Skip to content

easonlai/pneumonia_classification_challenge_databricks

Repository files navigation

Pneumonia Classification Challenge Databricks

This is sample repos for how to use Keras Tuner to perform hyper-parameter tuning in Databricks.

  • pneumonia_classification_challenge_v1.ipynb <-- Origional model training without Keras Tuner.
  • pneumonia_classification_challenge_v2.ipynb <-- Using Keras Tuner for hyper-paramter tuning, to search best combination for highest validation accurary.
  • pneumonia_classification_challenge_v3_resnet50.ipynb <-- Using ResNet50 model architecture to train model.
  • Keras Tuner documentation
  • Hyper-parameter Auto Tuning (Keras Tuner with Databricks)