Papers › Transformer tricks: Precomputing the first layer

Transformer tricks: Precomputing the first layer

20 Feb 2024arXiv:2402.13388archive 2025-07-28

Nils Graef

This micro-paper describes a trick to speed up inference of transformers with RoPE (such as LLaMA, Mistral, PaLM, and Gemma). For these models, a large portion of the first transformer layer can be precomputed, which results in slightly lower latency and lower cost-per-token. Because this trick optimizes only one layer, the relative savings depend on the total number of layers. For example, the maximum savings for a model with only 4 layers (such as Whisper tiny) is limited to 25%, while a 32-layer model is limited to 3% savings. See https://github.com/OpenMachine-ai/transformer-tricks for code and more transformer tricks.

PaperPDFCode

Code

openmachine-ai/transformer-tricks officialmentioned in papermentioned on GitHubpytorchMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

PaLMSPEED

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