Machine learning optimization technique that uses. Used to find the optimal weighting for Hyperparameters.

  • A high-learning rate can cause model to converge too quickly.
  • Low-learning rate means model can behave in too closely to the prompt, perhaps missing a key solution early.