Skip to content

Instantly share code, notes, and snippets.

View in0vik's full-sized avatar
⛩️
Focusing

Ilya Novik in0vik

⛩️
Focusing
6 followers · 13 following
View GitHub Profile
@in0vik
in0vik / README.md
Created March 3, 2026 10:11
Whisper quantization benchmark: WER + speed for FP16 vs quantized models (whisper.cpp, M1, jiwer)

Whisper Quantization Benchmark

Compares WER (Word Error Rate) and inference speed across FP16 and quantized Whisper models using whisper.cpp — the same engine Handy ships.

Models tested

Model Size Source
small (FP16) 487 MB ggerganov/whisper.cpp
small-q8_0 264 MB ggerganov/whisper.cpp
@in0vik
in0vik / PRIVACY.md
Created December 22, 2025 18:20
Privacy Policy for Shortcut Reactions for Google Meet Chrome Extension

Privacy Policy for Shortcut Reactions for Google Meet

Last updated: December 22, 2025

Data Collection

This extension does NOT collect, store, or transmit any user data.

How It Works