NVIDIA/TensorRT-LLM/trtllm-moe-develop
>- Review, design, and refactor TensorRT-LLM PyTorch MoE code for architecture fit, clean code, maintainability, and testability.
How to get this skill
Agent Skill by NVIDIA. Download or clone it, then install it in your agent.
Setup & Installation
- Clone the repository:
git clone https://github.com/NVIDIA/skills.git - Copy the skill folder (which contains
SKILL.md) into your agent skills folder, e.g..claude/skills/. - Restart or reload the agent to auto-discover the skill.
- Check
SKILL.mdfor any special instructions or requirements.
Related skills
NVIDIA/TensorRT-LLM/ad-accuracy-debug
> Debug AutoDeploy accuracy regressions vs a reference score (PyTorch backend or published baseline).
NVIDIA/TensorRT-LLM/ad-add-fusion-transformation
> Claude Code skill (trtllm-agent-toolkit): implement or extend TensorRT-LLM AutoDeploy fusion transforms under transform/library/ in a TensorRT-LLM checkout.
NVIDIA/TensorRT-LLM/ad-conf-check
> Check whether AutoDeploy YAML configs were actually applied by analyzing server logs and optionally graph dumps (AD_DUMP_GRAPHS_DIR).
NVIDIA/TensorRT-LLM/ad-graph-dump
> Enable and interpret TensorRT-LLM AutoDeploy FX graph text dumps via AD_DUMP_GRAPHS_DIR.