ProTuner：使用Monte Carlo树搜索优化程序（CS DC）---蔡秋纯

我们探索将Monte Carlo树搜索（MCTS）算法应用于一个众所周知的难题：高性能深度学习和图像处理的优化程序。我们在卤化物的基础上建立了我们的框架，并证明了MCTS的性能优于目前最先进的波束搜索算法。与beam搜索不同，MCTS通过贪婪的中间性能比较部分和不太有意义的调度，比较完整的调度，并在做出任何中间调度决策之前进行展望。我们进一步探讨对标准MCTS算法的修改，以及将实时执行时间测量与成本模型相结合。我们的结果表明，MCTS在16个真实的基准上可以优于beam搜索。

原文标题：ProTuner: Tuning Programs with Monte Carlo Tree Search

原文：We explore applying the Monte Carlo Tree Search (MCTS) algorithm in a notoriously difficult task: tuning programs for high-performance deep learning and image processing. We build our framework on top of Halide and show that MCTS can outperform the state-of-the-art beam-search algorithm. Unlike beam search, which is guided by greedy intermediate performance comparisons between partial and less meaningful schedules, MCTS compares complete schedules and looks ahead before making any intermediate scheduling decision. We further explore modifications to the standard MCTS algorithm as well as combining real execution time measurements with the cost model. Our results show that MCTS can outperform beam search on a suite of 16 real benchmarks.

原文作者：Ameer Haj-Ali, Hasan Genc, Qijing Huang, William Moses, John Wawrzynek, Krste Asanović, Ion Stoica

原文地址：https://arxiv.org/abs/2005.13685