Logo Lanfrica

LangBridge Multilingual Translation Dataset

Domaine:

natural language processing

Type de record:

dataset
Créateur:
cod
Hôte:
This dataset contains parallel translation pairs between Yoruba (yor_Latn), Hausa (hau_Latn), Igbo (ibo_Latn), and English (eng_Latn). It was designed for fine-tuning multilingual translation models such as facebook/nllb-200-distilled-600M, in order for it to achieve state of the art performance at translation tasks. 📦 Dataset Summary