Skip to content
반응형

Inference Performance2

Google DeepMind Unleashes Gemma 4 12B: Open-Source Multimodal AI Revolutionizes On-Device & Personal Computing for Laptops 🚀 Key TakeawaysGoogle DeepMind has unveiled Gemma 4 12B, a new 12 billion parameter multimodal AI model released under the Apache 2.0 license, specifically targeting the acceleration of personal and on-device AI markets.This intermediate model features an integrated (Encoder-Free) multimodal structure, allowing the language model to directly process images and audio, and is the first medium-siz.. 2026. 7. 7.
NVIDIA Nemotron 3 Ultra: Open-Source LLM Empowers Next-Gen AI Agents with 5x Faster Inference & Advanced Reasoning 🚀 Key TakeawaysNVIDIA Nemotron 3 Ultra is a new open-source, open-weight Mixture-of-Experts (MoE) model designed for AI agents to perform complex, long-duration tasks.It features 550 billion total parameters (55 billion activated for inference), supports up to 1 million tokens of context, and utilizes a hybrid Mamba-Attention architecture.The model offers up to 5 times faster inference performa.. 2026. 7. 7.
반응형