# -*- encoding:utf-8 -*-
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import numpy as np
from lightnn.models.models import Sequential, Model
from lightnn.layers.core import Dense, Flatten, Softmax, Input, Dropout, Activation
from lightnn.layers.convolutional import Conv2d
from lightnn.layers.pooling import MaxPooling, AvgPooling
from lightnn.base.activations import Relu, Selu
from lightnn.base.optimizers import SGD, Momentum, RMSProp, Adam, Adagrad, Adadelta
def mlp_random():
"""test MLP with random data and Sequential
"""
input_size = 600
input_dim = 20
label_size = 10
train_X = np.random.random((input_size, input_dim))
train_y = np.zeros((input_size, label_size))
for _ in xrange(input_size):
train_y[_,np.random.randint(0, label_size)] = 1
model = Sequential()
model.add(Input(input_shape=input_dim))
model.add(Dense(100, activator='selu'))
model.add(Softmax(label_size))
model.compile('CCE')
model.fit(train_X, train_y, verbose=1)
def mlp_mnist():
"""test MLP with MNIST data and Sequential
"""
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets('/tmp/data', one_hot=True)
training_data = np.array([image.flatten() for image in mnist.train.images])
training_label = mnist.train.labels
valid_data = np.array([image.flatten() for image in mnist.validation.images])
valid_label = mnist.validation.labels
input_dim = training_data.shape[1]
label_size = training_label.shape[1]
model = Sequential()
model.add(Input(input_shape=(input_dim, )))
model.add(Dense(300, activator='selu'))
model.add(Dropout(0.2))
model.add(Softmax(label_size))
model.compile('CCE', optimizer=SGD())
model.fit(training_data, training_label, validation_data=(valid_data, valid_label))
def cnn_random():
"""test CNN with random data and Sequential
"""
input_size = 600
input_dim = 28
input_depth = 1
label_size = 10
train_X = np.random.random((input_size, input_dim, input_dim, input_depth))
train_y = np.zeros((input_size, label_size))
for _ in xrange(input_size):
train_y[_,np.random.randint(0, label_size)] = 1
model =Sequential()
model.add(Input(batch_input_shape=(None, 28, 28, 1)))
model.add(Conv2d((3, 3), 1, activator='relu'))
model.add(AvgPooling((2, 2), stride=2))
model.add(Conv2d((4, 4), 2, activator='relu'))
model.add(AvgPooling((2, 2), stride=2))
model.add(Flatten())
model.add(Softmax(label_size))
model.compile('CCE', optimizer=SGD(1e-2))
model.fit(train_X, train_y)
def cnn_mnist():
"""test CNN with MNIST data and Sequential
"""
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets('/tmp/data', one_hot=True)
training_data = np.array([image.reshape(28, 28, 1) for image in mnist.train.images])
training_label = mnist.train.labels
valid_data = np.array([image.reshape(28, 28, 1) for image in mnist.validation.images])
valid_label = mnist.validation.labels
label_size = training_label.shape[1]
model =Sequential()
model.add(Input(batch_input_shape=(None, 28, 28, 1)))
model.add(Conv2d((3, 3), 1, activator='selu'))
model.add(AvgPooling((2, 2), stride=2))
model.add(Conv2d((4, 4), 2, activator='selu'))
model.add(AvgPooling((2, 2), stride=2))
model.add(Flatten())
model.add(Softmax(label_size))
model.compile('CCE', optimizer=SGD(lr=1e-2))
model.fit(training_data, training_label, validation_data=(valid_data, valid_label), verbose=2)
def model_mlp_random():
"""test MLP with random data and Model
"""
input_size = 600
input_dim = 20
label_size = 10
train_X = np.random.random((input_size, input_dim))
train_y = np.zeros((input_size, label_size))
for _ in xrange(input_size):
train_y[_,np.random.randint(0, label_size)] = 1
input = Input(input_shape=input_dim)
d1 = Dense(100, activator='selu')(input)
s1 = Softmax(label_size)(d1)
model = Model(input, s1)
model.compile('CCE')
model.fit(train_X, train_y, verbose=1)
def model_mlp_mnist():
"""test MLP with MNIST data and Model
"""
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets('/tmp/data', one_hot=True)
training_data = np.array([image.flatten() for image in mnist.train.images])
training_label = mnist.train.labels
valid_data = np.array([image.flatten() for image in mnist.validation.images])
valid_label = mnist.validation.labels
input_dim = training_data.shape[1]
label_size = training_label.shape[1]
dense_1 = Dense(300, input_dim=input_dim, activator=None)
dense_2 = Activation('selu')(dense_1)
dropout_1 = Dropout(0.2)(dense_2)
softmax_1 = Softmax(label_size)(dropout_1)
model = Model(dense_1, softmax_1)
model.compile('CCE', optimizer=Adadelta())
model.fit(training_data, training_label, validation_data=(valid_data, valid_label))
if __name__ == '__main__':
# mlp_random()
# mlp_mnist()
# cnn_random()
# cnn_mnist()
# model_mlp_random()
model_mlp_mnist()
