AutoEncoder/train_ae.py at master · eleferrand/AutoEncoder · 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
131
132
133
134
135
136
import tensorflow as tf
import data
import batching
from tensorflow.keras.datasets import mnist
import numpy as np
import os
import _pickle as pickle
tf.enable_eager_execution()


class AE_layer(tf.keras.Model):
    def __init__(self):
        super(AE_layer, self).__init__()
        self.Encoder = tf.keras.layers.Dense(100, activation=tf.nn.tanh)
        self.bottleneck = tf.keras.layers.Dense(100, activation=tf.nn.tanh)
        self.Decoder = tf.keras.layers.Dense(100, activation=tf.nn.tanh)

    def call(self, inp):
        x = self.Encoder(inp)
        bn = self.bottleneck(x)
        x = self.Decoder(bn)
        return x, bn

class AutoEncoder(tf.keras.Model):
    def __init__(self):
        super(AutoEncoder, self).__init__()
        self.ae1 = AE_layer()
        self.ae2 = AE_layer()
        self.ae3 = AE_layer()
        self.ae4 = AE_layer()
        self.ae5 = AE_layer()
        self.ae6 = AE_layer()
        self.ae7 = AE_layer()

        self.LastEnc = tf.keras.layers.Dense(100, activation=tf.nn.tanh)

        self.LastBN = tf.keras.layers.Dense(39, activation=tf.nn.tanh)

        self.LastDec = tf.keras.layers.Dense(100, activation=tf.nn.tanh)

        self.dense_final = tf.keras.layers.Dense(512, activation=tf.nn.relu)


    def call(self, inp):
        x, bn = self.ae1(inp)
        x, bn = self.ae2(bn)
        x, bn = self.ae3(bn)
        x, bn = self.ae4(bn)
        x, bn = self.ae5(bn)
        x, bn = self.ae6(bn)
        x, bn = self.ae7(bn)
        x = self.LastEnc(bn)
        bn = self.LastBN(x)
        x = self.LastDec(bn)

        x = self.dense_final(x)
        return x, bn

def trunc_and_limit_dim(x, lengths, d_frame, max_length):
    for i, seq in enumerate(x):
        x[i] = x[i][:max_length, :d_frame]

        lengths[i] = min(lengths[i], max_length)


def loss(x, x_bar):
    return tf.losses.mean_squared_error(x, x_bar)
def grad(model, inputs):
    with tf.GradientTape() as tape:
        x, bn = model(inputs)
        loss_value = loss(inputs, x)
    return loss_value, tape.gradient(loss_value, model.trainable_variables), inputs, x

def main():
    batch_size = 300
    input_x = tf.keras.Input(shape=(100,39))
    autoencoder = AutoEncoder()
    optimizer = tf.train.AdamOptimizer(learning_rate=0.001)
    global_step = tf.Variable(0)
    n_epochs = 150

    if os.path.isfile("train_wtv.pkl"):
        with open("train_wtv.pkl", mode="rb") as pfile:
            f = pickle.load(pfile)
        train_x = f[0]
        train_labels = f[1]
        train_lengths = f[2]
        train_keys = f[3]
    else:
        train_x, train_labels, train_lengths, train_keys = data.load_utt("train", "wtv", "smp")
    if os.path.isfile("dev_wtv.pkl"):
        with open("dev_wtv.pkl", mode="rb") as pfile:
            f = pickle.load(pfile)
        val_x = f[0]
        val_labels = f[1]
        val_lengths = f[2]
        val_keys = f[3]
    else:
        val_x, val_labels, val_lengths, val_keys = data.load_utt("val", "wtv", "smp")

    trunc_and_limit_dim(train_x, train_lengths, 512, 100)
    trunc_and_limit_dim(val_x, val_lengths, 512, 100)

    train_x=np.asarray(train_x)
    val_x= np.asarray(val_x)
    best_val = np.inf
    losses = []
    for epoch in range(n_epochs):
        print("Epoch: ", epoch)
        for x in range(0, len(train_x), batch_size):
            x_inp = train_x[x : x + batch_size]

            loss_value, grads, inputs, reconstruction = grad(autoencoder, x_inp)
            optimizer.apply_gradients(zip(grads, autoencoder.trainable_variables),
                                global_step)
            print("batch nb. {}      loss: {}".format(int(x/300), loss_value))
        v, bn = autoencoder(val_x)
        val_loss = loss(val_x, v)
        losses.append(loss_value)
        print("validation:          Loss: {} train loss : {}".format(loss(val_x, v),np.mean(losses)))
        if val_loss<best_val:
            best_val=val_loss
            if not os.path.isdir("./models"):
                os.mkdir("models")
            if not os.path.isdir("./models/AE_ckpt"):
                os.mkdir("models/AE_ckpt")
            autoencoder.save_weights("./models/AE_wtv_ckpt/ae_best_model.ckpt")
            print("new best model")
        if epoch == 0:
            autoencoder.save_weights("./models/AE_wtv_ckpt/ae_rand_model.ckpt")
        else:
            autoencoder.save_weights("./models/AE_wtv_ckpt/ae_last_model.ckpt")


if __name__ == "__main__":
    main()