Setup model training using CML
In this exercise, you will use CML GitHub Action to train a Random Forest Classifier to predict rainfall. CML is a GitHub Action that abstracts generating reports for ML experiments.
The training will trigger when you open a PR against the main
branch.
You'll continue working with the weather dataset; the preprocess_dataset.py
file contains helper functions to pre-process the dataset as before.
The output from running train.py
is a metrics.json
file containing model metrics, and confusion_matrix.png
file containing a plot of the confusion matrix.
Your task is to finish the scaffolded .github/workflows/train_cml.yaml
to formulate a high-level model training flow.
NOTE: Use python3
instead of python
to run Python scripts.
This exercise is part of the course
CI/CD for Machine Learning
Hands-on interactive exercise
Turn theory into action with one of our interactive exercises
