Nvidia has just quietly unveiled its new artificial intelligence model. This artificial intelligence can challenge similar models of OpenAI and Anthropic companies. It could also change Nvidia’s prospects in the market and make the company a serious competitor for artificial intelligence software development companies.
According to VentureBeat, Nvidia’s new artificial intelligence model is called Llama-3.1-Nemotron-70B-Instruct. The company has released it on the Hugging Face artificial intelligence platform without any special background. This model has been able to attract a lot of attention in a short period of time due to its exceptional performance in several benchmarks.
Nvidia’s new artificial intelligence model has an exceptional speed in performing tasks
According to Nvidia, the company’s AI model managed to achieve high scores in various benchmarks, including 85.0 points in Arena Hard benchmark, 57.6 points in AlpacaEval 2 LC and 8.98 points in GPT-4-Turbo MT-Bench. . These points are higher than the points obtained by models such as GPT-4o and Claude 3.5 Sonnet, and show that Nvidia has been able to launch a serious competitor.
Nvidia’s approach to creating the Llama-3.1-Nemotron-70B-Instruct AI model has involved modifying the open-source Llama 3.1 meta-model using advanced training techniques, including reinforcement learning from human feedback (RLHF). This approach allows artificial intelligence to learn from human preferences so that it can ultimately provide more natural and appropriate responses to users.
The release of this artificial intelligence model can be a turning point for Nvidia. Nvidia is currently the absolute king of the market of artificial intelligence chips and computer graphics, and based on this, “Jensen Huang” has been able to stabilize his company’s position in the market. Thanks to artificial intelligence, Nvidia’s value has reached more than 3.4 trillion dollars and has practically eliminated companies like Intel from the pit of competition.
RCO NEWS