def test_skip_concat_over_pooling_stub():
model = to_stub_model(get_pooling_model())
graph = Graph(model)
layer_num = graph.n_layers
graph.to_concat_skip_model(graph.layer_to_id[model.layers[4]], graph.layer_to_id[model.layers[11]])
assert graph.n_layers == layer_num + 2
def test_produce_keras_model():
for graph in [get_conv_dense_model(),
get_add_skip_model(),
get_pooling_model(),
get_concat_skip_model()]:
model = graph.produce_keras_model()
assert isinstance(model, keras.models.Model)
def test_skip_concat_over_pooling_stub():
model = get_pooling_model()
graph = Graph(model, False)
layer_num = graph.n_layers
graph.to_concat_skip_model(2, 15)
assert graph.n_layers == layer_num + 2
def test_skip_concat_over_pooling_stub():
graph = get_pooling_model()
graph.weighted = False
layer_num = graph.n_layers
graph.to_concat_skip_model(1, 14)
assert graph.n_layers == layer_num + 6
def test_skip_concat_over_pooling_stub():
graph = get_pooling_model()
graph.weighted = False
layer_num = graph.n_layers
graph.to_concat_skip_model(1, 11)
assert graph.n_layers == layer_num + 5
def test_skip_add_over_pooling_stub():
model = get_pooling_model()
graph = Graph(model, False)
layer_num = graph.n_layers
graph.to_add_skip_model(1, 10)
assert graph.n_layers == layer_num + 3
def test_keras_model():
for graph in [get_conv_dense_model(),
get_add_skip_model(),
get_pooling_model(),
get_concat_skip_model()]:
keras_model = KerasModel(graph)
keras_model.set_weight_to_graph()
assert isinstance(keras_model, KerasModel)
def test_produce_keras_model():
for graph in [
get_conv_dense_model(),
get_add_skip_model(),
get_pooling_model(),
get_concat_skip_model()
]:
model = graph.produce_keras_model()
assert isinstance(model, keras.models.Model)
def test_skip_add_over_pooling():
model = get_pooling_model()
graph = Graph(model)
new_model = graph.to_add_skip_model(model.layers[4], model.layers[11])
input_data = get_conv_data()
output1 = model.predict_on_batch(input_data).flatten()
output2 = new_model.predict_on_batch(input_data).flatten()
assert np.array_equal(output1, output2)
def test_keras_model():
for graph in [
get_conv_dense_model(),
get_add_skip_model(),
get_pooling_model(),
get_concat_skip_model()
]:
keras_model = KerasModel(graph)
keras_model.set_weight_to_graph()
assert isinstance(keras_model, KerasModel)
def test_skip_concat_over_pooling():
model = get_pooling_model()
graph = Graph(model)
new_model = graph.to_concat_skip_model(model.layers[4], model.layers[11])
input_data = get_conv_data()
output1 = model.predict_on_batch(input_data).flatten()
output2 = new_model.predict_on_batch(input_data).flatten()
assert abs(np.sum(output1 - output2)) < 1e-4
def test_skip_add_over_pooling():
model = get_pooling_model()
graph = Graph(model, True)
graph.to_add_skip_model(2, 10)
new_model = graph.produce_model()
input_data = get_conv_data()
output1 = model.predict_on_batch(input_data).flatten()
output2 = new_model.predict_on_batch(input_data).flatten()
assert np.array_equal(output1, output2)
def test_skip_add_over_pooling():
model = get_pooling_model()
graph = NetworkMorphismGraph(model)
graph.to_add_skip_model(graph.layer_to_id[model.layers[4]], graph.layer_to_id[model.layers[11]])
new_model = graph.produce_model()
input_data = get_conv_data()
output1 = model.predict_on_batch(input_data).flatten()
output2 = new_model.predict_on_batch(input_data).flatten()
assert np.array_equal(output1, output2)
def test_skip_concat_over_pooling():
