neptune-tensorflow-keras 2.0.0

Neptune.ai tensorflow-keras integration library

Neptune + TensorFlow/Keras Integration

Experiment tracking, model registry, data versioning, and live model monitoring for TensorFlow/Keras 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
• Collaborate with a team

What will be logged to Neptune?

• hyperparameters for every run,
• learning curves for losses and metrics during training,
• hardware consumption and stdout/stderr output during training,
• TensorFlow tensors as images to see model predictions live,
• training code and git commit information,
• model weights
• other metadata

Example charts in the Neptune UI with logged accuracy and loss

Resources

• Documentation
• Code example on GitHub
• Runs logged in the Neptune app
• Run example in Google Colab

Example

# On the command line:
pip install tensorflow neptune-client neptune-tensorflow-keras

# In Python:
import neptune.new as neptune
from neptune.new.integrations.tensorflow_keras import NeptuneCallback


# Start a run
run = neptune.init(project="common/tf-keras-integration",
                   api_token="ANONYMOUS")


# Create a NeptuneCallback instance
neptune_cbk = NeptuneCallback(run=run, base_namespace="metrics")


# Pass the callback to model.fit()
model.fit(x_train, y_train,
          epochs=5,
          batch_size=64,
          callbacks=[neptune_cbk])


# Stop the run
run.stop()

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 [e-mail]