The series contains four models with 1.5, 7, 14 and 32 billion parameters. All of them are created on the basis of a large model of 671 Milliard Deepseek, but they are much more compact, which allows you to run them on ordinary game graphics cards.
For learning, Nvidia used 5 million tasks in the field of mathematics, science and programming created using the NEMO Skills platform.
Models have been examined only with the use of controlled learning – without the use of RLHF, which has made them a suitable basis for further research.
According to the test results, the largest 32B model reaches 89.2 points in AIME24 and 73.8 Olympics and the youngest 1.5B at HMMT. In parallel, the performance of the 32B model can be compared or even over Openai, in the Gensect mode, which produces a few answers and chooses the best.
All four NVIDIA models are already available on the face of hugging for loading and local use.
Source: Ferra
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
