Harris Notes

Home

❯

concepts

❯

Model Quantization

Model Quantization

May 02, 20261 min read

  • concept
  • general
  • technique

Model Quantization

Reducing model precision (e.g., Q5_0 5-bit integer format) to shrink memory footprint by ~65% with minimal accuracy loss, enabling larger Whisper models to run comfortably on 16GB unified memory.

Related

whisper-model-variants local-offline-transcription whisper-cpp


Graph View

  • Model Quantization
  • Related

Backlinks

  • Whisper Model Variants
  • whisper.cpp
  • 20260502-0656-local-llm-for-audio-transcription-on-mac-mini-m2-1
  • index

Created with Quartz v4.5.2 © 2026

  • GitHub
  • Discord Community