On October 15, Nvidia quietly introduced its new artificial intelligence model, Nemotron, which has already surpassed such powerful systems as GPT-4 by OpenAI and Claude-3 by Anthropic. The model, called Llama-3.1-Nemotron-70B-Instruct, is an enhanced version of the open-source Llama-3.1-70B-Instruct model by Meta, renowned for its performance.
According to Nvidia’s developers, Nemotron became the leader in tests on the Chatbot Arena platform. The model received high marks in the “complex” tasks section, placing it at the top of the ranking among competitors. This result was made possible by carefully selected datasets, specialized fine-tuning, and the use of Nvidia’s advanced hardware computing capabilities.
Despite having fewer parameters (70 billion) compared to giant models like GPT-4 (1 trillion parameters), Nemotron demonstrates strong efficiency and usefulness in its responses, making it a promising tool in the field of AI.
