Recently, I was looking at the algorithm of MKR. After the reproduction runs through, save the trained model.
Just add the following code
saver = tf.train.Saver()
saver.save(sess, '../model/MKRmodel.ckpt')
print('模型已保存')
The above code should be indented with sess.run(), otherwise an error will be reported.
The intermediate code is too long, the first time I wrote to the indented place in front of sess.run(), an error message told me that sess has been closed. . .