Train on 406708 samples, validate on 174304 samples
Epoch 1/26
406708/406708 [==============================] - 9s 23us/step - loss: 3.2084 -
accuracy: 0.5765 - val_loss: 1.1794 - val_accuracy: 0.6811
Epoch 2/26
406708/406708 [==============================] - 10s 23us/step - loss: 1.1763 -
accuracy: 0.6387 - val_loss: 0.8447 - val_accuracy: 0.6538
Epoch 3/26
406708/406708 [==============================] - 10s 25us/step - loss: 0.8390 -
accuracy: 0.6687 - val_loss: 0.7099 - val_accuracy: 0.7046
Epoch 4/26
406708/406708 [==============================] - 10s 23us/step - loss: 0.6936 -
accuracy: 0.7037 - val_loss: 0.6578 - val_accuracy: 0.7269
Epoch 5/26
406708/406708 [==============================] - 10s 25us/step - loss: 0.6383 -
accuracy: 0.7252 - val_loss: 0.5917 - val_accuracy: 0.7517
...(중간 생략)...
Epoch 21/26
406708/406708 [==============================] - 11s 27us/step - loss: 0.5133 -
accuracy: 0.7821 - val_loss: 0.5160 - val_accuracy: 0.7827
Epoch 22/26
406708/406708 [==============================] - 10s 25us/step - loss: 0.5144 -
accuracy: 0.7819 - val_loss: 0.4919 - val_accuracy: 0.7941
Epoch 23/26
406708/406708 [==============================] - 11s 27us/step - loss: 0.5103 -
accuracy: 0.7838 - val_loss: 0.4969 - val_accuracy: 0.7932
Epoch 24/26
406708/406708 [==============================] - 11s 27us/step - loss: 0.5086 -
accuracy: 0.7847 - val_loss: 0.4857 - val_accuracy: 0.7963
Epoch 25/26
406708/406708 [==============================] - 10s 25us/step - loss: 0.5065 -
accuracy: 0.7852 - val_loss: 0.5235 - val_accuracy: 0.7809
Epoch 26/26
406708/406708 [==============================] - 10s 26us/step - loss: 0.5052 - accuracy: 0.7856 - val_loss: 0.5003 - val_accuracy: 0.7881