Neptune + LightGBM Integration

Experiment tracking, model registry, data versioning, and live model monitoring for LightGBM trained models.

What will you get with this integration?

• Log, display, organize, and compare ML experiments in a single place
• Version, store, manage, and query trained models, and model building metadata
• Record and monitor model training, evaluation, or production runs live

What will be logged to Neptune?

• training and validation metrics,
• parameters,
• feature names, num_features, and num_rows for the train set,
• hardware consumption (CPU, GPU, memory),
• stdout and stderr logs,
• training code and Git commit information,
• other metadata

Example dashboard with train-valid metrics and selected parameters

Resources

• Documentation
• Code example on GitHub
• Example of a run logged in the Neptune app
• Run example in Google Colab

Example

# On the command line:
pip install neptune-lightgbm

# In Python:
import lightgbm as lgb
import neptune
from neptune.integrations.lightgbm import NeptuneCallback

# Start a run
run = neptune.init_run(
    project="common/lightgbm-integration",
    api_token=neptune.ANONYMOUS_API_TOKEN,
)

# Create a NeptuneCallback instance
neptune_callback = NeptuneCallback(run=run)

# Prepare datasets
...
lgb_train = lgb.Dataset(X_train, y_train)

# Define model parameters
params = {
    "boosting_type": "gbdt",
    "objective": "multiclass",
    "num_class": 10,
    ...
}

# Train the model
gbm = lgb.train(
    params,
    lgb_train,
    callbacks=[neptune_callback],
)

Support

If you got stuck or simply want to talk to us, here are your options:

• Check our FAQ page
• You can submit bug reports, feature requests, or contributions directly to the repository.
• Chat! When in the Neptune application click on the blue message icon in the bottom-right corner and send a message. A real person will talk to you ASAP (typically very ASAP),
• You can just shoot us an email at [email protected]