documentation
Examples
API reference
Integrations
Search…
Home
Changelog
Getting Started
Installation and setup
Quickstart
Examples
How to add Neptune to your code
Getting help
You should know
Core concepts
Logging metadata
Displaying metadata
What can you log and display?
Best practices
Comparing runs
Organizing and filtering Runs
Querying and downloading metadata
Connection modes
Collaboration in Neptune
How-to guides
Experiment tracking
Model registry
Model monitoring
CI/CD
Data versioning
Neptune API
Neptune UI
Integrations and Supported Tools
Introduction
Languages
Model training
Hyperparameter Optimization
Model visualization and debugging
Automation pipelines
Experiment Tracking
IDEs and notebooks
Model monitoring
API reference
API reference guide
Neptune
Project
Run
System ("sys") namespace
Model
Model version
Field types
Management
Integrations
Environment variables
Command Line Interface
Administration
Workspaces
Projects
User management
Service accounts
Logging in to Neptune with a third-party account
Deploying Neptune on your server (on-premises)
Troubleshooting
Waiting for remaining operations
Max string attribute length exceeded
Float attribute value malformed
Links
Neptune Website
Neptune App
Legacy Neptune Documentation
Migrating to neptune.new
Powered By
GitBook
Home
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.
Explainer Video
Neptune in 30 seconds
Step 1: Install the Neptune client
Depending on your operating system, open a Terminal or Command Prompt window and enter the following command:
pip
conda
pip
install
neptune-client
conda
install
-c conda-forge neptune-client
For more help, see
Installation and setup
.
Step 2: Connect Neptune to your code
import
neptune
.
new
as
neptune
run
=
neptune
.
init
(
project
=
"corp_space/fraud_detection"
)
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 app
Model building metadata in the Neptune app
Getting started
Sign up and explore yourself
Run interactive tutorial (no signup needed)
Hello World
Core concepts
If you're planning on
using Neptune with R
instead of Python, head to
our dedicated section for R
.
Integrations with the ML ecosystem
PyTorch Lightning
PyTorch
Tensorflow / Keras
Optuna
Scikit Learn
See also
the full list of integrations
.
Experiment tracking
Logging metadata
Displaying metadata
What you can log and display
Comparing Runs
Organize ML experiments
Monitor model training runs live
Model registry
Model registry overview
Need help, have questions?
Getting help
Next
Changelog
Last modified
3mo ago
Export as PDF
Copy link
Outline
Neptune in 30 seconds
Getting started
Integrations with the ML ecosystem
Experiment tracking
Model registry
Need help, have questions?