Hvass-Labs / TensorFlow-Tutorials

TensorFlow Tutorials with YouTube Videos
MIT License
9.27k stars 4.19k forks source link

Tensor flow compile error #51

Closed prabirsinha closed 6 years ago

prabirsinha commented 6 years ago

Hi ,

I am having the following error in running the CIFAR10 dataset training . Can you give me some where to fix this issue. " File "C:\Users\Prabir Sinha\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\ops\array_ops.py", line 1001, in concat ).assert_is_compatible_with(tensor_shape.scalar()) File "C:\Users\Prabir Sinha\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\tensor_shape.py", line 756, in assert_is_compatible_with raise ValueError("Shapes %s and %s are incompatible" % (self, other)) ValueError: Shapes (2, 1) and () are incompatible "

Thanks Prabir

Hvass-Labs commented 6 years ago

I think this is the first time I've heard about this error. Have you modified the code or perhaps not downloaded everything? Or are you not running the Notebook from the beginning?

You also don't write how this error is connected with the code for the CIFAR10 dataset. These two errors are inside TensorFlow. Please try and debug / make print-statements to find out where in the CIFAR10 code the problem is.

prabirsinha commented 6 years ago

Hi,

Let me explain what I did . I had download your tutorial from your url link https://github.com/Hvass-Labs/TensorFlow-Tutorials

and follow the steps shown in https://github.com/Hvass-Labs/TensorFlow-Tutorials/blob/master/06_CIFAR-10.ipynb .

Following is my tool setup

tf.version '0.12.0-rc0' pt.version '0.7.4'

As per your steps I copy all the lines and put into a file cifar10_summaries.py inside the folder where I download your tutorial. Now from Python IDLE window I source this file with no modification . When I run this file I encounter this error

" File "C:\Users\Prabir Sinha\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\ops\array_ops.py", line 1003, in concat ).assert_is_compatible_with(tensor_shape.scalar()) File "C:\Users\Prabir Sinha\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\tensor_shape.py", line 756, in assert_is_compatible_with raise ValueError("Shapes %s and %s are incompatible" % (self, other)) ValueError: Shapes (2, 1) and () are incompatible "

It may sound strange or weired to you but this is the fact. .

Appreciate if you can help here.

Thanks Prabir

On Sat, Oct 7, 2017 at 4:27 PM, Hvass-Labs notifications@github.com wrote:

Closed #51 https://github.com/Hvass-Labs/TensorFlow-Tutorials/issues/51.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Hvass-Labs/TensorFlow-Tutorials/issues/51#event-1282821183, or mute the thread https://github.com/notifications/unsubscribe-auth/AX1aUCGwm-HjNgVvcn5OCvzQ0A0DTl3Dks5sp1kzgaJpZM4PvTnl .

Hvass-Labs commented 6 years ago

You copied the contents of the Notebook into a .py file? Did the Notebook run OK? Perhaps you made a mistake when copying the code?

The two error messages are inside TensorFlow, not in my code.

"Shapes (2, 1) and () are incompatible" tells you that two tensors have the wrong shapes. The error appears to be a call to concat but I did a search and I don't seem to use concat anywhere in my code.

Please try and debug this yourself, there is really nothing I can do to help. When you find the solution then please write it here in case others have the same problem in the future.

felipeardilac commented 6 years ago

I had the same issue running the 06_CIFAR-10 tutorial, and I fix it updating my tensorflow version to 1.3.0 and now it works

python -m pip uninstall tensorflow pip install --upgrade https://storage.googleapis.com/tensorflow/windows/cpu/tensorflow-1.3.0-cp35-cp35m-win_amd64.whl

Also, I had another issue with prettytensor '0.7.4' TypeError: zeros_initializer() missing 1 required positional argument: 'shape' and I fix it changing : tf.zeros_initializer() to tf.constant_initializer(0.0) In the files: pretty_tensor_image_methods.py, pretty_tensor_loss_methods.py, pretty_tensor_methods.py and recurrent_networks.py located in the folder C:\Python35\Lib\site-packages\prettytensor

Hope it works for you

prabirsinha commented 6 years ago

Hi felipeardilac,

Thanks for the update . I am able to make it run successfully.

Regards Prabir

On Tue, Nov 14, 2017 at 6:47 PM, felipeardilac notifications@github.com wrote:

I had the same issue running the 06_CIFAR-10 tutorial, and I fix it updating my tensorflow version to 1.3.0 and now it works

python -m pip uninstall tensorflow pip install --upgrade https://storage.googleapis. com/tensorflow/windows/cpu/tensorflow-1.3.0-cp35-cp35m-win_amd64.whl

Also, I had another issue with prettytensor '0.7.4' TypeError: zeros_initializer() missing 1 required positional argument: 'shape' and I fix it changing : tf.zeros_initializer() to tf.constant_initializer(0.0) In the files: pretty_tensor_image_methods.py, pretty_tensor_loss_methods.py, pretty_tensor_methods.py and recurrent_networks.py located in the folder C:\Python35\Lib\site-packages\ prettytensor

Hope it works for you

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Hvass-Labs/TensorFlow-Tutorials/issues/51#issuecomment-344255742, or mute the thread https://github.com/notifications/unsubscribe-auth/AX1aUFd64wLKt5gpW8PBfbeHtfpljFuhks5s2ZLjgaJpZM4PvTnl .