访问集至关重要：为可扩展的权重空间模型融合进行专家读取预算

摘要

权值空间模型合并通常被形式化为检查点上的代数操作，但在大语言模型（LLM）规模下，限制性资源往往是必须读取的专家权值集合。本文提出MergePipe，一种预算感知的执行层，将LLM合并问题转化为专家访问集合问题：在给定合并算子与共享权值坐标系下的检查点族时，在显式I/O预算下选择需要访问的专家增量块。MergePipe对参数块建立索引、构建确定性访问计划，并通过可回放清单执行诱导出的有预算合并。该计划本身具备预算合理性，在满预算下可恢复全量读取合并；对于固定系数加法算子，遗漏更新误差受遗漏增量范数的约束。在Qwen和Llama的合并工作负载中，MergePipe将专家读取I/O降低至多一个数量级，并实现最高11倍加速。代表性预算扫描显示，与全量读取合并相比，参数偏差为O(10^{-3})，且下游基准测试无单调退化。

English

Weight-space model merging is usually formulated as an algebraic operation on checkpoints, yet at LLM scale the limiting resource is often the set of expert weights that must be read. We introduce MergePipe, a budget-aware execution layer that casts LLM merging as an expert access-set problem: given a merge operator and a checkpoint family in a shared weight coordinate system, choose which expert delta blocks to access under an explicit I/O budget. MergePipe indexes parameter blocks, builds deterministic access plans, and executes the induced budgeted merge with replayable manifests. The plan is budget-sound by construction and recovers the full-read merge at full budget; for fixed-coefficient additive operators, the omitted-update error is bounded by the norm of omitted deltas. Across Qwen and Llama merging workloads, MergePipe reduces expert-read I/O by up to an order of magnitude and achieves up to 11times speedups. Representative budget sweeps show O(10^{-3}) parameter deviation from full-read merges and no monotonic degradation on downstream benchmarks.