"ValueError: Only call sparse_softmax_cross_entropy_with_logits with named arguments" encountered in the training loop

[bernt73]

Hi,

I'm using TF1.3 with Cuda 8.0 and CuDNN 6.0. May I know if this error could be caused by incompatible Tensorflow version? If so, could you share with us the version of TF/Cuda/CuDnn that you used to run the code?

[cullinap]

I get this same issue but when I am using CNN

[cullinap]

What I did to make it work was the following:

[Linkersem]

after this, i met another question,do you also has this question? or could you tell me how to solve this?

Traceback (most recent call last):
File "C:\Users\Administrator\Desktop\test\CNN_G.py", line 257, in
run_training()
File "C:\Users\Administrator\Desktop\test\CNN_G.py", line 166, in run_training
loss = IndianPinesCNN.loss(logits, labels_placeholder)
File "C:\Users\Administrator\Desktop\test\IndianPinesCNN.py", line 135, in loss
logits=logits,labels=labels, name='xentropy')
File "E:\python3.5\lib\site-packages\tensorflow\python\ops\nn_ops.py", line 2653, in sparse_softmax_cross_entropy_with_logits
logits.get_shape()))
ValueError: Shape mismatch: The shape of labels (received (100,)) should equal the shape of logits except for the last dimension (received (225, 16)).