mlflow.keras

Note

Autologging is known to be compatible with the following package versions: 2.7.4 <= tensorflow <= 2.17.0. Autologging may not succeed when used with package versions outside of this range.

Enables autologging for tf.keras. Note that only tensorflow>=2.3 are supported. As an example, try running the Keras/TensorFlow example.

For each TensorFlow module, autologging captures the following information:

tf.keras

  • Metrics and Parameters

  • Training and validation loss.

  • User-specified metrics.

  • Optimizer config, e.g., learning_rate, momentum, etc.

  • Training configs, e.g., epochs, batch_size, etc.

  • Artifacts

  • Model summary on training start.

  • Saved Keras model in MLflow Model format.

  • TensorBoard logs on training end.

tf.keras.callbacks.EarlyStopping

  • Metrics and Parameters

  • Metrics from the EarlyStopping callbacks: stopped_epoch, restored_epoch, restore_best_weight, etc

  • fit() or fit_generator() parameters associated with EarlyStopping: min_delta, patience, baseline, restore_best_weights, etc

Refer to the autologging tracking documentation for more information on TensorFlow workflows.

Note that autologging cannot be used together with explicit MLflow callback, i.e., mlflow.tensorflow.MlflowCallback, because it will cause the same metrics to be logged twice. If you want to include mlflow.tensorflow.MlflowCallback in the callback list, please turn off autologging by calling mlflow.tensorflow.autolog(disable=True).

param every_n_iter

deprecated, please use log_every_epoch instead. Per every_n_iter steps, metrics will be logged.

param log_models

If True, trained models are logged as MLflow model artifacts. If False, trained models are not logged.

param log_datasets

If True, dataset information is logged to MLflow Tracking. If False, dataset information is not logged.

param disable

If True, disables the TensorFlow autologging integration. If False, enables the TensorFlow integration autologging integration.

param exclusive

If True, autologged content is not logged to user-created fluent runs. If False, autologged content is logged to the active fluent run, which may be user-created.

param disable_for_unsupported_versions

If True, disable autologging for versions of tensorflow that have not been tested against this version of the MLflow client or are incompatible.

param silent

If True, suppress all event logs and warnings from MLflow during TensorFlow autologging. If False, show all events and warnings during TensorFlow autologging.

param registered_model_name

If given, each time a model is trained, it is registered as a new model version of the registered model with this name. The registered model is created if it does not already exist.

param log_input_examples

If True, input examples from training datasets are collected and logged along with tf/keras model artifacts during training. If False, input examples are not logged.

param log_model_signatures

If True, ModelSignatures describing model inputs and outputs are collected and logged along with tf/keras model artifacts during training. If False, signatures are not logged. Note that logging TensorFlow models with signatures changes their pyfunc inference behavior when Pandas DataFrames are passed to predict(). When a signature is present, an np.ndarray (for single-output models) or a mapping from str -> np.ndarray (for multi-output models) is returned; when a signature is not present, a Pandas DataFrame is returned.

param saved_model_kwargs

a dict of kwargs to pass to tensorflow.saved_model.save method.

param keras_model_kwargs

a dict of kwargs to pass to keras_model.save method.

param extra_tags

A dictionary of extra tags to set on each managed run created by autologging.

param log_every_epoch

If True, training metrics will be logged at the end of each epoch.

param log_every_n_steps

If set, training metrics will be logged every n training steps. log_every_n_steps must be None when log_every_epoch=True.

param checkpoint

Enable automatic model checkpointing.

param checkpoint_monitor

In automatic model checkpointing, the metric name to monitor if you set model_checkpoint_save_best_only to True.

param checkpoint_save_best_only

If True, automatic model checkpointing only saves when the model is considered the “best” model according to the quantity monitored and previous checkpoint model is overwritten.

param checkpoint_mode

one of {“min”, “max”}. In automatic model checkpointing, if save_best_only=True, the decision to overwrite the current save file is made based on either the maximization or the minimization of the monitored quantity.

param checkpoint_save_weights_only

In automatic model checkpointing, if True, then only the model’s weights will be saved. Otherwise, the optimizer states, lr-scheduler states, etc are added in the checkpoint too.

param checkpoint_save_freq

“epoch” or integer. When using “epoch”, the callback saves the model after each epoch. When using integer, the callback saves the model at end of this many batches. Note that if the saving isn’t aligned to epochs, the monitored metric may potentially be less reliable (it could reflect as little as 1 batch, since the metrics get reset every epoch). Defaults to “epoch”.

