[EMNLP'23, ACL'24] To speed up LLMs' inference and enhance LLM's perceive of key information, compress the prompt and KV-Cache, which achieves up to 20x compression with minimal performance loss.
An agentic skills framework & software development methodology that works.
Skills for Real Engineers. Straight from my .agents directory.
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
DigitalPlat FreeDomain: Free Domain For Everyone
Python tool for converting files and office documents to Markdown.