|
--- |
|
license: cc-by-sa-4.0 |
|
--- |
|
RwilaNMT is a parallel dataset composed of 8,000 sentences in Swahili and Rwila. It is intended to be used for fine-tuning Neural Machine Translation models and Large Language Models for Rwila. |
|
|
|
Rwila is a low-resource Bantu language spoken in Tanzania. |
|
|
|
This dataset was created and curated using a proprietary method developed by XRI Global which ensures coverage of a conceptual space when doing data collection. This method was developed in order to create the fastest and most affordable way to collect pristine in-domain data for low-resource languages optimized for fine-tuning language models. |
|
|
|
The contributors gave proper consent, were hired by a local agency, and were compensated fairly. |
|
|
|
For data collection, we utilized our mobile data collection app, Echonet, and our custom translation management system. |
|
|
|
The domain is somewhat generalized and the dataset is most effective at literary and narrative texts. It is less capable in other domains such as technical, scientific, or colloquial. |
|
|
|
If you are interested in creating ideal datasets for new languages, reach out to us at [email protected] |
