claudeindex
Plugin

tune-ik-llama

Six-phase procedure for discovering optimal --n-cpu-moe, -rtr, and context settings on a VRAM-constrained NVIDIA GPU. Detects auto-spill, applies -rtr only when needed, pushes context to native max. NVIDIA only (CUDA).

Installation

1

Add the marketplace

/plugin marketplace add pmaeria/tune-ik-llama
2

Install plugins

/plugin

Run these commands in Claude Code to add this plugin to your environment. The marketplace must be added before you can install its plugins.

Repository & Links

Plugin Source

View Plugin Code

GitHub Repository

pmaeria/tune-ik-llama

Details & Metadata