NN_work/mlp.py at master · lkshay/NN_work · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
#!/usr/bin/env python
# -*- coding: utf-8 -*-
# File: mlp.py
# Author: Qian Ge <qge2@ncsu.edu>

import sys
import argparse
import numpy as np
import matplotlib.pyplot as plt
from time import sleep
'''
sys.path.append('../')
import src.network2 as network2
import src.mnist_loader as loader
import src.activation as act
'''

import network2 as network2
import mnist_loader as loader
import activation as act


DATA_PATH = '../MNIST'

def get_args():
    parser = argparse.ArgumentParser()
    parser.add_argument('--input', action='store_true',
                        help='Check data loading.')
    parser.add_argument('--sigmoid', action='store_true',
                        help='Check implementation of sigmoid.')
    parser.add_argument('--gradient', action='store_true',
                        help='Gradient check')
    parser.add_argument('--train', action='store_true',
                        help='Train the model')


    return parser.parse_args()

def load_data():
    train_data, valid_data, test_data = loader.load_data_wrapper(DATA_PATH)
    print('Number of training: {}'.format(len(train_data[0])))
    print('Number of validation: {}'.format(len(valid_data[0])))
    print('Number of testing: {}'.format(len(test_data[0])))
    return train_data, valid_data, test_data

def test_sigmoid():
    z = np.arange(-10, 10, 0.1)
    y = act.sigmoid(z)
    y_p = act.sigmoid_prime(z)

    plt.figure()
    plt.subplot(1, 2, 1)
    plt.plot(z, y)
    plt.title('sigmoid')

    plt.subplot(1, 2, 2)
    plt.plot(z, y_p)
    plt.title('derivative sigmoid')
    plt.show()

def gradient_check():
    train_data, valid_data, test_data = load_data()
    model = network2.Network([784, 50, 10])
    model.gradient_check(training_data=train_data, layer_id=1, unit_id=5, weight_id=3)

def main():
    epochs = 1000
    # load train_data, valid_data, test_data
    train_data, valid_data, test_data = load_data()
    # construct the network

    # model = network2.Network([784, 60, 10])
    model = network2.load("mymodel2.json")
    #train the network using SGD

    [evaluation_cost, evaluation_accuracy,training_cost, training_accuracy] = model.SGD(
        training_data=train_data,
        epochs=epochs,
        mini_batch_size=500,
        eta=1.9e-3,
        lmbda = 0.00001,
        evaluation_data=test_data,
        monitor_evaluation_cost=True,
        monitor_evaluation_accuracy=True,
        monitor_training_cost=True,
        monitor_training_accuracy=True)

    epoch_list = []

    for i in range(epochs):
        epoch_list.append(i)

    # print((epoch_list),(training_cost))
    plt.figure()
    plt.plot(epoch_list,training_cost,"r-")
    plt.ylabel("Training Cost")
    plt.xlabel("Epochs")
    plt.figure()
    plt.plot(epoch_list,training_accuracy,"r-")
    plt.ylabel("Training Accuracy")
    plt.xlabel("Epochs")
    plt.figure()
    plt.plot(epoch_list,evaluation_cost,"b-")
    plt.ylabel("Evaluation Cost")
    plt.xlabel("Epochs")
    plt.figure()
    plt.plot(epoch_list,evaluation_accuracy,"b-")
    plt.ylabel("Evaluation Accuracy")
    plt.xlabel("Epochs")

    model.save("mymodel2.json")

    # sleep(2)

    # trained_model = network2.load("mymodel.json")
    # test_result = trained_model.accuracy(test_data,convert = False)
    # print test_result

    plt.show()

if __name__ == '__main__':
    FLAGS = get_args()
    if FLAGS.input:
        load_data()
    if FLAGS.sigmoid:
        test_sigmoid()
    if FLAGS.train:
        main()
    if FLAGS.gradient:
        gradient_check()