Keras 3 callback to log information to MLflow.

class mlflow.keras.callback.MlflowCallback(log_every_epoch=True, log_every_n_steps=None)[source]

Bases: keras.callbacks.Callback

Note

Experimental: This class may change or be removed in a future release without warning.

Callback for logging Keras metrics/params/model/… to MLflow.

This callback logs model metadata at training begins, and logs training metrics every epoch or every n steps (defined by the user) to MLflow.

Args:
log_every_epoch: bool, defaults to True. If True, log metrics every epoch. If False,

log metrics every n steps.

log_every_n_steps: int, defaults to None. If set, log metrics every n steps. If None,

log metrics every epoch. Must be None if log_every_epoch=True.

Example 
import keras
import mlflow
import numpy as np

# Prepare data for a 2-class classification.
data = np.random.uniform([8, 28, 28, 3])
label = np.random.randint(2, size=8)
model = keras.Sequential(
    [
        keras.Input([28, 28, 3]),
        keras.layers.Flatten(),
        keras.layers.Dense(2),
    ]
)
model.compile(
    loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    optimizer=keras.optimizers.Adam(0.001),
    metrics=[keras.metrics.SparseCategoricalAccuracy()],
)
with mlflow.start_run() as run:
    model.fit(
        data,
        label,
        batch_size=4,
        epochs=2,
        callbacks=[mlflow.keras.MlflowCallback()],
    )
on_batch_end(batch, logs=None)[source]

Log metrics at the end of each batch with user specified frequency.

on_epoch_end(epoch, logs=None)[source]

Log metrics at the end of each epoch.

on_test_end(logs=None)[source]

Log validation metrics at validation end.

on_train_begin(logs=None)[source]

Log model architecture and optimizer configuration when training begins.

Functions for loading Keras models saved with MLflow.

class mlflow.keras.load.KerasModelWrapper(model, signature, save_exported_model=False)[source]

Bases: object

get_model_call_method()[source]
get_raw_model()[source]

Returns the underlying model.

predict(data, **kwargs)[source]
mlflow.keras.load.load_model(model_uri, dst_path=None, custom_objects=None, load_model_kwargs=None)[source]

Note

Experimental: This function may change or be removed in a future release without warning.

Load Keras model from MLflow.

This method loads a saved Keras model from MLflow, and returns a Keras model instance.

Parameters

  • model_uri –

    The URI of the saved Keras model in MLflow. For example:

    • /Users/me/path/to/local/model

    • relative/path/to/local/model

    • s3://my_bucket/path/to/model

    • runs:/<mlflow_run_id>/run-relative/path/to/model

    • models:/<model_name>/<model_version>

    • models:/<model_name>/<stage>

    For more information about supported URI schemes, see Referencing Artifacts.

  • dst_path – The local filesystem path to which to download the model artifact. If unspecified, a local output path will be created.

  • custom_objects – The custom_objects arg in keras.saving.load_model.

  • load_model_kwargs – Extra args for keras.saving.load_model.

Example 
import keras
import mlflow
import numpy as np

model = keras.Sequential(
    [
        keras.Input([28, 28, 3]),
        keras.layers.Flatten(),
        keras.layers.Dense(2),
    ]
)
with mlflow.start_run() as run:
    mlflow.keras.log_model(model)

model_url = f"runs:/{run.info.run_id}/{model_path}"
loaded_model = mlflow.keras.load_model(model_url)

# Test the loaded model produces the same output for the same input as the model.
test_input = np.random.uniform(size=[2, 28, 28, 3])
np.testing.assert_allclose(
    keras.ops.convert_to_numpy(model(test_input)),
    loaded_model.predict(test_input),
)
Returns

A Keras model instance.

Functions for saving Keras models to MLflow.

mlflow.keras.save.get_default_conda_env()[source]
Returns

The default Conda environment for MLflow Models produced by calls to save_model() and log_model().

mlflow.keras.save.get_default_pip_requirements()[source]
Returns

A list of default pip requirements for MLflow Models produced by Keras flavor. Calls to save_model() and log_model() produce a pip environment that, at minimum, contains these requirements.

mlflow.keras.save.log_model(model, artifact_path, save_exported_model=False, conda_env=None, signature: mlflow.models.signature.ModelSignature = None, input_example: Union[pandas.core.frame.DataFrame, numpy.ndarray, dict, list, csr_matrix, csc_matrix, str, bytes, tuple] = None, registered_model_name=None, await_registration_for=300, pip_requirements=None, extra_pip_requirements=None, save_model_kwargs=None, metadata=None)[source]

Note

Experimental: This function may change or be removed in a future release without warning.

