COMPARING METRICS FOR LLM MODELS ON AVIATION TEXTS IN SLOVAK LANGUAGE

Abstract

In this article, we present a comprehensive evaluation of three large language models (LLMs) – LLaMA3, Gemma7B, and Aya – focusing on their performance in generating aviation-related texts in the Slovak language. The study employs three widely recognized evaluation metrics: METEOR, BLEU, and ROUGE, to systematically assess the quality of the generated texts.