Getting started

You're viewing docs for Neptune 2.x

The documentation on this site is for the following Neptune versions:

• App version 2.x
• neptune-client version <2.0

For the new docs, visit docs.neptune.ai →

The following is a quick procedure for getting set up in a typical environment. Fore more installation details, see Install Neptune.

1. Create a project.

2. Install the Neptune client library:

   pip install -U neptune
   

3. Set your Neptune API token to the NEPTUNE_API_TOKEN environment variable.

   

   export NEPTUNE_API_TOKEN="uyVrZXkiOiIzNTd0Zj..ifQ=="
   

   To have API tokens for sharing or non-human accounts, you can create a service account.

4. Set the name of your Neptune project to the NEPTUNE_PROJECT environment variable.

   You can copy the full name in your project's Details & privacy menu.

   

   export NEPTUNE_PROJECT="workspace-name/project-name"
   

For more detailed instructions, see the pages in the left menu.

All set

You can now import Neptune in your code and start logging.

import neptune

run = neptune.init_run() # (1)!

run["my/namespace/structure"] = some_metadata

1. We don't recommend it, but you can also pass your info when initializing Neptune:

   import neptune
   
   run = neptune.init_run(
       project="your-workspace-name/your-project-name",
       api_token="YourNeptuneApiToken",
   )
   

Neptune also integrates directly with the most popular ML frameworks. Instead of manual logging, you can use the integrations to log typical metadata. For more, see Integrations

Next steps:

• Do the Quickstart to run a basic logging example.
• See Add Neptune to your code for beginner tips.
• View how-to guides and examples of Using Neptune