Comparative Analysis for Open-Source Large Language Models

Large Language Models (LLMs) have significantly advanced the field of Natural Language Processing (NLP), demonstrating exceptional performance across diverse language tasks such as content summarization, sentiment analysis, and conversational AI. The advent of these models has profoundly impacted human-computer interaction research, marking the onset of a new era in the field. The pioneering model, ChatGPT, was introduced by OpenAI in November 2022, catalyzing the development of other commercial tools like Bing Chat and Google Bard. Subsequently, the emergence of open-source LLMs has democratized access to these powerful tools, enabling end-users to deploy them internally with relative ease. As the landscape of open-source LLMs continues to expand and evolve, researchers and practitioners are presented with a plethora of choices for NLP applications. This paper presents a comparative analysis for various open-source LLMs, assessing their unique features, strengths, and limitations. Our focus of comparisons will include licensing, training methods, computing resources needed, available application programming interfaces (APIs), robustness, and bias.