comet-ml/kangas

Health v2 · maintenance only

Adoption not part of the score

1077 stars · 50 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Kangas is a Python library and tool for exploring, analyzing, and visualizing large-scale multimedia datasets via its DataGrid format. It offers a Python API for logging large tables of data plus a visual interface for querying, grouping, sorting, and filtering millions of rows, with built-in computer vision support.

Use cases

• explore large multimedia datasets with images and bounding boxes
• visualize and filter millions of rows of data in a UI
• analyze computer vision datasets with labels and metadata
• load CSV or data files into an interactive datagrid
• run a local server to browse a dataset in a browser
• inspect ML training data from a notebook

When to choose

• you need to explore image-heavy or multimedia datasets at scale
• you want a fast visual UI for grouping, sorting, and filtering large tables
• you work in Jupyter notebooks and want integrated dataset exploration
• you need built-in visualization of bounding boxes and CV labels

When to avoid

• you need a general-purpose dataframe library for heavy numeric computation (use pandas or polars)
• you need production dashboards or BI reporting tools
• your data is purely tabular text without multimedia
• you need actively developed features - the project has seen limited recent activity

Facets

library · maturity maintenance

data-visualization analytics computer-vision data-science data-science machine-learning data-visualization computer-vision python cli cross-platform dataframe datagrid multimedia-datasets exploratory-data-analysis jupyter web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="comet-ml/kangas")