Log a Keras model along with metadata to MLflow.

This method saves a Keras model along with metadata such as model signature and conda environments to MLflow.

Parameters

  • model – an instance of keras.Model. The Keras model to be saved.

  • artifact_path – the run-relative path to which to log model artifacts.

  • save_exported_model – defaults to False. If True, save Keras model in exported model format, otherwise save in .keras format. For more information, please refer to [Keras doc](https://keras.io/guides/serialization_and_saving/).

  • conda_env –

    Either a dictionary representation of a Conda environment or the path to a conda environment yaml file. If provided, this describes the environment this model should be run in. At a minimum, it should specify the dependencies contained in get_default_conda_env(). If None, a conda environment with pip requirements inferred by mlflow.models.infer_pip_requirements() is added to the model. If the requirement inference fails, it falls back to using get_default_pip_requirements(). pip requirements from conda_env are written to a pip requirements.txt file and the full conda environment is written to conda.yaml. The following is an example dictionary representation of a conda environment:

    {
    "name": "mlflow-env",
    "channels": ["conda-forge"],
    "dependencies": [
        "python=3.8.15",
        {
            "pip": [
                "keras==x.y.z"
            ],
        },
    ],
}

  • signature –

    an instance of the ModelSignature class that describes the model’s inputs and outputs. If not specified but an input_example is supplied, a signature will be automatically inferred based on the supplied input example and model. To disable automatic signature inference when providing an input example, set signature to False. To manually infer a model signature, call infer_signature() on datasets with valid model inputs, such as a training dataset with the target column omitted, and valid model outputs, like model predictions made on the training dataset, for example:

    from mlflow.models import infer_signature

train = df.drop_column("target_label")
predictions = ...  # compute model predictions
signature = infer_signature(train, predictions)

  • input_example – one or several instances of valid model input. The input example is used as a hint of what data to feed the model. It will be converted to a Pandas DataFrame and then serialized to json using the Pandas split-oriented format, or a numpy array where the example will be serialized to json by converting it to a list. Bytes are base64-encoded. When the signature parameter is None, the input example is used to infer a model signature.

  • registered_model_name – defaults to None. If set, create a model version under registered_model_name, also create a registered model if one with the given name does not exist.

  • await_registration_for – defaults to mlflow.tracking._model_registry.DEFAULT_AWAIT_MAX_SLEEP_SECONDS. Number of seconds to wait for the model version to finish being created and is in READY status. By default, the function waits for five minutes. Specify 0 or None to skip waiting.

  • pip_requirements – Either an iterable of pip requirement strings (e.g. ["keras", "-r requirements.txt", "-c constraints.txt"]) or the string path to a pip requirements file on the local filesystem (e.g. "requirements.txt"). If provided, this describes the environment this model should be run in. If None, a default list of requirements is inferred by mlflow.models.infer_pip_requirements() from the current software environment. If the requirement inference fails, it falls back to using get_default_pip_requirements(). Both requirements and constraints are automatically parsed and written to requirements.txt and constraints.txt files, respectively, and stored as part of the model. Requirements are also written to the pip section of the model’s conda environment (conda.yaml) file.

  • extra_pip_requirements –

    Either an iterable of pip requirement strings (e.g. ["pandas", "-r requirements.txt", "-c constraints.txt"]) or the string path to a pip requirements file on the local filesystem (e.g. "requirements.txt"). If provided, this describes additional pip requirements that are appended to a default set of pip requirements generated automatically based on the user’s current software environment. Both requirements and constraints are automatically parsed and written to requirements.txt and constraints.txt files, respectively, and stored as part of the model. Requirements are also written to the pip section of the model’s conda environment (conda.yaml) file.

    Warning

    The following arguments can’t be specified at the same time:

    • conda_env

    • pip_requirements

    • extra_pip_requirements

    This example demonstrates how to specify pip requirements using pip_requirements and extra_pip_requirements.

  • save_model_kwargs – defaults to None. A dict of kwargs to pass to keras.Model.save method.

  • metadata – Custom metadata dictionary passed to the model and stored in the MLmodel file.

