Neptune comes with 25+ integrations with Python libraries popular in machine learning, deep learning and reinforcement learning.

How integrations work?

Integrations are written using neptune-client and provide a convenient way to jump-start working with Neptune and a library that you are using. There is no need to integrate your code manually using neptune-client (it’s easy though).

For each integration, the installation, scope and usage example are explained in detail in the documentation (i.e.: TensorFlow-Keras integration).

List of all integrations

• ​Altair​

• ​Bokeh​

• ​Catalyst​

• ​Dalex​

• ​Fastai​

• ​HiPlot​

• ​HTML​

• ​TensorFlow / Keras​

• ​Keras Tuner​

• ​Lightgbm​

• ​Matplotlib​

• ​MLflow​

• ​Optuna​

• ​Pandas​

• ​Plotly​

• ​Pytorch​

• ​Pytorch Ignite​

• ​Pytorch Lightning​

• ​Sacred​

• ​Scikit-learn​

• ​Scikit Optimize​

• ​Skorch​

• ​Tensorboard​

• ​XGBoost​

• Ray/Tune coming soon…

• Hydra coming soon…

• ACME coming soon…

• Spinning up coming soon…

• Stable baselines coming soon…

My library is not here. What now?

There are two common paths:

1. You can always use neptune-client , our open source Python library for logging all kinds of data and metadata to runs.

2. Contact us directly via mail ([email protected]) or chat (that little thing in the lower right corner) to discuss what you need and how we can deliver it.