1. Learn
    2. /
  2. Courses
    3. /
  3. Recurrent Neural Networks (RNN) for Language Modeling in Python

Exercise

Transfer learning starting point

In this exercise you will see the benefit of using pre-trained vectors as a starting point for your model.

You will compare the accuracy of two models trained with two epochs. The architecture of the models is the same: One embedding layer, one LSTM layer with 128 units and the output layer with 5 units which is the number of classes in the sample data. The difference is that one model uses pre-trained vectors on the embedding layer (transfer learning) and the other doesn't.

The pre-trained vectors used were the GloVE with 200 dimension. The training accuracy history of the validation set of both models are available in the variables history_no_emb and history_emb.

Instructions

100 XP
  • Import module matplotlib.pyplot as plt.
  • Add the list of accuracy from the model without embeddings to the plot.
  • Add the list of accuracy from the model with embeddings to the plot.
  • Display the plot using the method .show().