Logging tuned models
You have been experimenting with different model hyperparameters and need to log your latest round of experiment results to MLflow, let's do it!
Deze oefening maakt deel uit van de cursus
Designing Forecasting Pipelines for Production
Oefeninstructies
- Set the experiment name as
"hyperparameter_tuning". - Loop over the index and rows of
df. - Start an MLflow run.
Praktische interactieve oefening
Probeer deze oefening eens door deze voorbeeldcode in te vullen.
# Set the experiment name
experiment_name = "____"
experiment_id = mlflow.create_experiment(experiment_name)
# Loop through the DataFrame
for idx, row in df.____():
# Start a run
with mlflow.____(experiment_id=____):
model_params = ml_models[row["model_label"]].get_params()
model_params["model_name"] = row["model_name"]
model_params["model_label"] = row["model_label"]
model_params["partition"] = row["partition"]
model_params["lags"] = list(range(1, 24))
model_params["date_features"] = ["month", "day", "dayofweek", "week", "hour"]
mlflow.log_params(model_params)
mlflow.log_metric("mape", row["mape"])
mlflow.log_metric("rmse", row["rmse"])