LLM KV cache compression made easy - View it on GitHub