菜单

关于 🐙 GitHub
arXiv 提交日期: 2025-12-26
📄 Abstract - TimeBill: Time-Budgeted Inference for Large Language Models

Large Language Models (LLMs) are increasingly deployed in time-critical systems, such as robotics, autonomous driving, embodied intelligence, and industrial automation, where generating accurate responses within a given time budget is crucial for decision-making, control, or safety-critical tasks. However, the auto-regressive generation process of LLMs makes it challenging to model and estimate the end-to-end execution time. Furthermore, existing efficient inference methods based on a fixed key-value (KV) cache eviction ratio struggle to adapt to varying tasks with diverse time budgets, where an improper eviction ratio may lead to incomplete inference or a drop in response performance. In this paper, we propose TimeBill, a novel time-budgeted inference framework for LLMs that balances the inference efficiency and response performance. To be more specific, we propose a fine-grained response length predictor (RLP) and an execution time estimator (ETE) to accurately predict the end-to-end execution time of LLMs. Following this, we develop a time-budgeted efficient inference approach that adaptively adjusts the KV cache eviction ratio based on execution time prediction and the given time budget. Finally, through extensive experiments, we demonstrate the advantages of TimeBill in improving task completion rate and maintaining response performance under various overrun strategies.

顶级标签: llm systems model evaluation
详细标签: time-budgeted inference kv cache eviction response length prediction execution time estimation real-time systems 或 搜索:

TimeBill:面向大语言模型的时间预算推理框架 / TimeBill: Time-Budgeted Inference for Large Language Models


1️⃣ 一句话总结

本文提出了TimeBill,一个新颖的时间预算推理框架,旨在解决大语言模型在严格时间约束(如机器人、自动驾驶等实时系统)下,难以在给定预算内完成推理并保证响应性能的问题,其核心是通过细粒度响应长度预测和端到端执行时间估计来动态调整KV缓存淘汰率,从而在满足时间预算的同时最大化模型输出质量。


2️⃣ 论文创新点

1. 时间预算推理问题形式化

2. 细粒度响应长度预测器

3. 工作负载引导的执行时间估计器

4. 自适应KV缓存淘汰策略

5. 并行化预测与预填充阶段


3️⃣ 主要结果与价值

结果亮点

实际价值


4️⃣ 术语表

源自 arXiv: 2512.21859