BanglaLlama: LLaMA for Bangla Language

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zehady, Abdullah Khan, Dipta, Shubhashis Roy, Islam, Naymul, Mamun, Safi Al, Karmaker, Santu
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918155261050880
author Zehady, Abdullah Khan
Dipta, Shubhashis Roy
Islam, Naymul
Mamun, Safi Al
Karmaker, Santu
author_facet Zehady, Abdullah Khan
Dipta, Shubhashis Roy
Islam, Naymul
Mamun, Safi Al
Karmaker, Santu
contents Bangla is a language spoken by approximately 240 million native speakers and around 300 million people worldwide. Despite being the 5th largest spoken language in the world, Bangla is still a "low-resource" language, and existing pretrained language models often struggle to perform well on Bangla Language Processing (BLP) tasks. This paper addresses this gap by: (1) introducing two high-quality translated Bangla-instruction datasets totaling 224k samples - Bangla-Orca (172k) and Bangla-Alpaca (52k); and (2) leveraging these datasets to develop BanglaLlama, an open-source family of Bangla-specific LLMs, consisting of five base and instruct variants. We present our methodology, two large datasets, and comprehensive benchmarking results showcasing the effectiveness of our dataset and model on multiple benchmarks. We believe our proposed datasets and models will serve as the new standard baseline for future research focused on this widely spoken yet "low-resource" language.
format Preprint
id arxiv_https___arxiv_org_abs_2410_21200
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle BanglaLlama: LLaMA for Bangla Language
Zehady, Abdullah Khan
Dipta, Shubhashis Roy
Islam, Naymul
Mamun, Safi Al
Karmaker, Santu
Computation and Language
Artificial Intelligence
Machine Learning
Bangla is a language spoken by approximately 240 million native speakers and around 300 million people worldwide. Despite being the 5th largest spoken language in the world, Bangla is still a "low-resource" language, and existing pretrained language models often struggle to perform well on Bangla Language Processing (BLP) tasks. This paper addresses this gap by: (1) introducing two high-quality translated Bangla-instruction datasets totaling 224k samples - Bangla-Orca (172k) and Bangla-Alpaca (52k); and (2) leveraging these datasets to develop BanglaLlama, an open-source family of Bangla-specific LLMs, consisting of five base and instruct variants. We present our methodology, two large datasets, and comprehensive benchmarking results showcasing the effectiveness of our dataset and model on multiple benchmarks. We believe our proposed datasets and models will serve as the new standard baseline for future research focused on this widely spoken yet "low-resource" language.
title BanglaLlama: LLaMA for Bangla Language
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2410.21200