import plot_service, data_service
import optimization, opt_utils
plot_service.visualize_dataset()
train_X, train_Y = data_service.load_dataset()
# train 3-layer model
layers_dims = [train_X.shape[0], 5, 2, 1]
learning_rate = 0.0007
optimizers = ['gd', 'momentum', 'adam']
for optimizer in optimizers:
parameters, costs = optimization.model(train_X,
train_Y,
layers_dims,
optimizer=optimizer)
plot_service.plot_loss_per_iteration_for_learning_rate(
costs, learning_rate)
predictions = opt_utils.predict(train_X, train_Y, parameters)
plot_service.plot_decision_boundary(
lambda x: opt_utils.predict_dec(parameters, x.T), train_X, train_Y,
"Model with {0} optimization".format(optimizer))
from res_net import ResNet50
import keras.backend as K
import data_service
from keras.utils import plot_model
from IPython.display import SVG
from keras.utils.vis_utils import model_to_dot
K.set_image_data_format('channels_last')
K.set_learning_phase(1)
model = ResNet50(input_shape=(64, 64, 3), classes=6)
model.compile(optimizer='adam',
loss='categorical_crossentropy',
metrics=['accuracy'])
X_train_orig, Y_train_orig, X_test_orig, Y_test_orig, classes = data_service.load_dataset(
)
X_train, Y_train, X_test, Y_test = data_service.preprocess_data(
X_train_orig, Y_train_orig, X_test_orig, Y_test_orig)
model.fit(X_train, Y_train, epochs=2, batch_size=32)
preds = model.evaluate(X_test, Y_test)
print("Loss = " + str(preds[0]))
print("Test Accuracy = " + str(preds[1]))
model.summary()
plot_model(model, to_file='plots/model.png')
SVG(model_to_dot(model).create(prog='dot', format='svg'))
import keras.backend as K
import data_service, plot_service
from happy_model import HappyModel
from keras.utils import plot_model
from IPython.display import SVG
from keras.utils.vis_utils import model_to_dot
K.set_image_data_format('channels_last')
X_train, Y_train, X_test, Y_test, classes = data_service.load_dataset()
plot_service.plot_training_image(2, X_train, Y_train)
X_train, Y_train, X_test, Y_test, = data_service.norm_and_reshape(
X_train,
Y_train,
X_test,
Y_test,
)
print("number of training examples = " + str(X_train.shape[0]))
print("number of test examples = " + str(X_test.shape[0]))
print("X_train shape: " + str(X_train.shape))
print("Y_train shape: " + str(Y_train.shape))
print("X_test shape: " + str(X_test.shape))
print("Y_test shape: " + str(Y_test.shape))
m, rows, cols, channels = X_train.shape
happyModel = HappyModel((rows, cols, channels))
happyModel.compile(optimizer="adam",
loss="binary_crossentropy",
metrics=["accuracy"])