search

johnb30 / py_crepe

Keras implementation of the Crepe character-level convolutional neural net.
74 stars 33 forks source link

Update main.py #4

Closed fgadaleta closed 7 years ago

fgadaleta commented 7 years ago

it makes more sense to print average loss and average accuracy for test in order to check when overfitting.