Meta发射用于图像和文本处理的开放源多式AI 模型Llama3.2。 Meta launches open-source multimodal AI model Llama 3.2 for image and text processing.

Meta已推出Llama 3.2号,这是其第一个能够处理图像和文字的开放源码多式联运AI模型。 Meta has launched Llama 3.2, its first open-source multimodal AI model capable of processing images and text. 它包括为各种硬件设计的具有110亿和900亿参数的愿景模型和具有10亿和30亿参数的轻量级文本模型。 It includes vision models with 11 billion and 90 billion parameters, and lightweight text models with 1 billion and 3 billion parameters, designed for diverse hardware. Llama 3.2旨在增强AI在诸如扩大现实和文件分析等领域的应用,在形象识别任务方面提供竞争性业绩,以对抗OpenAI和人类学等竞争对手。 Llama 3.2 aims to enhance AI applications in areas like augmented reality and document analysis, offering competitive performance in image recognition tasks against rivals like OpenAI and Anthropic.