1. Learn
    2. /
  2. Courses
    3. /
  3. Reinforcement Learning from Human Feedback (RLHF)
Connected

Exercise

Active learning loop

Now that you've set up your active learner, it's time to use it! In this exercise, you'll implement a loop that will allow to continuously improve the categorization of the data.

The dataset has been loaded with X_labeled for labeled training data, X_unlabeled for unlabeled training data, and y_labeled for labels.

The learner object has been pre-imported.

Instructions

100 XP
  • Implement a loop that will run 10 queries.
  • In each iteration, have the learner teach itself using the current labeled data.
  • Use the learner to query the most uncertain data points from the unlabeled data, setting the number of instances to 5.
  • Update the unlabeled dataset accordingly.