def test_all_hyperparameters(sagemaker_session):
ntm = NTM(sagemaker_session=sagemaker_session,
encoder_layers=[1, 2, 3],
epochs=3,
encoder_layers_activation='tanh',
optimizer='sgd',
tolerance=0.05,
num_patience_epochs=2,
batch_norm=False,
rescale_gradient=0.5,
clip_gradient=0.5,
weight_decay=0.5,
learning_rate=0.5,
**ALL_REQ_ARGS)
assert ntm.hyperparameters() == dict(num_topics=str(
ALL_REQ_ARGS['num_topics']),
encoder_layers='[1, 2, 3]',
epochs='3',
encoder_layers_activation='tanh',
optimizer='sgd',
tolerance='0.05',
num_patience_epochs='2',
batch_norm='False',
rescale_gradient='0.5',
clip_gradient='0.5',
weight_decay='0.5',
learning_rate='0.5')
def test_all_hyperparameters(sagemaker_session):
ntm = NTM(sagemaker_session=sagemaker_session,
encoder_layers=[1, 2, 3],
epochs=3,
encoder_layers_activation="tanh",
optimizer="sgd",
tolerance=0.05,
num_patience_epochs=2,
batch_norm=False,
rescale_gradient=0.5,
clip_gradient=0.5,
weight_decay=0.5,
learning_rate=0.5,
**ALL_REQ_ARGS)
assert ntm.hyperparameters() == dict(
num_topics=str(ALL_REQ_ARGS["num_topics"]),
encoder_layers="[1, 2, 3]",
epochs="3",
encoder_layers_activation="tanh",
optimizer="sgd",
tolerance="0.05",
num_patience_epochs="2",
batch_norm="False",
rescale_gradient="0.5",
clip_gradient="0.5",
weight_decay="0.5",
learning_rate="0.5",
)
def test_all_hyperparameters(sagemaker_session):
ntm = NTM(sagemaker_session=sagemaker_session,
encoder_layers=[1, 2, 3], epochs=3, encoder_layers_activation='tanh', optimizer='sgd',
tolerance=0.05, num_patience_epochs=2, batch_norm=False, rescale_gradient=0.5, clip_gradient=0.5,
weight_decay=0.5, learning_rate=0.5, **ALL_REQ_ARGS)
assert ntm.hyperparameters() == dict(
num_topics=str(ALL_REQ_ARGS['num_topics']),
encoder_layers='[1, 2, 3]',
epochs='3',
encoder_layers_activation='tanh',
optimizer='sgd',
tolerance='0.05',
num_patience_epochs='2',
batch_norm='False',
rescale_gradient='0.5',
clip_gradient='0.5',
weight_decay='0.5',
learning_rate='0.5'
)
