Python TrainingPipeline.view_schema示例

示例#1

                                     'parameters': {}
                                 }]
                             }
                         }))

# Add a trainer
training_pipeline.add_trainer(
    TFFeedForwardTrainer(loss='binary_crossentropy',
                         last_activation='sigmoid',
                         output_units=1,
                         metrics=['accuracy'],
                         epochs=20))

# Add an evaluator
training_pipeline.add_evaluator(
    TFMAEvaluator(
        slices=[['has_diabetes']],
        metrics={'has_diabetes': ['binary_crossentropy', 'binary_accuracy']}))

# Run the pipeline locally
training_pipeline.run()

# See schema of data
training_pipeline.view_schema()

# See statistics of train and eval
training_pipeline.view_statistics()

# Creates a notebook for evaluation
training_pipeline.evaluate()

示例#2

文件： smoke_test.py 项目： sjoerdteunisse/zenml

                  metrics={'has_diabetes': ['binary_crossentropy',
                                            'binary_accuracy']}))

# Run the pipeline locally
training_pipeline.run()

######################
# DO SOME EVALUATION #
######################
# Sample data
df = training_pipeline.sample_transformed_data()
print(df.shape)
print(df.describe())

# See schema of data and detect drift
print(training_pipeline.view_schema())

##########################
# CREATE SECOND PIPELINE #
##########################
training_pipeline_2 = training_pipeline.copy('Experiment 2')
training_pipeline_2.add_trainer(TFFeedForwardTrainer(
    loss='binary_crossentropy',
    last_activation='sigmoid',
    output_units=1,
    metrics=['accuracy'],
    epochs=15))
training_pipeline_2.run()

############################
# DO SOME REPOSITORY STUFF #