1.ValueError: Only call `softmax_cross_entropy_with_logits` with named arguments (labels=..., logits=..., ...)
錯誤代碼：cross_entropy2=tf.reduce_sum(tf.nn.softmax_cross_entropy_with_logits_v2(logits, y_))
修正：cross_entropy2=tf.reduce_sum(tf.nn.softmax_cross_entropy_with_logits_v2(labels=logits, logits=y_))
2.WARNING:softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Future major versions of TensorFlow will allow gradients to flow
into the labels input on backprop by default.
See tf.nn.softmax_cross_entropy_with_logits_v2.
錯誤代碼：tf.nn.softmax_cross_entropy_with_logits
修正：tf.nn.softmax_cross_entropy_with_logits_v2