May 2, 2026·1 min read

FlashAttention — Fast and Memory-Efficient Exact Attention

An IO-aware exact attention algorithm that is 2-4x faster than standard attention and uses 5-20x less memory. FlashAttention has become a foundational building block for training and serving modern large language models.

Anonymous
Anonymous · Community
Agent ready

Ready-to-run agent install

This asset can be installed after the agent chooses its runtime, checks the plan, and runs the matching command.

Native · 98/100Policy: allow
Agent surface
Any MCP/CLI agent
Kind
Skill
Install
Single
Trust
Trust: Community
Entrypoint
FlashAttention Overview
Direct install command
npx -y tokrepo@latest install b4c1950f-45df-11f1-9bc6-00163e2b0d79 --target codex

Run after dry-run confirms the install plan.

An IO-aware exact attention algorithm that is 2-4x faster than standard attention and uses 5-20x less memory. FlashAttention has become a foundational building block for training and serving modern large language models.

Discussion

Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.

Related Assets

Flash Attention — Fast Memory-Efficient Exact Attention for Transformers

Flash Attention is a CUDA kernel library that computes exact scaled dot-product attention 2-4x faster and with up to 20x less memory than standard implementations by using IO-aware tiling to minimize GPU memory reads and writes.

Skills
Script Depot

Apache DataFusion — Fast In-Process SQL Query Engine in Rust

An extensible query engine written in Rust that uses Apache Arrow as its in-memory format, enabling fast analytical SQL queries embeddable in any application.

Skills
Apache Software Foundation

SWC — Super-Fast Rust-Based JavaScript and TypeScript Compiler

SWC is an extensible Rust-based platform for the next generation of fast developer tools. It can be used for both compilation and bundling, offering 20x faster performance than Babel while maintaining compatibility with its ecosystem.

Skills
AI Open Source

Lit — Simple Library for Fast Lightweight Web Components

Lit is a simple library for building fast, lightweight web components. Built by Google on top of the standard Web Components APIs, it provides reactive properties, scoped styles, and a declarative templating system in about 5KB.

Skills
AI Open Source
◈Home🔍Search👤Me