Open yangyangwill opened 6 years ago
I have this problem too,but what is wrong is in file "train_faces", after running that file you didn`t get the savepath,you could change the code like this:
saver.save(sess, './train_faces.model', global_step=n*num_batch+i)
sys.exit(0)
print ('saver done')
and you will see if u have generate the path.
if you find a way to solve(or to say generate the path),plz tell me ,thanks!
sorry ,i make a mistake here. you should put the "print"before the function exit,or you won`t see the result either
I have this problem too?Have you solved this problem?Please reply me,thank you!
I ... emmm,delete the “if ”sentence at the end of the code,for that avoid me from generating the path...but i dont know why.there should be some reason in that sentence but i cant understand.You can try this way,but its not a good way to solve the prolem.
Excuse me, can you explain it clearly? Which "if" should I remove?
remove line 163,
if (n*num_batch+i) % 100 == 0
(i forget if this is the origin code,for i had changed it once)
and you could change the print in 161 and 166 to
`print(n*num_batch+i,loss,'loss')
print(n*num_batch+i,acc,'acc') `
then,if the acc>0.98,you will get the path
That's great, this is working, thank you!
Thank you, but when I run this program if I use the other people's faces it will also terms out 'True' Do you know why? Thank you so much!
为什么按照楼上的方法,我没能得到解决
you don't have enough samples. Acc can't reach 0.95
File "G:/pycharm/untitled1/venv/is_my_face.py", line 145, in
saver.restore(sess, tf.train.latest_checkpoint('.'))
File "G:\Anaconda\envs\tensorflow\lib\site-packages\tensorflow\python\training\saver.py", line 1751, in restore
raise ValueError("Can't load save_path when it is None.")
ValueError: Can't load save_path when it is None.
How to tackle this problem? Thanks!