TIMEGATE: 资源约束下可持续的限时促销门控用于持续ML适应
TIMEGATE: Sustainable Time-Boxed Promotion Gates for Continual ML Adaptation Under Resource Constraints
AI总结：
提出TIMEGATE策略层，通过预算时间、标注、训练和评估来管理持续ML适应，实现评估计算节省且无静默错误促销。
摘要：
随着机器学习（ML）系统向持续适应演进，每个重新训练周期都会消耗计算、标注和能源。我们引入TIMEGATE，一个通过预算时间、标注、训练和评估来管理适应的策略层。TIMEGATE发出一个度量可用性信号M，用于部分与完整评估决策。我们验证：（i）在Adult表格数据上，标注优于训练2.3倍；（ii）它迁移到LLaMA-3.1-8B + QLoRA在SST-2上（准确率从0.80到0.96；35/36次运行中M=1）；（iii）M具有信息性，28单元敏感性显示在严格阈值下M降至0.81；（iv）100周期模拟实现66%的评估计算节省，且无静默错误促销；（v）在单个H200上，LLaMA的10%切片评估使用89%更少的挂钟时间和能源（比率一致至0.2%）。
英文摘要：
As machine learning(ML) systems evolve to continual adaptation, each re-training cycle uses compute, annotation, and energy. We introduce TIMEGATE, a policy layer managing adaptation by budgeting time, labeling, training, and evaluation. TIMEGATE emits a metric-availability signal M for partial vs. full-evaluation decisions. We validate: (i) labeling outperforms training by 2.3x on Adult tabular; (ii) it transfers to LLaMA-3.1-8B + QLoRA on SST-2 (accuracy 0.80 to 0.96; M =1 in 35/36 runs); (iii) M is informative, 28-cell sensitivity shows M drops to 0.81 at tight thresholds; (iv) 100-cycle simulation achieves 66% evaluation-compute savings with no silent mis-promotions; (v) 10%-slice evaluation on LLaMA uses 89% less wall-clock and energy on a single H200 (ratios agree to 0.2%).
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。
