More Time Doesn't Always Help

Continuing on with Keras, there is a simple IMDB data set with movie ratings.  Building a quick model and training over a few epochs shows the training accuracy goes up to the high 90%. However, the validation accuracy hovers around 87%.

The model is over-fitting the data somehow. No matter how many more epochs I run, the validation won’t get better.