Star 历史趋势
数据来源: GitHub API · 生成自 Stargazers.cn
README.md

Adaptive Task Scheduling in Multi‑Core CPUs Using Deep Q‑Networks (DQN)

Research‑grade, config‑driven repo for a learning‑based CPU scheduler:

  • Gym‑like simulator of multi‑core CPUs with DVFS and thermal proxy
  • DQN agent (Double & Dueling), prioritized replay
  • Baselines: RR, EDF, SRTF
  • Reproducible experiments + plotting

Quick Start

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
python -m experiments.run_experiment --config configs/default.yaml --tag dqn_default
python -m src.evaluate --checkpoint artifacts/dqn_default/best.pt --episodes 50
python scripts/plot_results.py --runs artifacts/dqn_default/metrics.json artifacts/baselines/metrics.json

关于 About

Adaptive task scheduling for multi-core CPUs using DQN with DVFS & thermal-aware reward; includes simulator, baselines, training, evaluation, and plotting.

语言 Languages

Python100.0%

提交活跃度 Commit Activity

代码提交热力图
过去 52 周的开发活跃度
2
Total Commits
峰值: 2次/周
Less
More

核心贡献者 Contributors