解决 tensorflow softmax_cross_entropy_with_logits() 报错 Only call `softmax_cross_entropy_with_logits`

运行程序报错：

Traceback (most recent call last):
File "demo.py", line 74, in <module>
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(pred, y))
File "/python2.7/site-packages/tensorflow/python/ops/nn_ops.py", line 1558, in softmax_cross_entropy_with_logits
labels, logits)
File "/python2.7/site-packages/tensorflow/python/ops/nn_ops.py", line 1512, in _ensure_xent_args
"named arguments (labels=..., logits=..., ...)" % name)
ValueError: Only call `softmax_cross_entropy_with_logits` with named arguments (labels=..., logits=..., ...)

tf.nn.softmax_cross_entropy_with_logits()函数已经更新，官方API如下：

tf.nn.softmax_cross_entropy_with_logits(_sentinel=None, labels=None, logits=None, dim=-1, name=None)

至少需要label和logistics参数。

label是神经网络目标输出

logistics是神经网络实际输出

因此需要将

cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(pred, y))

修改为

cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, labels=y))

Ref:

http://blog.csdn.net/caimouse/article/details/61208940