AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

AMP: Automatically Finding Model Parallel Strategies with Heterogeneity Awareness

Dacheng Li, Hongyi Wang, Eric Xing, Hao Zhang

NeurIPS 2022 · Architectures and efficiency › RNNs and normalization · approach: Neural scaling laws · 12 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
use AMP instead of Megatron heuristicsGPT-2 · heterogeneous clustermin iteration time: 1.97 -> 1.28 s (1.54x speedup)Improved
use AMP instead of Megatron heuristicsTransGAN generator · heterogeneous modelmin iteration time: 1.89 -> 1.07 s (1.77x)Improved
use AMP instead of Megatron on homogeneous setupGPT-2 · homogeneous clustermin iteration time: 1.32 -> 1.20 sImproved
use DP-based layer assignment instead of uniform assignment (ablation)AMP best strategy · heterogeneous modelmin iteration time: 1.25 -> 1.07 s (1.17x)Improved
use DP-based layer assignment instead of parameter-balanced assignment (ablation)AMP best strategy · heterogeneous modelmin iteration time: 2.19 -> 1.07 sImproved
add simulated annealing device placement search (ablation)AMP · homogeneous clustermin iteration time: 1.20 -> 1.57 sHurt
add simulated annealing device placement search (ablation)AMP · heterogeneous clustermin iteration time: 1.28 -> 1.46 sHurt
cost-model guided search instead of random strategy selection (ablation)AMP · heterogeneous clustermean iteration time: 3.04 -> 1.73 sImproved

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