Introduction to Neptune – your ML experiments tracking tool

Learn how to compare and analyze multiple experiments in Neptune. There are few ways to compare experiments in Neptune depending on how deep you need to go with your analysis, and how much of a detail you need to inspect.

python
import neptune

neptune.init(‘Me/MyProject’)
neptune.create_experiment(params={‘lr’:0.1, ‘dropout’:0.4})
# training and evaluation logic
neptune.log_metric(‘test_accuracy’, 0.84)