Layer weight shape (32, 80, 8, 8) not compatible with provided weight shape (8, 8, 4, 32)

Question

Layer weight shape (32, 80, 8, 8) not compatible with provided weight shape (8, 8, 4, 32)

tangyudi opened this issue 7 years ago · comments

I run your code but it occurs a error,Layer weight shape (32, 80, 8, 8) not compatible with provided weight shape (8, 8, 4, 32)
Can you help me to solve this problem?

Ben Lau · Answer 1 · Fri Mar 17 2017 17:13:39 GMT+0800 (China Standard Time)

Are you using theano or tensorflow?

…

On Fri, Mar 17, 2017 at 4:28 PM, TangYudi ***@***.***> wrote: I run your code but it occurs a error,Layer weight shape (32, 80, 8, 8) not compatible with provided weight shape (8, 8, 4, 32) Can you help me to solve this problem? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#11>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AO1sY4xRqakFw_wzHpbMuKsDrB40Rhcyks5rmkQngaJpZM4MgTX8> .

TangYudi · Answer 2 · Fri Mar 17 2017 17:24:25 GMT+0800 (China Standard Time)

@yanpanlau I use tensorflow.
Thanks for help me!

Ben Lau · Answer 3 · Fri Mar 17 2017 19:30:06 GMT+0800 (China Standard Time)

Can you try to download everything from beginning?

Delete the original folder

git clone https://github.com/yanpanlau/Keras-FlappyBird.git

and then type

python qlearn.py -m "Run"

TangYudi · Answer 4 · Sat Mar 18 2017 17:19:59 GMT+0800 (China Standard Time)

@yanpanlau I clone it again.But it still occurs the error,Layer weight shape (32, 80, 8, 8) not compatible with provided weight shape (8, 8, 4, 32).
Can you help me to solve it? Thank you!

Arinc Elhan · Answer 5 · Thu Mar 23 2017 19:11:14 GMT+0800 (China Standard Time)

I have the same error, i am using tensorflow with keras 1.0.7 version. Did you guys find any solution to this ?

gitricko · Answer 6 · Wed Mar 29 2017 09:52:34 GMT+0800 (China Standard Time)

same problem for me

Ben Lau · Answer 7 · Wed Mar 29 2017 19:45:39 GMT+0800 (China Standard Time)

Can you please let me know what version of tensorflow and keras using so i can replicate the problem? Thanks 2017年3月28日 21:52 於 "gitricko" <notifications@github.com> 寫道：

…

same problem for me — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#11 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AO1sY4xiBySmKrcJ3z4Tgz3Ba8DIqigVks5rqbljgaJpZM4MgTX8> .

Duy Do · Answer 8 · Tue Jun 06 2017 18:36:18 GMT+0800 (China Standard Time)

Hi, I use keras 1.0.0, tensorflow 1.1.0 and has the same issue as above. In addition, I have another error:

  File "qlearn.py", line 210, in <module>
    K.set_session(sess)
AttributeError: 'module' object has no attribute 'set_session'

The code runs fine without set_session line. I don't know what that line means and how to fix it. Can you help me?! Thanks for supporting!

Alaa Megahed · Answer 9 · Wed Jul 05 2017 03:45:33 GMT+0800 (China Standard Time)

I have the same problem. I'm using tensorflow and keras 1.0.0. Any solution was found?
UPDATE: it works fine with keras version 2.0.5

Duc Cao · Answer 10 · Wed Feb 21 2018 00:12:43 GMT+0800 (China Standard Time)

The code at commit 25daf5fc1adfb9dc95854e355eafefc76da72331 works fine for me.

wmonges · Answer 11 · Thu Mar 29 2018 10:26:31 GMT+0800 (China Standard Time)

I am trying to run it but I got the same error.
My tensorflow version is 1.6.0-cp27
Keras version is 2.1.5
python 2.7

Ben Lau · Answer 12 · Wed Apr 04 2018 16:59:38 GMT+0800 (China Standard Time)

I just update the code and re-test it. Can you try again?