forked from gxwangupc/FAAL
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathlosses.py
46 lines (27 loc) · 1.3 KB
/
losses.py
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
import tensorflow as tf
def classifier_loss(y_true, y_pred):
loss = tf.keras.losses.categorical_crossentropy(y_true, y_pred, from_logits=False, label_smoothing=0)
return loss
def dcgan_loss():
criterion = tf.keras.losses.BinaryCrossentropy(from_logits=False)
def d_loss(real_logits, fake_logits):
real_loss = criterion(tf.ones_like(real_logits), real_logits)
fake_loss = criterion(tf.zeros_like(fake_logits), fake_logits)
return real_loss + fake_loss
def g_loss(fake_logits):
return criterion(tf.ones_like(fake_logits), fake_logits)
return d_loss, g_loss
def d_loss(real_logits, fake_logits):
criterion = tf.keras.losses.BinaryCrossentropy(from_logits=False)
real_loss = criterion(tf.ones_like(real_logits), real_logits)
fake_loss = criterion(tf.zeros_like(fake_logits), fake_logits)
return real_loss + fake_loss
def g_loss(fake_logits):
criterion = tf.keras.losses.BinaryCrossentropy(from_logits=False)
return criterion(tf.ones_like(fake_logits), fake_logits)
def wgan_loss():
def d_loss(real_logits, fake_logits):
return tf.reduce_mean(fake_logits) - tf.reduce_mean(real_logits)
def g_loss(fake_logits):
return -tf.reduce_mean(fake_logits)
return d_loss, g_loss