classifier

Classifier Context.

Classes

ClassifierContext

ashpy.ClassifierContext provide the standard functions to test a classifier.

class ashpy.contexts.classifier.ClassifierContext(classifier_model=None, loss=None, dataset=None, metrics=None, log_eval_mode=<LogEvalMode.TEST: 0>, global_step=<tf.Variable 'global_step:0' shape=() dtype=int64, numpy=0>, ckpt=None)

Bases: ashpy.contexts.base_context.BaseContext

ashpy.ClassifierContext provide the standard functions to test a classifier.

__init__(classifier_model=None, loss=None, dataset=None, metrics=None, log_eval_mode=<LogEvalMode.TEST: 0>, global_step=<tf.Variable 'global_step:0' shape=() dtype=int64, numpy=0>, ckpt=None)

Instantiate the ashpy.contexts.classifier.ClassifierContext context.

Parameters

• classifier_model (tf.keras.Model) – A tf.keras.Model model.

• loss (ashpy.losses.Executor) – Loss function, format f(y_true, y_pred).

• dataset (tf.data.Dataset) – The test dataset.

• metrics (list of [ashpy.metrics.metric.Metric]) – List of ashpy.metrics.metric.Metric with which to measure training and validation data performances.

• log_eval_mode (ashpy.modes.LogEvalMode) – Models’ mode to use when evaluating and logging.

• global_step (tf.Variable) – tf.Variable that keeps track of the training steps.

• ckpt (tf.train.Checkpoint) – checkpoint to use to keep track of models status.

Return type

None

property classifier_model

Retrieve the Model Object.

Return type

Model

Returns

tf.keras.Model.

property loss

Retrieve the loss value.

Return type

Optional[Executor]