TigerLLM - A Family of Bangla Large Language Models
This addresses the linguistic disparity in LLMs for Bangla speakers, providing a new baseline for future development.
The authors tackled the lack of high-performance Bangla large language models by introducing TigerLLM, a family of models that surpass all open-source alternatives and outperform larger proprietary models like GPT3.5 across standard benchmarks.
The development of Large Language Models (LLMs) remains heavily skewed towards English and a few other high-resource languages. This linguistic disparity is particularly evident for Bangla - the 5th most spoken language. A few initiatives attempted to create open-source Bangla LLMs with performance still behind high-resource languages and limited reproducibility. To address this gap, we introduce TigerLLM - a family of Bangla LLMs. Our results demonstrate that these models surpass all open-source alternatives and also outperform larger proprietary models like GPT3.5 across standard benchmarks, establishing TigerLLM as the new baseline for future Bangla language modeling.