問題一：TypeError: Expected int32, got list containing Tensors of type ‘_Message’ instead.
tensorflow 函數tf.cocat([fw,bw],2)出錯:
Expected int32, got list containing Tensors of type ‘_Message’ inst
查看原因是11版本的函數形式爲:tf.concat(2,[fw,bw]),即應把串聯的維度與串聯值位置調換即可.
問題二：Input ‘split_dim’ of ‘Split’ Op has type float32 that does not match expected type of int32
This is because in Tensorflow versions < 0.12.0 the split function takes the arguments as:
x = tf.split(0, n_steps, x) # tf.split(axis, num_or_size_splits, value)
The tutorial you are working from was written for versions > 0.12.0, which has been changed to be consistent with Numpy’s split syntax:
x = tf.split(x, n_steps, 0) # tf.split(value, num_or_size_splits, axis)
問題三：TypeError: concat() got an unexpected keyword argument ‘axis’
tf.concat(concat_dim=axis, values=inputs, name=name)
修改爲： tf.concat(inputs,1,name=name)
問題四：ValueError: ‘size’ must be a 1-D Tensor of 2 elements
img = tf.image.resize_images(img, new_shape[0], new_shape[1])
改爲
img = tf.image.resize_images(img, new_shape)
問題五： ‘module’ object has no attribute ‘pack’
因爲TF後面的版本修改了這個函數的名稱，把 tf.pack 改爲 tf.stack。
問題六：The value of a feed cannot be a tf.Tensor object. Acceptable feed values include Python scalars, strings, lists, or numpy ndarrays
數據集是feed輸入的，feed的數據格式是有要求的
解決：img,label = sess.run[img,label],用返回值
問題七：AttributeError: ‘module’ object has no attribute ‘per_image_whitening’
For anyone else who has this problem, per_image_whitening was replaced by per_image_standardization in v0.12.
問題八：AttributeError: ‘module’ object has no attribute ‘image_summary’
tf.image_summary should be renamed to tf.summary.image;
問題九：AttributeError: ‘module’ object has no attribute ‘mul’
tf.mul(a,b) 這裏的矩陣a和矩陣b的shape必須相等 tf.mul()是矩陣的element-wise相乘（即Hadamard乘積）
tf.matmul(a,b) 這裏的矩陣a和矩陣b的shape應是a的行數對等與b的列數，tf.matmul()是矩陣的一般相乘。
解決：[tf.mul，tf.sub ] 和 [tf.neg] 不再使用，改爲 [tf.multiply]，[tf.subtract] 和 [tf.negative]。
問題十：AttributeError: ‘module’ object has no attribute ‘scalar_summary’
修改爲:tf.summary.scalar(‘batch_loss’, loss)原因:新版本做了調整 …
