Fanar 2.0：阿拉伯生成AI堆栈
Fanar 2.0: Arabic Generative AI Stack
AI总结：
Fanar 2.0在资源受限条件下实现卓越性能，通过数据质量优先、持续预训练和模型融合，提升阿拉伯语知识、语言、方言及英语能力，同时引入多项新功能。
摘要：
我们介绍了Fanar 2.0，卡塔尔阿拉伯语中心生成AI平台的第二代。主权是首要设计原则：从数据管道到部署基础设施，均由QCRI哈马德·本·卡西姆大学设计和运营。Fanar 2.0是资源受限下的卓越体现：在256块NVIDIA H100 GPU上运行，阿拉伯语仅占网络数据的约0.5%，尽管有4亿名母语者。Fanar 2.0采用数据质量优先于数量、定向持续预训练和模型融合策略，在受限条件下取得显著进展。核心是Fanar-27B，基于Gemma-3-27B骨架在1200亿高质量令牌的三大数据食谱上持续预训练。尽管预训练令牌数量仅为Fanar 1.0的1/8，但其在阿拉伯语知识（+9.1分）、语言（+7.3分）、方言（+3.5分）和英语能力（+7.6分）方面均取得显著改进。除核心LLM外，Fanar 2.0还引入了丰富的新增功能。FanarGuard是先进的4B双语审核过滤器，用于阿拉伯语安全和文化对齐。语音家族Aura获得用于小时级音频的长形式ASR模型。Oryx视觉家族增加阿拉伯语意识的图像和视频理解，以及基于文化基础的图像生成。代理工具调用框架支持多步骤工作流。Fanar-Sadiq利用多代理架构生成伊斯兰内容。Fanar-Diwan提供古典阿拉伯语诗歌生成。FanarShaheen提供LLM驱动的双语翻译。重新设计的多层协调器通过意图感知路由和纵深安全验证协调所有组件。总体而言，Fanar 2.0证明了主权、资源受限的AI开发可以产生与大规模开发相媲美的系统。
英文摘要：
We present Fanar 2.0, the second generation of Qatar's Arabic-centric Generative AI platform. Sovereignty is a first-class design principle: every component, from data pipelines to deployment infrastructure, was designed and operated entirely at QCRI, Hamad Bin Khalifa University. Fanar 2.0 is a story of resource-constrained excellence: the effort ran on 256 NVIDIA H100 GPUs, with Arabic having only ~0.5% of web data despite 400 million native speakers. Fanar 2.0 adopts a disciplined strategy of data quality over quantity, targeted continual pre-training, and model merging to achieve substantial gains within these constraints. At the core is Fanar-27B, continually pre-trained from a Gemma-3-27B backbone on a curated corpus of 120 billion high-quality tokens across three data recipes. Despite using 8x fewer pre-training tokens than Fanar 1.0, it delivers substantial benchmark improvements: Arabic knowledge (+9.1 pts), language (+7.3 pts), dialects (+3.5 pts), and English capability (+7.6 pts). Beyond the core LLM, Fanar 2.0 introduces a rich stack of new capabilities. FanarGuard is a state-of-the-art 4B bilingual moderation filter for Arabic safety and cultural alignment. The speech family Aura gains a long-form ASR model for hours-long audio. Oryx vision family adds Arabic-aware image and video understanding alongside culturally grounded image generation. An agentic tool-calling framework enables multi-step workflows. Fanar-Sadiq utilizes a multi-agent architecture for Islamic content. Fanar-Diwan provides classical Arabic poetry generation. FanarShaheen delivers LLM-powered bilingual translation. A redesigned multi-layer orchestrator coordinates all components through intent-aware routing and defense-in-depth safety validation. Taken together, Fanar 2.0 demonstrates that sovereign, resource-constrained AI development can produce systems competitive with those built at far greater scale.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
