← back to the directory
Hosted Service Observability & EvaluationFine-tuning & Training

Weights & Biases

1Track 2and 3version 4machine 5learning 6experiments

the six Ws · specification

W1 Who

Maintained by Weights & Biases.

W2 What

Tracks experiments, models, and datasets for ML.

W3 Where

A Python SDK logging to the W&B platform.

W4 When

When training and comparing machine learning runs.

W5 Why

The de facto standard for experiment tracking.

W6 With

A W&B account and the Python SDK.

W7 Watch

Company-maintained; open-source SDK (MIT) backed by a hosted SaaS. Metrics and artifacts go to the W&B cloud by default unless you self-manage the server.

Experiment-trackingMLOpsMonitoring

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.