Testing the revised approach

In this section, we will perform testing of the revised approach. Before performing actual testing and seeing how good or bad the chatbot conversation is, we need to understand the basic testing metrics that we will be using for this approach and for the best approach. These testing metrics help us evaluate the model accuracy. Let's understand the testing metrics first, and then we will move on to the testing of the revised approach.

Understanding the testing metrics

In this section, we need to understand the following testing metrics:

  • Perplexity
  • Loss


In the NLP domain, perplexity is also referred to as per-word perplexity. Perplexity is a measurement of how well a trained model predicts the output for unseen ...

Get Machine Learning Solutions now with the O’Reilly learning platform.

O’Reilly members experience live online training, plus books, videos, and digital content from nearly 200 publishers.