Neptune is a metadata store for MLOps, built for teams that run a lot of experiments.‌

It gives you a single place to log, store, display, organize, compare, and query all your model-building metadata.

‌Neptune is used for:‌

  • Experiment tracking: Log, display, organize, and compare ML experiments in a single place.

  • Model registry: Version, store, manage, and query trained models, and model building metadata.

  • Monitoring ML runs live: Record and monitor model training, evaluation, or production runs live

Neptune in 30 seconds

Step 1: Install neptune-client

Depending on your operating system open a terminal or CMD and run this command:

pip install neptune-client
conda install -c conda-forge neptune-client

For more help see installing neptune-client.

Step 2: Connect Neptune to your code

import neptune.new as neptune
run = neptune.init(project="corp_space/fraud_detection")

This is tested with neptune-client==0.9.16.

Step 3: Log metadata

run["parameters"] = {"lr": 0.001, "optim": "Adam"} #parameters
run["f1_score"] = 0.66 #metrics
run["roc_curve"].upload("roc_curve.png") #charts
run["model"].upload("model.h5") #models

Step 4: See it live in the UI

Model building metadata logged to the Neptune UI

Getting started

Integrations with the ML ecosystem

Experiment Tracking

Model Registry

Need help, questions?