有可能的致错原因
tf.summary.scalar(‘loss’,loss)
merged = tf.summary.merge_all()#save
saver = tf.train.Saver(tf.global_variables())
sv = tf.train.Supervisor(logdir=xx,saver=saver）with sv.managed_session(config=config) as sess:
这样写没有报上述错误
然而：下述顺序会报错
tf.summary.scalar(‘loss’,loss)
#save
saver = tf.train.Saver(tf.global_variables())
sv = tf.train.Supervisor(logdir=xx,saver=saver）merged = tf.summary.merge_all()
with sv.managed_session(config=config) as sess:
可能的原因是：
https://stackoverflow.com/questions/41798311/tensorflow-graph-is-finalized-and-cannot-be-modified
中提到The root cause for your error seems to be that MonitoredTrainingSession has finalized (frozen) the graph and your
tf.global_variable_initializer()
is no longer able to modify it.
可能是saver中涉及冻结图标的操作，但是这和merge_all有何关系扔不懂:-D….