#model.add(Dense(nb_actions))
#model.compile(RMSprop(), 'MSE')
agent = Agent(model=model, memory_size=-1, nb_frames=nb_frames)
#model.save('/tmp/snake1.hdf5')
#agent.train(snake, batch_size=64, nb_epoch=10000, gamma=0.8)
#model.save('/tmp/snake2.hdf5')
#agent.play(snake)
snake.reset()
agent.clear_frames()
S = agent.get_game_data(snake)
game_over = False
frames = list()
frames.append(S[0])
while not game_over:
q = model.predict(S)[0]
possible_actions = snake.get_possible_actions()
q = [q[i] for i in possible_actions]
action = possible_actions[np.argmax(q)]
snake.play(action)
S = agent.get_game_data(snake)
frames.append(S[0])
game_over = snake.is_over()
print(np.asarray(frames).shape)
kerasimo.ToSVG('snake',
model,
np.array(frames),
showarrows=False,
columns=[1, 3, 3, 10, 10, 1])
X = list()
for i in range(0, ntest):
X.append(np.empty((5, 7, 1)))
for jj in range(0, 7):
for ii in range(0, 5):
if (digits[i * 7 + jj][ii] == '*'):
X[i][ii][jj][0] = 1.
else:
X[i][ii][jj][0] = 0.
X = np.array(X)
Y = np.identity(ntest)
model = Sequential()
model.add(ZeroPadding2D(padding=1, input_shape=(5, 7, 1)))
model.add(Conv2D(2, (2, 2), activation='relu'))
model.add(MaxPooling2D(pool_size=(3, 3)))
model.add(Dropout(0.10))
model.add(Flatten())
model.add(Dense(ntest, activation='sigmoid'))
sgd = SGD(lr=0.1)
model.compile(loss='binary_crossentropy', optimizer=sgd)
model.fit(X, Y, batch_size=1, epochs=2000)
print(model.predict_proba(X))
kerasimo.ToSVG('digits_cnn', model, X)
from keras.models import Sequential
from keras.layers.core import Dense, Dropout, Activation
from keras.optimizers import SGD
import numpy as np
from lib import kerasimo
X = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])
y = np.array([[0], [1], [1], [0]])
model = Sequential()
model.add(Dense(2, activation='tanh', input_dim=2))
model.add(Dense(1, activation='sigmoid'))
sgd = SGD(lr=0.1)
model.compile(loss='binary_crossentropy', optimizer=sgd)
model.fit(X, y, batch_size=1, epochs=1000)
print(model.predict_proba(X))
kerasimo.ToSVG('xor', model, X)
for ii in range(0, 5):
if (digits[i*7+jj][ii] == '*'): X[i][jj*5+ii] = 1
X = np.array(X);
print(X);
Y = np.array([
[1,0,0,0,0,0,0,0,0,0],
[0,1,0,0,0,0,0,0,0,0],
[0,0,1,0,0,0,0,0,0,0],
[0,0,0,1,0,0,0,0,0,0],
[0,0,0,0,1,0,0,0,0,0],
[0,0,0,0,0,1,0,0,0,0],
[0,0,0,0,0,0,1,0,0,0],
[0,0,0,0,0,0,0,1,0,0],
[0,0,0,0,0,0,0,0,1,0],
[0,0,0,0,0,0,0,0,0,1]
])
model = Sequential()
model.add(Dense(10, activation='sigmoid', input_dim=35))
sgd = SGD(lr=0.1)
model.compile(loss='binary_crossentropy', optimizer=sgd)
model.fit(X, Y, batch_size=1, epochs=2000)
print(model.predict_proba(X))
kerasimo.ToSVG('digits', model, X, columns=[5, 1])
from keras.models import Sequential
from keras.layers.core import Dense, Dropout, Activation
from keras.optimizers import SGD
import numpy as np
from lib import kerasimo
X = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])
y = np.array([[0], [0], [0], [1]])
model = Sequential()
model.add(Dense(1, activation='sigmoid', input_dim=2))
sgd = SGD(lr=0.1)
model.compile(loss='binary_crossentropy', optimizer=sgd)
model.fit(X, y, batch_size=1, epochs=1000)
print(model.predict_proba(X))
kerasimo.ToSVG('and', model, X)
model.add(
Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=input_shape))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(num_classes, activation='softmax'))
model.compile(loss=keras.losses.categorical_crossentropy,
optimizer=keras.optimizers.Adadelta(),
metrics=['accuracy'])
model.fit(X_train,
Y_train,
batch_size=batch_size,
epochs=epochs,
verbose=1,
validation_data=(X_test, Y_test))
score = model.evaluate(X_test, Y_test, verbose=0)
print('Test score:', score[0])
print('Test accuracy:', score[1])
kerasimo.ToSVG('mnist_cnn',
model,
X_test[20:40],
columns=[1, 4, 4, 4, 1, 1, 1, 1, 1, 1, 1],
showarrows=False)
from keras.models import Sequential
from keras.layers.core import Dense, Dropout, Activation
from keras.optimizers import SGD
import numpy as np
from lib import kerasimo
X = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])
y = np.array([[0], [1], [1], [0]])
model = Sequential()
model.add(Dense(1, activation='sigmoid', input_dim=2))
sgd = SGD(lr=0.1)
model.compile(loss='binary_crossentropy', optimizer=sgd)
model.fit(X, y, batch_size=1, epochs=1000)
print(model.predict_proba(X))
kerasimo.ToSVG('xor_wrong', model, X)
import keras
print(sys.argv)
X = np.array([
[1, 0, 0, 0, 0, 0, 0, 0],
[0, 1, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0, 0, 0],
[0, 0, 0, 1, 0, 0, 0, 0],
[0, 0, 0, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0],
[0, 0, 0, 0, 0, 0, 0, 1],
])
model = Sequential()
model.add(Dense(int(sys.argv[1]), activation='tanh', input_dim=8))
if (int(sys.argv[2]) != 0):
model.add(Dense(int(sys.argv[2]), activation='tanh'))
model.add(Dense(8, activation='sigmoid'))
sgd = SGD(lr=0.1)
model.compile(loss='binary_crossentropy', optimizer=sgd)
model.fit(X, X, batch_size=1, epochs=10000)
print(model.predict_proba(X))
kerasimo.ToSVG('encode%d%d' % (int(sys.argv[1]), int(sys.argv[2])), model, X)