Updated 1d ago
by alesha-pro
inference-optimization-playbook
97 tested hypotheses on making huge MoE LLMs fast on consumer GPUs (4x RTX 3090 + experts in RAM): what worked, what failed and why. A playbook for people and a ready skill for coding agents.
54 GitHub stars (54)
+54 in 2 days
- 1 skill (1 skills found in its SKILL.md files)
- 9 forks (9 forks)
- 1d ago last update (Last updated 1d ago)
Install
Skills CLI
npx skills add alesha-pro/inference-optimization-playbook
Asks which agents to install it for. Add -g to install it for every project.
Claude Code
git clone https://github.com/alesha-pro/inference-optimization-playbookcp -r "inference-optimization-playbook" ~/.claude/skills/inference-optimization-playbook
Copies the inference-optimization-playbook skill into ~/.claude/skills/ for every project (use .claude/skills/ for one project).
Plugin
/plugin marketplace add alesha-pro/inference-optimization-playbook
Run inside Claude Code if the repository ships a plugin marketplace, then pick plugins from /plugin.
Skills can run code on your machine — read the SKILL.md first.
What it does
In the author's own words, from each skill's SKILL.md.
-
inference-optimization-playbook
Use when the task is to make local LLM inference faster or to explain why it is slow, especially large MoE models that do not fit in VRAM: expert offload to RAM, hot expert placement, live expert cache, CPU lane vs PCIe copy, multi-GPU (layer split, tensor parallel, expert parallel, all-reduce), prefill and re-read speed, speculative decoding and drafters, quant choice for speed, expert pruning,…
SKILL.md on GitHub (opens in a new tab)
No skill matches that filter.
Review before you install
A skill is a set of instructions and scripts your agent will follow, and it can run code on your machine. Read the SKILL.md and any scripts first — this listing is not a security review.