model = get_pooling_model()
graph = Graph(model, True)
graph.to_concat_skip_model(6, 11)
graph.to_concat_skip_model(6, 11)
graph = Graph(graph.produce_model(), True)
new_model = graph.produce_model()
input_data = get_conv_data()
output1 = model.predict_on_batch(input_data).flatten()
output2 = new_model.predict_on_batch(input_data).flatten()
assert np.sum(np.abs(output1 - output2)) < 4e-1
def test_skip_concat_over_pooling():
model = get_pooling_model()
graph = NetworkMorphismGraph(model)
graph.to_concat_skip_model(graph.layer_to_id[model.layers[4]], graph.layer_to_id[model.layers[8]])
graph.to_concat_skip_model(graph.layer_to_id[model.layers[4]], graph.layer_to_id[model.layers[8]])
graph = NetworkMorphismGraph(graph.produce_model())
new_model = graph.produce_model()
input_data = get_conv_data()
output1 = model.predict_on_batch(input_data).flatten()
output2 = new_model.predict_on_batch(input_data).flatten()
assert np.sum(np.abs(output1 - output2)) < 4e-1
def test_skip_add_over_pooling():
graph = get_pooling_model()
model = graph.produce_model()
graph = deepcopy(graph)
graph.to_add_skip_model(1, 10)
new_model = graph.produce_model()
input_data = torch.Tensor(get_conv_data())
model.eval()
new_model.eval()
output1 = model(input_data)
output2 = new_model(input_data)
assert (output1 - output2).abs().sum() < 1e-4
def test_skip_add_over_pooling():
graph = get_pooling_model()
model = graph.produce_model()
graph = deepcopy(graph)
graph.to_add_skip_model(1, 8)
new_model = graph.produce_model()
input_data = torch.Tensor(get_conv_data())
model.eval()
new_model.eval()
output1 = model(input_data)
output2 = new_model(input_data)
assert (output1 - output2).abs().sum() < 1e-4
def test_legal_graph2():
graph = get_pooling_model()
graph.to_concat_skip_model(1, 5)
assert legal_graph(graph)
graph.to_concat_skip_model(1, 5)
assert not legal_graph(graph)
def test_legal_graph():
graph = get_pooling_model()
graph.to_add_skip_model(1, 4)
assert legal_graph(graph)
graph.to_add_skip_model(1, 4)
assert not legal_graph(graph)
def test_transform():
models = transform(Graph(to_stub_model(get_pooling_model())))
assert len(models) == constant.N_NEIGHBORS
def test_legal_graph2():
graph = get_pooling_model()
graph.to_concat_skip_model(1, 4)
assert legal_graph(graph)
graph.to_concat_skip_model(1, 4)
assert not legal_graph(graph)
def test_legal_graph():
graph = Graph(get_pooling_model(), False)
graph.to_add_skip_model(1, 5)
assert legal_graph(graph)
graph.to_add_skip_model(1, 5)
assert not legal_graph(graph)
def test_transform():
models = transform(get_pooling_model())
assert len(models) == Constant.N_NEIGHBOURS
def test_wider_dense():
model = to_wider_graph(Graph(to_stub_model(get_pooling_model())))
assert isinstance(model, Graph)
def test_transform():
models = transform(Graph(get_pooling_model(), False))
assert len(models) == 12
def test_legal_graph2():
graph = Graph(get_pooling_model(), False)
graph.to_concat_skip_model(2, 6)
assert legal_graph(graph)
graph.to_concat_skip_model(2, 6)
assert not legal_graph(graph)
def test_skip():
model = to_skip_connection_graph(get_pooling_model())
assert isinstance(model, Graph)
def test_wider_dense():
model = to_wider_graph(Graph(get_pooling_model(), False))
assert isinstance(model, Graph)
def test_wider():
model = to_wider_graph(get_pooling_model())
assert isinstance(model, Graph)
def test_transform():
models = transform(Graph(get_pooling_model(), False))
assert len(models) == constant.N_NEIGHBOURS
