optimizing-attention-flash

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

Install

KitScout test notes

Not tested by KitScout yet. When our harness runs it, pros, cons and the tasks we tried appear here, with the date and evidence.

Similar tools

• reverse-engineer-anything 456,653 ★
  Reverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and b
• scaffold-exercises 323,485 ★
  Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use
• wait-what 323,485 ★
  Stop. That last message did not land: re-pitch it.
• template-skill 193,970 ★
  Replace with description of the skill and when Claude should use it.
• ponytail-debt 190,107 ★
  List every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing