Class TrainingOptions (2.1.0)

TrainingOptions(mapping=None, *, ignore_unknown_fields=False, **kwargs)

Attributes

Name	Description
max_iterations	`int` The maximum number of iterations in training. Used only for iterative training algorithms.
loss_type	`.gcb_model.Model.LossType` Type of loss function used during training run.
learn_rate	`float` Learning rate in training. Used only for iterative training algorithms.
l1_regularization	`.wrappers.DoubleValue` L1 regularization coefficient.
l2_regularization	`.wrappers.DoubleValue` L2 regularization coefficient.
min_relative_progress	`.wrappers.DoubleValue` When early_stop is true, stops training when accuracy improvement is less than 'min_relative_progress'. Used only for iterative training algorithms.
warm_start	`.wrappers.BoolValue` Whether to train a model from the last checkpoint.
early_stop	`.wrappers.BoolValue` Whether to stop early when the loss doesn't improve significantly any more (compared to min_relative_progress). Used only for iterative training algorithms.
input_label_columns	`Sequence[str]` Name of input label columns in training data.
data_split_method	`.gcb_model.Model.DataSplitMethod` The data split type for training and evaluation, e.g. RANDOM.
data_split_eval_fraction	`float` The fraction of evaluation data over the whole input data. The rest of data will be used as training data. The format should be double. Accurate to two decimal places. Default value is 0.2.
data_split_column	`str` The column to split data with. This column won't be used as a feature. 1. When data_split_method is CUSTOM, the corresponding column should be boolean. The rows with true value tag are eval data, and the false are training data. 2. When data_split_method is SEQ, the first DATA_SPLIT_EVAL_FRACTION rows (from smallest to largest) in the corresponding column are used as training data, and the rest are eval data. It respects the order in Orderable data types: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#data-type-properties
learn_rate_strategy	`.gcb_model.Model.LearnRateStrategy` The strategy to determine learn rate for the current iteration.
initial_learn_rate	`float` Specifies the initial learning rate for the line search learn rate strategy.
label_class_weights	Sequence[`.gcb_model.Model.TrainingRun.TrainingOptions.LabelClassWeightsEntry`] Weights associated with each label class, for rebalancing the training data. Only applicable for classification models.
distance_type	`.gcb_model.Model.DistanceType` Distance type for clustering models.
num_clusters	`int` Number of clusters for clustering models.
model_uri	`str` [Beta] Google Cloud Storage URI from which the model was imported. Only applicable for imported models.
optimization_strategy	`.gcb_model.Model.OptimizationStrategy` Optimization strategy for training linear regression models.
kmeans_initialization_method	`.gcb_model.Model.KmeansEnums.KmeansInitializationMethod` The method used to initialize the centroids for kmeans algorithm.
kmeans_initialization_column	`str` The column used to provide the initial centroids for kmeans algorithm when kmeans_initialization_method is CUSTOM.

Inheritance

builtins.object > proto.message.Message > TrainingOptions

Classes

LabelClassWeightsEntry

LabelClassWeightsEntry(mapping=None, *, ignore_unknown_fields=False, **kwargs)

The abstract base class for a message.

Parameters

Name	Description
kwargs	`dict` Keys and values corresponding to the fields of the message.
mapping	Union[dict, `.Message`] A dictionary or message to be used to determine the values for this message.
ignore_unknown_fields	`Optional(bool)` If True, do not raise errors for unknown fields. Only applied if `mapping` is a mapping type or there are keyword parameters.

Methods

delattr

__delattr__(key)

Delete the value on the given field.

This is generally equivalent to setting a falsy value.

eq

__eq__(other)

Return True if the messages are equal, False otherwise.

ne

__ne__(other)

Return True if the messages are unequal, False otherwise.

setattr

__setattr__(key, value)

Set the value on the given field.

For well-known protocol buffer types which are marshalled, either the protocol buffer object or the Python equivalent is accepted.