2019
Learning Generalizable Device Placement Algorithms for Distributed Machine Learning
NeurIPS 2019poster
We present Placeto, a reinforcement learning (RL) approach to efficiently find device placements for distributed neural network training.
2 accepted papers
We present Placeto, a reinforcement learning (RL) approach to efficiently find device placements for distributed neural network training.
We consider reinforcement learning in input-driven environments, where an exogenous, stochastic input process affects the dynamics of the system. Input processes arise in many applications, including queuing systems, robotics control with disturbances, and object tracking. Since the state dynamics a…