研究人员开发了 FMplex，这是一个新颖的系统，旨在通过将基础模型 (FM) 视为虚拟化底层来优化其服务。这种方法允许多个下游任务共享单个物理 FM 实例，从而减少内存浪费并摊销与批处理和加载相关的成本。FMplex 可实现任务特定的扩展和隔离，同时提高效率，其延迟显著降低和任务托管能力增强得到了证明。 AI
影响 优化基础模型的部署，可能降低人工智能应用的基础设施成本并提高延迟。
排序理由 该集群包含一篇详细介绍基础模型服务新系统的研究论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2606.09643v1 Announce Type: cross Abstract: Foundation models (FMs) are increasingly used as backbones for downstream tasks across language, vision, time-series, and multimodal applications. Yet existing model-serving systems deploy each customized task as an independent mo…
Foundation models (FMs) are increasingly used as backbones for downstream tasks across language, vision, time-series, and multimodal applications. Yet existing model-serving systems deploy each customized task as an independent model instance, thereby replicating heavyweight back…