Deepnote

You can easily execute runs in Deepnote notebooks and log them to Neptune. To do that you need to:
Step 1: Install neptune-client
! pip install neptune-client
Step 2: Set up an environment variable for the API token
Create a new environment variable integration in the left tab where you set the NEPTUNE_API_TOKEN. Alternatively, you can initialize neptune with the API token directly with the snippet:
# Alternative version to initialising Neptune
run = neptune.init(project='', api_token='')
See how to get your Neptune API token here.
Step 3: Replace the project name and log metrics into a Neptune dashboard
import neptune.new as neptune
​
# The init() function called this way assumes that
# NEPTUNE_API_TOKEN environment variable is defined by the integration.
run = neptune.init(project='my_workspace/my_project',
                   name='minimal_example')
​
# Track some metadata and hyperparameters
run["JIRA"] = "NPT-952"
run["algorithm"] = "ConvNet"
​
params = {
    "batch_size": 64,
    "dropout": 0.2,
    "learning_rate": 0.001,
    "optimizer": "Adam"
}
run["parameters"] = params
​
# Track the training process by logging your training metrics
for epoch in range(100):
    run["train/accuracy"].log(epoch * 0.6)
    run["train/loss"].log(epoch * 0.4)
​
# Log the final results
run["f1_score"] = 0.66
Amazon SageMaker
Google Colab
Last updated 3 weeks ago