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
Perplexity
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 data. It is also used to compare probability models. A low perplexity indicates that the probability distribution is good at predicting the sample. Even during training, you can see that for each checkpoint, perplexity is decreasing. Ideally, when there...
