L4S-LLM Experience Pool

该数据集是从基于FreeBSD的虚拟机测试床上收集的，旨在利用大型语言模型来增强活动队列管理。数据包含了多种度量指标，如损失、准确度、奖励、GPU/CPU利用率、计算时间、内存使用情况以及总的训练/测试时间。该数据集还涉及在多种拥塞控制算法下，从多个客户端和路由器收集的数据，其任务是针对低延迟、低损失和可扩展吞吐量（L4S）进行活动队列管理。

This dataset was collected from a FreeBSD-based virtual machine testbed, with the objective of leveraging large language models (LLMs) to enhance active queue management. It includes a variety of metrics such as loss, accuracy, reward, GPU/CPU utilization, computation time, memory usage, and total training and testing time. Additionally, it contains data gathered from multiple clients and routers under various congestion control algorithms, targeting active queue management for low latency, low packet loss, and scalable throughput (L4S).