This is an adaptation from this Google Cloud official tutorial. I adapted the models trained to make this tutorial more realistic and useful for customers with no deep learning experience.
The original tutorial uses the Vertex AI Python SDK to create a new experiment, train three models:
- a scikit-learn one
- a TensorFlow one
- a PyTorch one
and then automatically logs the results of the training experiments.
My adapted tutorial -notebook Experiments - Autolog 5 models' parameters and metrics in this repo- leverages the same API to train five simple models:
- a scikit-learn model
- an LGBM one
- three XGBoost models with different hyperparameters
and automatically log the experiment's input and output results to be later reviewed in Vertex AI Experiments UI.
