wehub-resource-sync
29 lines
848 B
Python
29 lines
848 B
Python
from sklearn.datasets import load_diabetes
|
|
from sklearn.linear_model import LinearRegression
|
|
from sklearn.model_selection import train_test_split
|
|
|
|
import mlflow
|
|
|
|
diabetes_dataset = load_diabetes()
|
|
|
|
X_train, X_test, y_train, y_test = train_test_split(
|
|
diabetes_dataset.data, diabetes_dataset.target, test_size=0.33, random_state=42
|
|
)
|
|
|
|
with mlflow.start_run() as run:
|
|
model = LinearRegression().fit(X_train, y_train)
|
|
model_info = mlflow.sklearn.log_model(model, name="model")
|
|
|
|
result = mlflow.evaluate(
|
|
model_info.model_uri,
|
|
X_test,
|
|
targets=y_test,
|
|
model_type="regressor",
|
|
evaluators="default",
|
|
feature_names=diabetes_dataset.feature_names,
|
|
evaluator_config={"explainability_nsamples": 1000},
|
|
)
|
|
|
|
print(f"metrics:\n{result.metrics}")
|
|
print(f"artifacts:\n{result.artifacts}")
|