Quark Torch LLM Ptq

Runs an end-to-end AMD Quark post-training quantization workflow for PyTorch / Hugging Face LLMs: inspect a Hub or local model, choose a quantization plan, create reproducible artifacts, request execution approval, and produce a quantized model. Applies to Llama, Qwen, Mistral, and similar transformer LLM requests involving FP8, INT4, or another Quark scheme. Does not handle .onnx model inputs or ONNX PTQ.

500 stars 47 forksAI, LLMs & Agents Plugin marketplace

Install

Claude Code plugin
/plugin marketplace add amd/skills

The repo is a Claude Code plugin marketplace. Run /plugin afterwards to pick the plugin that includes this skill.

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add amd/skills --skill quark-torch-llm-ptq
Manual install
git clone --depth 1 https://github.com/amd/skills.git /tmp/skills
mkdir -p ~/.claude/skills
cp -r /tmp/skills/skills/quark-torch-llm-ptq ~/.claude/skills/quark-torch-llm-ptq

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Quark Torch PTQ

## Purpose

Take a PyTorch / Hugging Face LLM from model identification through confirmed
AMD Quark PTQ. Perform intake, planning, manifest generation, execution, and
output verification as one self-contained workflow.

This skill stops at the quantized model. It does not accept `.onnx` model input,
train or fine-tune a model, or modify Quark package/source files.

## Prerequisites

- Python 3.11 to 3.13, accelerator-matched PyTorch 2.2 or newer, `amd-quark[cli]`, and `datasets`.
- The required ROCm version and GPU architecture depend on the PyTorch build and selected quantization scheme. Record the actual runtime, GPU architecture (`gfx...` from `gcnArchName` on AMD), host kernel, and driver, and verify support for the confirmed plan.
- No container image is required. If running in a container, record its image name or digest when available and verify GPU device access.
- Inspect and preserve `HIP_VISIBLE_DEVICES`, `CUDA_VISIBLE_DEVICES`, `HSA_OVERRIDE_GFX_VERSION`, `PYTORCH_ROCM_ARCH`, and `PYTORCH_HIP_ALLOC_CONF`. Include any required changes to device visibility, architecture, or memory allocation in the confirmed plan.

## Inputs

- Model source: a Hugging Face repository ID or local model directory.
- Output directory.
- Quantization intent: requested precision/scheme, target hardware, accuracy
  priority, and optional calibration settings.
- Optional environment facts: Python, PyTorch, accelerator, available memory,
Read the full skill on GitHub

Details

Skill name
quark-torch-llm-ptq
Repository
amd/skills
Path
skills/quark-torch-llm-ptq/SKILL.md
License
MIT
Repo updated
18 November 2026

FAQ

What does the Quark Torch LLM Ptq skill do?

Runs an end-to-end AMD Quark post-training quantization workflow for PyTorch / Hugging Face LLMs: inspect a Hub or local model, choose a quantization plan, create reproducible artifacts, request execution approval, and produce a quantized model. Applies to Llama, Qwen, Mistral, and similar transformer LLM requests involving FP8, INT4, or another Quark scheme. Does not handle .onnx model inputs or ONNX PTQ.

How do I install the Quark Torch LLM Ptq skill?

Claude Code plugin: /plugin marketplace add amd/skills. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Quark Torch LLM Ptq popular?

It lives in amd/skills, which has 500 GitHub stars and 47 forks.

Related skills

Apu Memory Tuner

500

amd/skills

Inspects and tunes the shared-vs-dedicated memory split on AMD Ryzen APUs with unified memory (UMA) so larger LLMs and image-gen models fit on the iGPU, or so reserved GPU memory is returned to the CPU. Use when the user mentions Ryzen AI, Strix Halo / Strix Point / Krackan / Phoenix / Hawk Point, Ryzen AI Max, gfx1150 / gfx1151 / gfx1152, integrated Radeon, iGPU memory, UMA Frame Buffer Size, AMD Variable Graphics Memory, VGM, GTT, GART, TTM, pages_limit, amd-ttm, amd-debug-tools, "shared GPU memory", "dedicated GPU memory", carve-out, "not enough VRAM", "out of VRAM", "GPU OOM", llama.cpp on

Productivity & Notes47

Eaiapps Guided Install

500

amd/skills

Guides a user through installing the AI Workbench (AIWB) onto a Kubernetes cluster end-to-end — preflight cluster readiness (access, StorageClass, GPU operator, required operators/CRDs), install the Helm charts in the correct order, and verify the deployment is healthy with plain-language failure triage. Use when a user wants to install, deploy, or stand up AIWB / the AI Workbench / the enterprise AI stack on a K8s cluster, or asks why their install failed partway. Use PROACTIVELY when a user is following helm/INSTALL.md by hand. Do NOT activate for onboarding models to AIM, general kubectl/he

DevOps & Cloud47

Flydsl Kernel Authoring

500

amd/skills

Comprehensive reference for authoring FlyDSL GPU kernels on AMD GPUs. Covers the layout algebra, tiled copy/MMA, buffer ops, loop-carried range loops, SharedAllocator (LDS), autotuning, and common patterns. Use when writing, reviewing, or understanding FlyDSL kernel code.

Writing & Content47

Rocprofiler Systems Choosing a Profiling Preset

500

amd/skills

Recommends which rocprofiler-systems runtime profiling preset (balanced, trace-gpu, trace-hpc, sys-trace, workload-trace, etc.) to pass to `rocprof-sys-run --preset=<name>` or `rocprof-sys-sample --preset=<name>` based on the user's workload and profiling goal, and returns the ready-to-run command. Use when the user asks "which preset should I use", describes a workload (GPU kernel, MPI, OpenMP, AI/ML training, production overview) and wants profiling guidance, or invokes `rocprof-sys-run`/`rocprof-sys-sample` without picking a preset. Do NOT use for CMake build presets (`cmake --preset debug|

AI, LLMs & Agents47

Magpie Kernel Evaluator

500

amd/skills

Benchmarks LLM inference and drives GPU kernel optimization with Magpie. Use when the user wants to benchmark vLLM, SGLang, or Atom; capture torch traces; post-process inference traces with TraceLens into prefill/decode and roofline reports; identify top bottleneck kernels or map profiler names to source; analyze or compare HIP, CUDA, PyTorch, or Triton kernels; validate and rank optimized variants; run local, container, or Ray workloads; or mentions Magpie, TraceLens, gap analysis, TTFT, TPOT, kernel evaluation, or AMD GPU optimization.

AI, LLMs & Agents47

Serving Llms on Epyc

500

amd/skills

Serves an LLM on a supported AMD EPYC server CPU using vLLM with zentorch, in Docker, Podman, or conda. Use for "vLLM on CPU", "zentorch serving", or an EPYC CPU endpoint, including on a host that also has AMD Instinct GPUs. Detects the EPYC generation, validates the runtime, checks model support and RAM fit, sizes threads/KV/NUMA, confirms the plan, launches, and verifies the endpoint. Runs one instance on one socket and its memory. Reports and stops on failure; does not retry or debug. Use serving-llms-on-instinct when the endpoint should run on a GPU. Excludes multi-node, EPYC 4000, and pre

AI, LLMs & Agents47

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.