A crafty data engineer put 14 AI bots against each other in Street Fighter III matches to see which model is the best. Some weeks ago, French coder Stan Girard introduced LLM Colosseum - an open-source test application that allows users to evaluate the quality of Large Language Models (LLMs) and rank them. In his initial test, Girard put OpenAI against MistralAI models against each other to see which model performs best. A video showing off this benchmark tool for LLMS has been included below (courtesy of tech enthusiast Matthew Berman): Following this initial benchmark using LLM Colosseum, Amazon engineer […]
Read full article at https://wccftech.com/ai-battle-heres-14-llms-fighting-each-other-in-street-fighter-iii/
from Wccftech https://ift.tt/04RujqK
0 Comments