fsrs4anki: a machine-learning based spaced-repetition scheduler that optimizes Anki review intervals

fsrs4anki: a machine-learning based spaced-repetition scheduler that optimizes Anki review intervals

What it solves

FSRS4Anki replaces Anki's default scheduling algorithm with a more modern, data-driven approach. It aims to optimize the timing of card reviews to better align with how humans actually remember things, reducing unnecessary reviews while maintaining retention.

How it works

The project consists of two primary components:

  1. The Scheduler: This replaces the built-in Anki scheduler to determine when a card should be next reviewed based on the FSRS algorithm.
  2. The Optimizer: This uses machine learning to analyze a user's specific review history and calculate the parameters that best fit their individual memory patterns.

Who it’s for

Anki users who want a more efficient and more personalized spaced-repetition system based on recent memory research.

Highlights

  • Machine Learning Optimization: Learns from your personal review history to customize scheduling parameters.
  • Data-Driven Approach: Based on the Free Spaced Repetition Scheduler (FSRS) algorithm, which balances interpretability and verifiability.
  • Anki Integration: Integrated into Anki 23.10 or newer, with support for older versions via a standalone scheduler.

Sources