AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Inductive-bias-driven Reinforcement Learning For Efficient Schedules in Heterogeneous Clusters

Subho Banerjee, Saurabh Jha, Zbigniew Kalbarczyk, Ravishankar Iyer

ICML 2020 · Reinforcement learning › Visual and representation RL · approach: RL with tree search for discovery · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Encode architectural resource dependencies in a Bayesian belief model instead of using a model-free RNNSymphony scheduling · Scheduling workloads on heterogeneous clusterssample efficiency: ~2.2× worse than SymphonyImproved
Use sampling-based Bayesian gradient computation instead of full belief-propagation inference (ablation)Symphony training · Training to reach 30% mean loss relative to oraclesample efficiency: at least 4.3×Improved
Use task batching rather than scheduling single tasks (ablation)Symphony · Scheduler tail latencytail latency: 12×Improved
Set task batch size to 128 (ablation)Symphony · Task batchingoracle-normalized workload latency: 128 tasks per batchImproved
Use random task assignment rather than Symphony's learned scheduling policyScheduling workloads on heterogeneous clusters · Scheduling latency and workload latencyworkload latency: 3.6× lower median latencyMixed
Account for system topology and low-level resource contention in schedulingSymphony · Three workloads, 500 executionsoracle-normalized kernel runtime: 32%Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy