• Bio
  • Blog
  • Experience
  • Portfolio
  • OSS
  • Publication
  • Uses
  • Blog
    • 📈 Golang profiling made easy
    • 📈 Run an ArchLinux sandbox with Lima on MacOS
    • 🧠 Build a personal website with just two dollars
    • 📈 Level up your dotfiles with scripting
  • Uses
  • Contributed-Projects
    • HQQ
    • bashdb
    • podman-compose
    • golang-migrate
    • jupyter_server
    • terraform-provider-ucloud
    • titra
    • WeKan
    • RedisFullCheck
    • PowerJob
    • RocketMQ
    • instaviz
    • fastjson
    • lua-cjson
    • lua-resty-websocket
    • mybatis/generator
    • fcgiwrap
  • Publications
    • Towards Superior Quantization Accuracy: A Layer-sensitive Approach
    • A Mixed Quantization Approach for Data-Free Quantization of LLMs
  • Created-Projects
    • lm-quant-toolkit
    • ResearchBuddy
    • ghminer
    • dotfiles
    • localenv
    • doxgen-maven-plugin
    • grpc-converter-maven-plugin
    • jar-probe
    • hawks
  • Projects
  • Experience
  • Portfolio

HQQ

2025-06-16 · 1 min read
Go to Project Site

This project extends HQQ to support the mixed quantization (MXQ) method I proposed.

Last updated on 2025-06-16
LLM Quantization AI LLM Python PyTorch
Justin Zhang
Authors
Justin Zhang
Software Engineer/Researcher
bashdb 2025-06-04 →

© 2025 Justin Zhang. All rights reserved.

Published with Hugo Blox Builder — the free, open source website builder that empowers creators.