Methods › General › Data Parallel Methods › Accordion

Accordion

4 papers tagged archive 2025-07-28

Introduced by Saurabh Agarwal et al. in Accordion: Adaptive Gradient Communication via Critical Learning Regime Identification

archive 2025-07-28 Description, source and code snippet are the archive's method entry.

Accordion is a gradient communication scheduling algorithm that is generic across models while imposing low computational overheads. Accordion inspects the change in the gradient norms to detect critical regimes and adjusts the communication schedule dynamically. Accordion works for both adjusting the gradient compression rate or the batch size without additional parameter tuning.

PaperSource

Papers archive 2025-07-28

4 shown of 4, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.

Tasks archive 2025-07-28

2 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.

TaskPapers
3D geometry1
Quantization1

Usage over time archive 2025-07-28

Papers per year tagged with Accordion: 2020 to 2025, peak 2 2 0 2020: 1 paper 2020 2021: 0 papers 2021 2022: 0 papers 2022 2023: 2 papers 2023 2024: 0 papers 2024 2025: 1 paper 2025
Papers per year the archive tags with this method, by the paper's archive date (4 dated). Bars are counts, not a trend claim.

Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).

Categories archive 2025-07-28

Data Parallel MethodsDistributed Methods

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections