Keras Tuner further possibilities


I was doing the C3_W1 Lab of the Keras Autotuner, and I wonder if these techniques could be used using Transfer Learning on custom datasets (which I guess it is definetly possible and will probably try in the near future), but furthermore, I would like to know if any of you has tried using these kind of auto hyperparameter tuning for the TensorFlow Object Detection API (or something similar).

Thank you guys, I wish you good fortune in the courses to come.

1 Like

Hyperparameter tuning doesn’t place any restrictions on the underlying model or data. The goal is to pick the best hyperparameters based on the evaluation criteria of your model. Please go ahead with applying keras tuner for your transfer learning problem.