fossasia/visdom

Tool for real-time visualization, monitoring and collaborative analysis of AI/ML experiments and live data. Supports Python, PyTorch/Torch, NumPy, TensorFlow/Keras https://visdom.dev

What it solves

Visdom provides a flexible, web-based dashboard for visualizing live data, which is particularly useful for monitoring scientific experiments, tracking machine learning model training, or debugging experimental code in real-time.

How it works

It consists of a Python client and a web server. You run the server locally or on a remote machine, and then use the Python library to push plots, images, and text to the dashboard. The UI organizes these inputs into "windows" within "environments." You can interact with these windows by dragging, resizing, or closing them, and the state is persisted to disk so your dashboard remains consistent across sessions.

Who it’s for

Researchers, data scientists, and developers who need a lightweight, programmable way to visualize experimental results or model training metrics without building a custom web application.

Highlights

  • Live Dashboards: Broadcast plots, images, and text to a browser-based UI in real-time.
  • State Management: Automatically caches your visualization layout and data, allowing you to save, fork, and reload specific dashboard configurations.
  • Interactive Callbacks: Supports event handling (e.g., key presses, clicks, or property updates) to allow the dashboard to trigger actions in your Python code.
  • Environment Organization: Partition your workspace into different environments to manage multiple experiments or collaborators simultaneously.
  • Async Support: Includes an asynchronous client for high-concurrency plotting needs.

Related

  • Project
    VIAME/VIAME

    VIAME is a BSD‑licensed, open‑source computer‑vision toolkit that provides a modular C++/Python pipeline framework, a large set of ready‑made detection/tracking/measurement algorithms, and both desktop and web GUIs. It can be installed via pre‑built binaries, Docker, or pip, and supports GPU acceleration through CUDA, PyTorch, ONNX, TensorRT, etc. Ideal for researchers and developers building video analytics pipelines.

  • Project
    ContextLab/hypertools

    A Python toolbox for dimensionality reduction and visual exploration of high-dimensional data, enabling users to project complex datasets into 2D or 3D geometric insights.

  • Project
    zju3dv/Wis3D

    Wis3D is a web-based 3D visualization tool for computer vision researchers that allows for the direct import and viewing of 3D bounding boxes, point clouds, and meshes from Python.

  • Project
    kiyoon/jupynium.nvim

    Jupynium.nvim is a Neovim plugin that syncs a Jupytext‑style file (`*.ju.py`) to a live Jupyter Notebook opened in Firefox via Selenium. It provides one‑way (Neovim → browser) live preview, execution shortcuts, optional completions, folding, and a small Lua API, enabling a full notebook workflow without leaving the editor.

  • Project
    Farama-Foundation/ViZDoom

    A Doom-based AI research platform for developing and testing bots that play the game using only visual information, primarily for reinforcement learning research.