Methods › General › Learning Rate Schedules › Linear Warmup With Cosine Annealing

Linear Warmup With Cosine Annealing

3,797 papers tagged archive 2025-07-28

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

Linear Warmup With Cosine Annealing is a learning rate schedule where we increase the learning rate linearly for n updates and then anneal according to a cosine schedule afterwards.

Papers archive 2025-07-28

30 shown of 3,797, 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

20 shown of 994 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
Language Modelling796
Language Modeling587
Question Answering300
Large Language Model296
Text Generation285
Retrieval202
Sentence167
In-Context Learning162
Prompt Engineering133
Few-Shot Learning111
Decoder108
Code Generation106
Decision Making100
Text Classification90
RAG89
Translation89
Math87
Retrieval-augmented Generation87
text-classification82
Machine Translation80

Usage over time archive 2025-07-28

Papers per year tagged with Linear Warmup With Cosine Annealing: 2018 to 2025, peak 1,372 1,372 0 2018: 1 paper 2018 2019: 82 papers 2019 2020: 164 papers 2020 2021: 274 papers 2021 2022: 376 papers 2022 2023: 1172 papers 2023 2024: 1372 papers 2024 2025: 356 papers 2025
Papers per year the archive tags with this method, by the paper's archive date (3,797 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

Learning Rate Schedules

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