Big difference acc and val_acc about cnn-for-sentence-classification-in-keras HOT 2 CLOSED

Using Theano backend.
WARNING (theano.sandbox.cuda): The cuda backend is deprecated and will be removed in the next release (v0.10).  Please switch to the gpuarray backend. You can get more information about how to switch at this URL:
 https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28gpuarray%29

Using gpu device 0: GeForce GTX 980 (CNMeM is enabled with initial size: 90.0% of memory, cuDNN 5110)
Load data...
x_train shape: (25000, 400)
x_test shape: (25000, 400)
Model type is CNN-rand
Train on 25000 samples, validate on 25000 samples
Epoch 1/10
7s - loss: 0.5782 - acc: 0.6365 - val_loss: 0.3192 - val_acc: 0.8677
Epoch 2/10
7s - loss: 0.3063 - acc: 0.8735 - val_loss: 0.2846 - val_acc: 0.8827
Epoch 3/10
7s - loss: 0.2539 - acc: 0.8977 - val_loss: 0.2727 - val_acc: 0.8882
Epoch 4/10
7s - loss: 0.2334 - acc: 0.9064 - val_loss: 0.2768 - val_acc: 0.8868
Epoch 5/10
7s - loss: 0.2143 - acc: 0.9162 - val_loss: 0.2894 - val_acc: 0.8825
Epoch 6/10
7s - loss: 0.2065 - acc: 0.9189 - val_loss: 0.2821 - val_acc: 0.8848
Epoch 7/10
7s - loss: 0.2007 - acc: 0.9218 - val_loss: 0.2855 - val_acc: 0.8844
Epoch 8/10
7s - loss: 0.1926 - acc: 0.9234 - val_loss: 0.2929 - val_acc: 0.8839
Epoch 9/10
7s - loss: 0.1822 - acc: 0.9274 - val_loss: 0.2986 - val_acc: 0.8810
Epoch 10/10
7s - loss: 0.1835 - acc: 0.9292 - val_loss: 0.2924 - val_acc: 0.8800

Process finished with exit code 0