调用tf.softmax_cross_entropy_with_logits函数出错解决
运行一个程序时提示出错如下：
Traceback (most recent call last):
File "/MNIST/softmax.py", line 12, in
cross_entropy2=tf.reduce_sum(tf.nn.softmax_cross_entropy_with_logits(logits, y_))#dont forget tf.reduce_sum()!!
File "C:\python35\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 1578, in softmax_cross_entropy_with_logits
labels, logits)
File "C:\python35\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 1533, in _ensure_xent_args
"named arguments (labels=..., logits=..., ...)" % name)
ValueError: Only call `softmax_cross_entropy_with_logits` with named arguments (labels=..., logits=..., ...)
原来这个函数，不能按以前的方式进行调用了，只能使用命名参数的方式来调用。原来是这样的：
tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(y, y_))
因此修改需要成这样：
tf.reduce_sum(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=y_))
例子完整代码如下：
#python 3.5.3
#2017-03-09 蔡军生 http://blog.csdn.net/caimouse
#
import tensorflow as tf
#our NN's output
logits = tf.constant([[1.0,2.0,3.0],[1.0,2.0,3.0],[1.0,2.0,3.0]])
#step1:do softmax
y = tf.nn.softmax(logits)
#true label
y_ = tf.constant([[0.0,0.0,1.0],[0.0,0.0,1.0],[0.0,0.0,1.0]])
#step2:do cross_entropy
cross_entropy = -tf.reduce_sum(y_*tf.log(y))
#do cross_entropy just one step
cross_entropy2 = tf.reduce_sum(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=y_))
with tf.Session() as sess:
softmax=sess.run(y)
c_e = sess.run(cross_entropy)
c_e2 = sess.run(cross_entropy2)
print("step1:softmax result=")
print(softmax)
print("step2:cross_entropy result=")
print(c_e)
print("Function(softmax_cross_entropy_with_logits) result=")
print(c_e2)
输出结果：
step1:softmax result=
[[ 0.09003057 0.24472848 0.66524094]
[ 0.09003057 0.24472848 0.66524094]
[ 0.09003057 0.24472848 0.66524094]]
step2:cross_entropy result=
1.22282
Function(softmax_cross_entropy_with_logits) result=
1.22282