Working with the Neptune UI

Neptune is a browser-enabled app that lets you visualize and browse experiments.

Experiments view

Supported Browsers

Neptune supports Chrome and Firefox, on all major operating systems.

Log in to the App

  1. In a web browser, navigate to ui.neptune.ai or, in the case of an on-prem deployment, to your Neptune instance address.

  2. In the window that appears, click Login in the upper right corner.

  3. In the Sign-in dialog that appears, either:

    • Click the Facebook, Google, or GitHub icon and sign in.

    • Type in your user name and password, then click Sign in.

    Login screen

Experiments View

The Experiments tab displays all the experiments in a specific Project in table form.

Filter experiments

You can perform the simplest filtering by typing into the search fields:

Experiment search

You can also use the Neptune Query Language to filter experiments for more advanced criteria.

Organize experiments

There are several ways to organize your experiments:

  • Use tags: You can create tag(s), which you assign to experiments. Later, you can quickly filter by these tags.

    Tag chooser
  • Customize views: You can create a custom view of the dashboard. For example, you can filter rows by parameter or metric values and select a subset of useful columns. Then save the view and quickly return to it by selecting it from the list of views.

    View list
  • Choose columns: You can configure several data types logged to Neptune so that they are displayed as columns in the dashboard. They are metrics, text logs, properties and parameters. However, all data can be seen in the single experiment view.

  • Group experiments: You can group experiments by one or more column(s). The dashboard displays the selected columns, allowing you to make in-group and across-groups analysis of the experiments. Each group is represented by the first experiment that appears according to the sorting order. After opening it, each group shows at most 10 experiments - all experiments can be viewed by clicking Show all.

    Group columns

Compare experiments

Neptune lets you compare up to 10 experiments using in-depth analysis in the specialized view. Simply select experiments and click Compare:

Compare experiments table
  • Overlaid charts: In the comparison view, all metrics with the same name are placed on a single chart with one curve per experiment. The customizable legend lets you select additional metrics and/or parameters to display. When hovering with the mouse over a particular area, the values for the selected metrics are displayed below:

    Charts legend
  • Interactive comparison table: Below the charts, details of the experiments being compared are shown in table form. Each column represents one experiment and each row represents a single property and the data associated with it.

    Compare experiments table

Single Experiment View

Click a line in the experiments table to see details of that experiment.

Single experiment

Inside the experiment, there are tabs in the left sidebar. Each displays specific content that is logged to Neptune for the specific experiment. Each tab has a unique URL.

  • Charts: All metrics (numeric type of data) are visualized as charts.

  • Logs: Logs are data that can be logged or tracked to the experiment. There are multiple types: Numeric, text, and images.

  • Monitoring: Displays information about utilization of the memory, CPU and GPU, as well as stdout and stderr.

  • Artifacts: Displays files uploaded to the experiment.

  • Source code: Displays sources uploaded to the experiment.

  • Parameters: Displays parameters uploaded to the experiment during experiment creation.

  • Details: Displays additional metadata information.

For more information about the experiment data that is logged, see Logging to Neptune.

Notebooks

Neptune lets you track Jupyter Notebooks.

For more information, see Using Jupyter Notebooks in Neptune.

Wiki

An interactive Wiki with collaborative editing features lets multiple users edit the Wiki pages and create reports with teammates in real time.

For more information, see Working with the Project Wiki.