Banglanlg And Banglat5: Benchmarks And Resources For Evaluating Low-resource Natural Language Generation In Bangla

Bhattacharjee Abhik, Hasan Tahmid, Ahmad Wasi Uddin, Shahriyar Rifat. Arxiv 2022

[Paper] [Code]    
Applications Has Code Language Modeling Model Architecture Pretraining Methods Transformer

This work presents BanglaNLG, a comprehensive benchmark for evaluating natural language generation (NLG) models in Bangla, a widely spoken yet low-resource language. We aggregate six challenging conditional text generation tasks under the BanglaNLG benchmark, introducing a new dataset on dialogue generation in the process. Furthermore, using a clean corpus of 27.5 GB of Bangla data, we pretrain BanglaT5, a sequence-to-sequence Transformer language model for Bangla. BanglaT5 achieves state-of-the-art performance in all of these tasks, outperforming several multilingual models by up to 9% absolute gain and 32% relative gain. We are making the new dialogue dataset and the BanglaT5 model publicly available at https://github.com/csebuetnlp/BanglaNLG in the hope of advancing future research on Bangla NLG.

Similar Work