Open-Source Large Language Models: A Comprehensive Survey

Large Language Models (LLMs) have emerged as strategic in the promotion of natural language processing (NLP) to the extent of having machines translate languages as well as making logical deductions. The role of the open-source team in advancing this progress is that LLMs are now available for development and research. This paper offers an elaborate literature review on some of the most widely used open-source LLMs and their architecture, training approach, and evaluation results. Certainly, the mathematical aspects behind these models are explored, the comparison of the models is given and the problems and perspectives of development of the models are described.