optuna/optuna-dashboard

Real-time Web Dashboard for Optuna.

What it solves

It provides a real-time visual interface for monitoring and analyzing hyperparameter optimization studies conducted with Optuna, removing the need to manually parse logs or database files to understand model performance.

How it works

The dashboard connects to an Optuna storage backend (such as SQLite, MySQL, or PostgreSQL) via a storage URL. It reads the study data and renders it in a web-based interface. It can be deployed as a standalone command-line tool, a Docker container, a Jupyter Lab extension, or a VS Code extension. There is also a browser-only version that uses SQLite3 Wasm and Rust to analyze files locally without a Python installation.

Who it’s for

Machine learning engineers and researchers who use Optuna for hyperparameter tuning and need a way to visualize optimization history and hyperparameter importance.

Highlights

  • Real-time visualization of optimization studies.
  • Support for multiple database backends including SQLite, MySQL, and PostgreSQL.
  • Integration with Jupyter Lab and VS Code.
  • Browser-only version for quick analysis of SQLite3 files without installing dependencies.

Related

  • Project
    optuna/optuna

    An automatic hyperparameter optimization framework for machine learning that uses a define-by-run API to dynamically tune model parameters.

  • Project
    fossasia/visdom

    A real-time visualization tool for scientific experimentation that allows users to broadcast live plots, images, and text to a web-based dashboard.

  • Project
    syne-tune/syne-tune

    A library for large-scale and reproducible hyperparameter optimization (HPO) that supports distributed execution and state-of-the-art tuning methods.

  • Project
    mljar/supertree

    An interactive decision tree visualization tool for Jupyter notebooks that supports scikit-learn, XGBoost, LightGBM, and ONNX models.