'NoneType' object is not iterable
产生原因:
当sess.run的第一个参数列表和返回值列表不匹配时,会出现这个错误。比如:
1 2 3 4 5
| with tf.Session() as sess: init = tf.global_variables_initializer() sess.run(init)
eval, cost = sess.run(evaluation_step, feed_dict={xxx})
|
修改之后就可以了:
1 2 3 4 5
| with tf.Session() as sess: init = tf.global_variables_initializer() sess.run(init)
eval, cost = sess.run([evaluation_step, cost], feed_dict={xxx})
|