Weights & Biases logo

Weights & Biases

Experiment tracking
Artifact tracking
Data versioning
Model registry

Weights & Biases is a tool to track and visualize your machine learning pipeline pieces, from datasets to production models.

Use it when

  • You want a lightweight Python library specialized in experiment tracking, artifact tracking, and visualization.
  • You want to get started quickly with a free (for personal use) hosted platform.
  • You want built-in integrations with popular ML Frameworks (Keras, Fastai, etc.), repositories (Hugging Face, XGBoost, etc.), and tools (Kubeflow Pipelines, OpenAI Gym, etc.)
  • You want project management and collaboration tools for machine learning projects.

Watch out

  • Only the client is open-source and requires signing up.
  • The client covers data logging to Weights & Biases, data queries, and downloads of your logged data.
  • The free plan is only for personal projects.
  • Self-hosted deployment requires an Enterprise plan.

Example stacks

Airflow + MLflow stack


pip install wandb