Skip to main content
App version: 3.4.6

Create an experiment

To create an experiment, use the Run() constructor.

For example, to create a run that becomes the head of the seabird-flying-skills experiment, use:

from neptune_scale import Run

run = Run(
    run_id="astute-kittiwake-23",
    experiment_name="seabird-flying-skills",
)

The run_id and experiment_name arguments are mandatory.

  • The custom run ID provided to the run_id argument must be unique within the project. It can't contain the / character.

  • To make the name easy to read in the app, ensure that it's at most 190 characters long.

Then, to track experiment metadata, call a logging method on the Run object:

for step in epoch:
    # your training loop
    acc = ...
    loss = ...

    run.log_metrics(
        data={"metrics/accuracy": acc, "metrics/loss": loss},
        step=1,
    )

To open the head run of a particular experiment in the app, construct a URL as follows:

...neptune.ai/{Workspace}/{Project}/runs/...&runIdentificationKey={ExperimentName}&type=experiment

For details, see Construct Neptune URLs.

Creating non-experiment runs​

Forking and history are only supported for experiment runs. However, it's possible to create stand-alone runs that aren't part of any experiment. In this case, only a unique run ID is required when creating the run:

Create a non-experiment run
run = Run(run_id="astute-kittiwake-178")
Related