Keras integration API reference

API reference for the Neptune-Keras integration.

You can use a Neptune callback to capture model training metadata when using TensorFlow with Keras.

NeptuneCallback

Captures model training metadata and logs them to Neptune.

Note

To use this package, you need to have Keras or TensorFlow 2 installed on your machine.

Parameters

Name	Type	Default	Description
run	RunorHandler	-	(required) An existing run reference, as returned byneptune.init_run(), or anamespace handler.
base_namespace	str, optional	training	Namespace under which all metadata logged by the Neptune callback will be stored.
log_model_diagram	bool, optional	False	Save the model visualization. Requirespydotto be installed.
log_on_batch	bool, optional	False	Log the metrics also for each batch, not only each epoch.
log_model_summary	bool, optional	True	Whether to log the model summary.

Examples

Creating a Neptune run and callback

Create a run:

import neptune

run = neptune.init_run()

As a best practice, you should save your Neptune API token and project name as environment variables:

export NEPTUNE_API_TOKEN="h0dHBzOi8aHR0cHM6Lkc78ghs74kl0jv...Yh3Kb8"

export NEPTUNE_PROJECT="ml-team/classification"

Alternatively, you can pass the information when using a function that takesapi_tokenandprojectas arguments:

run = neptune.init_run(
 api_token="h0dHBzOi8aHR0cHM6Lkc78ghs74kl0jv...Yh3Kb8", # (1)!
 project="ml-team/classification", # (2)!
)

In the bottom-left corner, expand the user menu and select Get my API token .
You can copy the path from the project details ( → Details & privacy ).

If you haven't registered, you can log anonymously to a public project:

api_token=neptune.ANONYMOUS_API_TOKEN
project="common/quickstarts"

Make sure not to publish sensitive data through your code!

Instantiate the Neptune callback:

from neptune.integrations.tensorflow_keras import NeptuneCallback

neptune_callback = NeptuneCallback(run=run)

Pass the callback to thecallbacksargument of model.fit()

model.fit(x_train, y_train, callbacks=[neptune_callback])

Logging with additional options

import neptune
from neptune.integrations.tensorflow_keras import NeptuneCallback

run = neptune.init_run()

neptune_callback = NeptuneCallback(
 run=run,
 base_namespace="visualizations", # optionally set a custom namespace name
 log_model_diagram=True,
 log_on_batch=True,
)

...