-assign

SamuelCahyawijaya commented 11 months ago

fhudi commented 10 months ago

@SamuelCahyawijaya @catlaughing Issues / bug found on this dataset.

Batak (btk) cannot be loaded
- See lines 103, 113, 123, and 133
- Due to line 17. Maybe forgot to include Indonesian (ind)?
Unable to load complete language pairs
- Basically this code block is never reached.

SamuelCahyawijaya commented 10 months ago

@fhudi : The issue has been fixed now. Kindly try updating to nusacrowd==0.1.2