moe skills

23 tagged moe, measured the same way as everything else here.

Browse within: cuda 63inference 34LLM 32llama 32transformer 32PyTorch 31blackwell 31llm-serving 31mixture-of-experts 10INT4 7DeepSpeed 6Efficient Training 6Expert Parallelism 6FP8 6

sgl-project/sglang

Skill Claude CodeCodex needs its repo

Use when choosing the fastest SGLang Diffusion flags for a model, GPU, and VRAM budget.

not rated 37k +256 changed 8d ago A Snyk: passSkillSpector: pass 29 tokens original Apache-2.0

perf-optimize

02

NVIDIA/TensorRT-LLM

Skill Claude Code ✓ vendor

Launch and operate this repo's perf-optimize workflow, which iteratively APPLIES TensorRT-LLM serving optimizations — baseline benchmark at one concurrency or a Pareto curve of them (tok/s/user vs tok/s/gpu), analytical SOL projection on by default (via the internal-perf-sol-analysis skill) sizing the headroom the…

not rated 15k +39 changed 1mo ago A SkillSpector: warn 224 tokens

yolo-master-agent

03

Tencent/YOLO-Master

Skill Claude CodeCodex ✓ vendor

Use when the user wants to run a YOLO-Master task (train/val/predict/track/export/benchmark) or use the Agent Skill dispatcher.

not rated 754 +9 changed 1mo ago A SkillSpector: warn 37 tokens AGPL-3.0

docx

04

dnnyngyen/kimi-agent-internals

Skill Claude CodeCodex

Provided in response to plain-English questions. Maintainer does not claim copyright. Included for research under CC0 1.0. See LICENSE for details. -->.

not rated 330 +1 6mo ago A 0 tokens

At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: