TigerBot: An Open Multilingual Multitask LLM

TL;DR

TigerBot introduces a family of large, open-source multilingual LLMs with significant performance improvements over existing models, emphasizing democratization and real-world applicability.

Contribution

The paper presents a new family of open-source multilingual LLMs with advanced training techniques and benchmark performance, expanding the capabilities of accessible large language models.

Findings

Achieved up to 6% performance gain in English and 20% in Chinese over SOTA models.

Models outperform existing open-source LLMs on major benchmarks.

Released models to promote democratized access to advanced LLM technology.

Abstract

We release and introduce the TigerBot family of large language models (LLMs), consisting of base and chat models, sized from 7, 13, 70 and 180 billion parameters. We develop our models embarking from Llama-2 and BLOOM, and push the boundary further in data, training algorithm, infrastructure, and application tools. Our models yield meaningful performance gain over SOTA open-source models, e.g., Llama-2, specifically 6% gain in English and 20% gain in Chinese. TigerBot model family also achieves leading performance in major academic and industrial benchmarks and leaderboards. We believe that TigerBot represents just a snapshot of lightning-fast progression in LLM open-source community. Therefore, we are thrilled to give back by publicly releasing our models and reporting our approach behind, with additional emphases on building SOTA LLMs in a democratized way and making LLMs of use in real-world applications.