ds = Repository.get_instance().get_datasource_by_name(
'Pima Indians Diabetes')
training_pipeline.add_datasource(ds)
# Add a split
training_pipeline.add_split(RandomSplit(split_map={'train': 0.7, 'eval': 0.3}))
# Add a preprocessing unit
training_pipeline.add_preprocesser(
StandardPreprocesser(features=[
'times_pregnant', 'pgc', 'dbp', 'tst', 'insulin', 'bmi', 'pedigree',
'age'
],
labels=['has_diabetes'],
overwrite={
'has_diabetes': {
'transform': [{
'method': 'no_transform',
'parameters': {}
}]
}
}))
# Add a trainer
training_pipeline.add_trainer(
TFFeedForwardTrainer(loss='binary_crossentropy',
last_activation='sigmoid',
output_units=1,
metrics=['accuracy'],
epochs=20))
except AlreadyExistsException:
ds = repo.get_datasource_by_name("my_csv_datasource")
training_pipeline.add_datasource(ds)
# Add a split
training_pipeline.add_split(CategoricalDomainSplit(
categorical_column="name",
split_map={'train': ["arnold", "nicholas"], 'eval': ["lülük"]}))
# Add a preprocessing unit
training_pipeline.add_preprocesser(
StandardPreprocesser(
features=["name", "age"],
labels=['gpa'],
overwrite={'gpa': {
'transform': [
{'method': 'no_transform', 'parameters': {}}]}}
))
# Add a trainer
training_pipeline.add_trainer(TFFeedForwardTrainer(
batch_size=1,
loss='binary_crossentropy',
last_activation='sigmoid',
output_units=1,
metrics=['accuracy'],
epochs=i))
# Run the pipeline locally
training_pipeline.run()