Measurement-Driven Sub-Network Selection for On-Premise Retrieval-Augmented Factory Agents· 用于工厂代理的测量驱动子网络选择
On-premise assistants can give factory workers conversational access to machine documentation, but models capable of the task rarely fit shop-floor hardware. We show that after structural compression and retrieval-grounded adaptation, model size is no longer a reliable predictor of adapted answer quality: general capability falls almost linearly with parameter count, while judged retrieval-augmented answer quality does not. We therefore treat deployment as a post-adaptation selection problem, committing one sub-network per device on judged answer quality and measured on-device throughput under a configurable general-capability floor and memory budget; rules that optimize size, speed, or quality alone each give up capability or throughput. A weight-shared supernetwork trained with sandwich-
研究工厂代理助手的子网络选择方法，提升小型设备上的问答性能。
- 核心方法
- 通过结构压缩和检索增强适应，提出基于测量的子网络选择策略，在一般能力和内存预算的限制下优化模型大小、速度和质量。
- 适合谁读
- 研究者、工程师
- 要解决的问题
- 工厂代理助手模型通常过大，难以在车间的小型硬件上运行，如何在有限资源下选择合适的子网络以保持较高性能。
- 关键实验
- 实验包括不同压缩比下的模型性能测试，以及实际设备上的线性与非线性质量下降对比。
- 主要贡献
- 展示了模型大小与适应后问答质量之间的非线性关系，提出了一种新的子网络选择方法，有效平衡了性能与资源消耗。
- 意义与局限
- 该研究有助于提高工厂车间等资源受限环境中的智能助手性能，但模型选择策略的泛化能力需进一步验证。