Inside a Frontier Open Model: Nemotron 3 Ultra Explained ft. NVIDIA's Chris Alexiuk
today we are taking a deeper look at how you actually build a frontier open AI model with Chris Alexiuk, senior product research engineer at NVIDIA! the nemotron team pretrained a 550B parameter hybrid mamba-transformer on 20 trillion tokens in 4-bit, built the whole thing around one idea (faster models are smarter models) and in nvidia fashion released everything openly: weights, datasets, RL environments, recipes and even their failed training runs! 👉 Chris Website: https://alexi.uk 👉 Twitter: https://x.com/llm_wizard 👉 Paper: arXiv: https://arxiv.org/abs/2606.15007 👉 Blog: https://de
Discussion
Rejoignez la communauté pour réagir et poser vos questions.
Aucun message pour l'instant. Lancez la discussion !