A Flexible Approach to Autotuning Multi-Pass Machine Learning Compilers

Search-based techniques have been demonstrated effective in solving complex optimization problems that arise in domain-specific compilers for machine learning (ML). Unfortunately, deploying such techniques in production compilers is impeded by two limitations. First, prior works require factorizatio...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	2021 30th International Conference on Parallel Architectures and Compilation Techniques (PACT) S. 1 - 16
Hauptverfasser:	Phothilimthana, Phitchaya Mangpo, Sabne, Amit, Sarda, Nikhil, Murthy, Karthik Srinivasa, Zhou, Yanqi, Angermueller, Christof, Burrows, Mike, Roy, Sudip, Mandke, Ketan, Farahani, Rezsa, Wang, Yu Emma, Ilbeyi, Berkin, Hechtman, Blake, Roune, Bjarke, Wang, Shen, Xu, Yuanzhong, Kaufman, Samuel J.
Format:	Tagungsbericht
Sprache:	Englisch
Veröffentlicht:	IEEE 01.09.2021
Schlagworte:	autotuning compiler Internet Layout Machine learning Production Search problems Tensors Training
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!