Example 
import keras
import mlflow

model = keras.Sequential(
    [
        keras.Input([28, 28, 3]),
        keras.layers.Flatten(),
        keras.layers.Dense(2),
    ]
)
with mlflow.start_run() as run:
    mlflow.keras.log_model(model, "model")
mlflow.keras.save.save_model(model, path, save_exported_model=False, conda_env=None, mlflow_model=None, signature: mlflow.models.signature.ModelSignature = None, input_example: Union[pandas.core.frame.DataFrame, numpy.ndarray, dict, list, csr_matrix, csc_matrix, str, bytes, tuple] = None, pip_requirements=None, extra_pip_requirements=None, save_model_kwargs=None, metadata=None)[source]

Note

Experimental: This function may change or be removed in a future release without warning.

Save a Keras model along with metadata.

This method saves a Keras model along with metadata such as model signature and conda environments to local file system. This method is called inside mlflow.keras.log_model().

Parameters

  • model – an instance of keras.Model. The Keras model to be saved.

  • path – local path where the MLflow model is to be saved.

  • save_exported_model – If True, save Keras model in exported model format, otherwise save in .keras format. For more information, please refer to https://keras.io/guides/serialization_and_saving/.

  • conda_env –

    Either a dictionary representation of a Conda environment or the path to a conda environment yaml file. If provided, this describes the environment this model should be run in. At a minimum, it should specify the dependencies contained in get_default_conda_env(). If None, a conda environment with pip requirements inferred by mlflow.models.infer_pip_requirements() is added to the model. If the requirement inference fails, it falls back to using get_default_pip_requirements(). pip requirements from conda_env are written to a pip requirements.txt file and the full conda environment is written to conda.yaml. The following is an example dictionary representation of a conda environment:

    {
    "name": "mlflow-env",
    "channels": ["conda-forge"],
    "dependencies": [
        "python=3.8.15",
        {
            "pip": [
                "keras==x.y.z"
            ],
        },
    ],
}

  • mlflow_model – an instance of mlflow.models.Model, defaults to None. MLflow model configuration to which to add the Keras model metadata. If None, a blank instance will be created.

  • signature –

    an instance of the ModelSignature class that describes the model’s inputs and outputs. If not specified but an input_example is supplied, a signature will be automatically inferred based on the supplied input example and model. To disable automatic signature inference when providing an input example, set signature to False. To manually infer a model signature, call infer_signature() on datasets with valid model inputs, such as a training dataset with the target column omitted, and valid model outputs, like model predictions made on the training dataset, for example:

    from mlflow.models import infer_signature

train = df.drop_column("target_label")
predictions = ...  # compute model predictions
signature = infer_signature(train, predictions)

  • input_example – one or several instances of valid model input. The input example is used as a hint of what data to feed the model. It will be converted to a Pandas DataFrame and then serialized to json using the Pandas split-oriented format, or a numpy array where the example will be serialized to json by converting it to a list. Bytes are base64-encoded. When the signature parameter is None, the input example is used to infer a model signature.

  • pip_requirements – Either an iterable of pip requirement strings (e.g. ["keras", "-r requirements.txt", "-c constraints.txt"]) or the string path to a pip requirements file on the local filesystem (e.g. "requirements.txt"). If provided, this describes the environment this model should be run in. If None, a default list of requirements is inferred by mlflow.models.infer_pip_requirements() from the current software environment. If the requirement inference fails, it falls back to using get_default_pip_requirements(). Both requirements and constraints are automatically parsed and written to requirements.txt and constraints.txt files, respectively, and stored as part of the model. Requirements are also written to the pip section of the model’s conda environment (conda.yaml) file.

  • extra_pip_requirements –

    Either an iterable of pip requirement strings (e.g. ["pandas", "-r requirements.txt", "-c constraints.txt"]) or the string path to a pip requirements file on the local filesystem (e.g. "requirements.txt"). If provided, this describes additional pip requirements that are appended to a default set of pip requirements generated automatically based on the user’s current software environment. Both requirements and constraints are automatically parsed and written to requirements.txt and constraints.txt files, respectively, and stored as part of the model. Requirements are also written to the pip section of the model’s conda environment (conda.yaml) file.

    Warning

    The following arguments can’t be specified at the same time:

    • conda_env

    • pip_requirements

    • extra_pip_requirements

    This example demonstrates how to specify pip requirements using pip_requirements and extra_pip_requirements.

  • save_model_kwargs – A dict of kwargs to pass to keras.Model.save method.

  • metadata – Custom metadata dictionary passed to the model and stored in the MLmodel file.

Example 
import keras
import mlflow

model = keras.Sequential(
    [
        keras.Input([28, 28, 3]),
        keras.layers.Flatten(),
        keras.layers.Dense(2),
    ]
)
with mlflow.start_run() as run:
    mlflow.keras.save_model(model, "./model")