Toolverse
All skills

hqq-quantization

by davila7

Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

Installation

Pick a client and clone the repository into its skills directory.

Installation

Quick info

Author
davila7
Category
Security

About this skill

Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

How to use

  1. Zainstaluj HQQ za pomocą pip install hqq. Jeśli chcesz konkretny backend (np. PyTorch), użyj pip install hqq[torch], pip install hqq[torchao], pip install hqq[bitblas] lub pip install hqq[marlin].

  2. Zaimportuj niezbędne moduły: from hqq.core.quantize import BaseQuantizeConfig, HQQLinear oraz torch.nn as nn.

  3. Zdefiniuj konfigurację kwantyzacji, określając precyzję (8, 4, 3, 2 lub 1 bit) i rozmiar grupy wag. Konfiguracja decyduje o kompromisie między rozmiarem modelu a jakością wyjścia.

  4. Zastosuj HQQLinear do warstw modelu, które chcesz skwantyzować. Proces przebiega bez kalibracji – wystarczy model i konfiguracja.

  5. Jeśli planujesz fine-tuning, użyj LoRA lub PEFT na skwantyzowanych wagach. HQQ jest w pełni kompatybilny z tymi metodami.

  6. Wdrażaj model z vLLM lub HuggingFace Transformers – oba frameworki natywnie wspierają HQQ i automatycznie wykorzystują zoptymalizowany backend do szybszej inferencji.

Related skills

qmd

by tobi

Search personal markdown knowledge bases, notes, meeting transcripts, and documentation using QMD - a local hybrid search engine. Combines BM25 keyword search, vector semantic search, and LLM re-ranking. Use when users ask to search notes, find documents, look up information in

Security
1951

academic-researcher

by Shubhamsaboo

Academic research assistant for literature reviews, paper analysis, and scholarly writing.\nUse when: reviewing academic papers, conducting literature reviews, writing research summaries,\nanalyzing methodologies, formatting citations, or when user mentions academic research,

Security
1260

1password

by openclaw

Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in (single or multi-account), or reading/injecting/running secrets via op.

Security
1174

software-security

by project-codeguard

A software security skill that integrates with Project CodeGuard to help AI coding agents write secure code and prevent common vulnerabilities. Use this skill when writing, reviewing, or modifying code to ensure secure-by-default practices are followed.

Security
1678

content-creator

by alirezarezvani

Create SEO-optimized marketing content with consistent brand voice. Includes brand voice analyzer, SEO optimizer, content frameworks, and social media templates. Use when writing blog posts, creating social media content, analyzing brand voice, optimizing SEO, planning content

Security
25124

ui-audit

by openclaw

AI skill for automated UI audits. Evaluate interfaces against proven UX principles for visual hierarchy, accessibility, cognitive load, navigation, and more. Based on Making UX Decisions by Tommy Geoco.

Security
1223