def shoubi_model():
model = load_model(os.path.join(modelPath, shoubi))
model = Model(inputs=model.input, outputs=model.layers[1].output, name="shoubi_model")
model = Sequential(name="shoubi_model")
model.add(LSTM(64, input_shape=(30, 9), return_sequences=True, kernel_regularizer=tf.keras.regularizers.l2(0.0001)))
model.add(LSTM(64, kernel_regularizer=tf.keras.regularizers.l2(0.0001)))
# model.trainable = False
return model
dense2 = Dense(512)(dense1)
dense2 = LeakyReLU(0.1)(dense2)
dense3 = Dense(10)(dense2)
dense3 = LeakyReLU(0.1)(dense3)
output_position = Dense(1)(dense3)
model = Model(inputs=input_image1, outputs=output_position)
opt = Adam(lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0001)
model.compile(loss='mse', optimizer=opt)
## Discriminator
model = Sequential()
model.add(
Conv2D(1, (8, 8),
8,
padding='valid',
input_shape=(image_size, image_size, 1),
name='conv1'))
model.add(Activation('relu'))
model.add(Conv2D(1, (8, 8), 8, padding='valid', name='conv2'))
model.add(Activation('relu'))
model.add(Flatten())
model.add(Dense(10, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(2, 'softmax', name='output'))
opt = Adam(lr=0.001,
beta_1=0.9,
beta_2=0.999,
epsilon=None,
decay=0.0001,
