# aimhubio/aim

Aim 💫 — An easy-to-use & supercharged open-source experiment tracker.

Repository: https://github.com/aimhubio/aim
Canonical: https://ross.abutalabs.com/products/aim
Homepage: https://aimstack.io
Language: Python
License: Apache-2.0
License Family: permissive
Topics: python, ai, data-science, data-visualization, experiment-tracking, machine-learning, metadata, metadata-tracking, ml, mlflow, mlops, pytorch, tensorboard, tensorflow, visualization, prompt-engineering
Last push: 2026-08-26T22:36:31+00:00

## Health v2 (maintenance only)
Score: 74/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 28, longevity 100
- inputs: {"age_days": 2651, "days_push": 7, "days_rel": 482, "gap_med": 34, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6243, forks 407 (observed 2026-08-28T04:09:40.895343+00:00)

## What it is
Aim is an open-source experiment tracker for AI/ML that logs training runs and metadata, with a web UI to compare runs and a Python API to query them programmatically. It integrates with frameworks like PyTorch, TensorFlow, and serves as a TensorBoard/MLflow alternative.

## Use cases
- track machine learning training experiments
- compare training runs across hyperparameters
- visualize training metrics like loss and accuracy
- log LLM prompts and AI metadata
- query experiment data programmatically
- replace TensorBoard or MLflow for experiment tracking
- manage MLOps experiment metadata

## When to choose
- you need a fast, self-hosted experiment tracker with a rich comparison UI
- you train models with PyTorch, TensorFlow, or Hugging Face and want minimal integration code
- you want to log and query large volumes of training metadata via an API

## When to avoid
- you need a fully managed SaaS tracking service with team collaboration out of the box
- your project is not ML/AI related and you just need general metrics monitoring
- you require deep integration with a specific MLOps platform like Weights & Biases

## Facets
- artifact type: library
- maturity: active
- function: data-visualization, logging, monitoring, analytics, sdk
- domain: machine-learning, data-science, artificial-intelligence, data-visualization, developer-tools
- platform: python, cross-platform
- tags: experiment-tracking, mlops, tensorboard-alternative, mlflow-alternative, training-runs, metadata-tracking, pytorch, tensorflow, linux, macos

## Member repositories
- aimhubio/aim (main) score 74

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:40.895343+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:47:10.486149+00:00, confidence not recorded.
  - readme: https://github.com/aimhubio/aim (fetched 2026-08-28T04:09:40.895343+00:00, sha b20f661a2575)
- Data as of 2026-08-30T08:39:29.467469+00:00.