AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

DreamShard: Generalizable Embedding Table Placement for Recommender Systems

Daochen Zha, Louis Feng, Qiaoyu Tan, Zirui Liu, Kwei-Herng Lai, Bhargav Bhushanam, Yuandong Tian, Arun Kejariwal, Xia Hu

NeurIPS 2022 · Reinforcement learning › Visual and representation RL · approach: RL with tree search for discovery · 7 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
remove pooling factor feature (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 47.2Hurt
remove cost features from state (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 46.3Hurt
remove dimension feature (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 40.9Hurt
remove hash size feature (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 40.6No clear effect
remove table size feature (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 40.6No clear effect
remove distribution feature (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 40.5No clear effect
add RNN on device representation in policy (ablation)DreamShard · DLRM-50 (4) testoverall cost (ms): 40.4 -> 40.5No clear effect
apply DreamShard trained on DLRM-20 (4) to DLRM-100 (4) without fine-tuningDreamShard · DLRM-100 (4) testoverall cost (ms): 77.8 -> 77.9No clear effect
apply DreamShard trained on DLRM-100 (4) to DLRM-20 (4)DreamShard · DLRM-20 (4) testoverall cost (ms): 17.6 -> 17.7No clear effect
apply DreamShard trained on DLRM-20 (2) to DLRM-20 (4)DreamShard · DLRM-20 (4) testoverall cost (ms): 17.6 -> 17.8No clear effect
use DreamShard instead of Lookup-based expert on large 8-GPU taskplacement · DLRM-200 (8) testoverall cost (ms): 38.0 -> 36.4Improved
use DreamShard instead of Dim-based expert on hardest Prod taskplacement · Prod-80 (8) testoverall cost (ms): 41.9 -> 35.2Improved

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