ptq

Official

Quantize pretrained models with ModelOpt PTQ.

AuthorNVIDIA
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Post-training quantization (PTQ) enables converting pretrained models into compact, deployment-friendly checkpoints with reduced memory footprint and faster inference using ModelOpt.

Core Features & Use Cases

  • Supports nvfp4, fp8, and int4_awq formats for HF, MoE, and VLM backbones.
  • Automates calibration, validation, and export of quantized checkpoints.
  • Use Case: quantize a Qwen3-0.6B to nvfp4 for deployment on GPU-constrained environments.

Quick Start

Quantize a pretrained model to produce a deployment-ready quantized checkpoint using ModelOpt PTQ.

Dependency Matrix

Required Modules

None required

Components

references

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: ptq
Download link: https://github.com/NVIDIA/skills/archive/main.zip#ptq

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.