The XGBoostModel class in ADS is designed to allow you to rapidly get a XGBoost model into production. The .prepare() method creates the model artifacts that are needed to deploy a functioning model without you having to configure it or write code. However, you can customize the required score.py file.
The following steps take your trained XGBoost model and deploy it into production with a few lines of code.
The XGBoostModel module in ADS supports serialization for models generated from both the Learning API using xgboost.train() and the Scikit-Learn API using xgboost.XGBClassifier(). Both of these interfaces are defined by XGBoost.
Create Learning API and Scikit-Learn Wrapper XGBoost Models
In the following several code snippets you will prepare the data and train XGBoost models. In the first snippet, the data will be prepared. This will involved loading a dataset, splitting it into dependent and independent variables and into test and training sets. The data will be encoded and a preprocessing pipeline will be defined. In the second snippet, the XGBoost Learning API will be used to train the model. In the third and final code snippet, the Scikit-Learn Wrapper interface is used to create another XGBoost model.
import pandas as pd
import os
import tempfile
import xgboost as xgb
from ads.model.framework.xgboost_model import XGBoostModel
from sklearn.compose import ColumnTransformer
from sklearn.impute import SimpleImputer
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import LabelEncoder, OrdinalEncoder
df_path = os.path.join("/", "opt", "notebooks", "ads-examples", "oracle_data", "orcl_attrition.csv")
df = pd.read_csv(df_path)
y = df["Attrition"]
X = df.drop(columns=["Attrition", "name"])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
# Label encode the y values
le = LabelEncoder()
y_train_transformed = le.fit_transform(y_train)
y_test_transformed = le.transform(y_test)
# Extract numerical columns and categorical columns
categorical_cols = []
numerical_cols = []
for i, col in X.iteritems():
if col.dtypes == "object":
categorical_cols.append(col.name)
else:
numerical_cols.append(col.name)
categorical_transformer = Pipeline(
steps=[('encoder', OrdinalEncoder())]
)
# Build a pipeline
preprocessor = ColumnTransformer(
transformers=[('cat', categorical_transformer, categorical_cols)]
)
preprocessor_pipeline = Pipeline(steps=[('preprocessor', preprocessor)])
preprocessor_pipeline.fit(X_train)
X_train_transformed = preprocessor_pipeline.transform(X_train)
X_test_transformed = preprocessor_pipeline.transform(X_test)Create an XGBoost model using the Learning API.
dtrain = xgb.DMatrix(X_train_transformed, y_train_transformed)
dtest = xgb.DMatrix(X_test_transformed, y_test_transformed)
model_learn = xgb.train(
params = {"learning_rate": 0.01, "max_depth": 3},
dtrain = dtrain,
)Create an XGBoost model using the Scikit-Learn Wrapper interface.
model = xgb.XGBClassifier(
n_estimators=100, max_depth=3, learning_rate=0.01, random_state=42,
use_label_encoder=False
)
model.fit(
X_train_transformed,
y_train_transformed,
)Instantiate a XGBoostModel() object with an XGBoost model. Each instance accepts the following parameters:
artifact_dir: str: Artifact directory to store the files needed for deployment.auth: (Dict, optional): Defaults toNone. The default authentication is set using theads.set_authAPI. To override the default, useads.common.auth.api_keys()orads.common.auth.resource_principal()and create the appropriate authentication signer and the**kwargsrequired to instantiate theIdentityClientobject.estimator: (Callable): Trained XGBoost model either using the Learning API or the Scikit-Learn Wrapper interface.properties: (ModelProperties, optional): Defaults toNone. TheModelPropertiesobject required to save and deploy a model.
The prepare step is performed by the .prepare() method. It creates several customized files used to run the model after it is deployed. These files include:
input_schema.json: A JSON file that defines the nature of the features of theX_sampledata. It includes metadata such as the data type, name, constraints, summary statistics, feature type, and more.model.json: This is the default filename of the serialized model. It can be changed with themodel_file_nameattribute. By default, the model is stored in a JSON file. You can use theas_onnxparameter to save in ONNX format, and the model name defaults tomodel.onnx.output_schema.json: A JSON file that defines the nature of the dependent variable in they_sampledata. It includes metadata such as the data type, name, constraints, summary statistics, feature type, and more.runtime.yaml: This file contains information needed to set up the runtime environment on the deployment server. It has information about what conda environment was used to train the model and what environment to use to deploy the model. The file also specifies what version of Python should be used.score.py: This script contains theload_model()andpredict()functions. Theload_model()function understands the format the model file was saved in and loads it into memory. The.predict()method is used to make inferences in a deployed model. There are also hooks that allow you to perform operations before and after inference. You can modify this script to fit your specific needs.
To create the model artifacts you use the .prepare() method. There are a number of parameters that allow you to store model provenance information.
To serialize the model to ONNX format, set the as_onnx parameter to True. You can provide the initial_types parameter, which is a Python list describing the variable names and types. Alternatively, the service tries to infer this information from the data in the X_sample parameter. X_sample supports List, Numpy array or Pandas dataframe. DMatrix class is not supported because this format can't convert into a JSON serializable format, see the ONNX docs.
When using the Scikit-Learn Wrapper interface, the .prepare() method accepts any parameter that skl2onnx.convert_sklearn accepts. When using the Learning API, the .prepare() method accepts any parameter that onnxmltools.convert_xgboost accepts.
data: Any: Data used to test if deployment works in a local environment.
data: Any: JSON serializable data used for making inferences.
The .predict() and .verify() methods take the same data formats. You must ensure that the data passed into and returned by the predict() function in the score.py file is JSON serializable.
You can restore serialization models from model artifacts, from model deployments or from models in the model catalog. This section provides details on how to restore serialization models.
from ads.model.framework.xgboost_model import XGBoostModel
model = XGBoostModel.from_model_artifact(
uri="/folder_to_your/artifact.zip",
model_file_name="model.joblib",
artifact_dir="/folder_store_artifact"
)from ads.model.framework.xgboost_model import XGBoostModel
model = XGBoostModel.from_model_catalog(model_id="<model_id>",
model_file_name="model.json",
artifact_dir=tempfile.mkdtemp())from ads.model.generic_model import XGBoostModel
model = XGBoostModel.from_model_deployment(
model_deployment_id="<model_deployment_id>",
model_file_name="model.pkl",
artifact_dir=tempfile.mkdtemp())import pandas as pd
import os
import tempfile
import xgboost as xgb
from ads.catalog.model import ModelCatalog
from ads.common.model_metadata import UseCaseType
from ads.model.framework.xgboost_model import XGBoostModel
from sklearn.compose import ColumnTransformer
from sklearn.impute import SimpleImputer
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import LabelEncoder, OrdinalEncoder
df_path = os.path.join("/", "opt", "notebooks", "ads-examples", "oracle_data", "orcl_attrition.csv")
df = pd.read_csv(df_path)
y = df["Attrition"]
X = df.drop(columns=["Attrition", "name"])
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
# Label encode the y values
le = LabelEncoder()
y_train_transformed = le.fit_transform(y_train)
y_test_transformed = le.transform(y_test)
# Extract numerical columns and categorical columns
categorical_cols = []
numerical_cols = []
for i, col in X.iteritems():
if col.dtypes == "object":
categorical_cols.append(col.name)
else:
numerical_cols.append(col.name)
categorical_transformer = Pipeline(
steps=[('encoder', OrdinalEncoder())]
)
# Build a pipeline
preprocessor = ColumnTransformer(
transformers=[('cat', categorical_transformer, categorical_cols)]
)
preprocessor_pipeline = Pipeline(steps=[('preprocessor', preprocessor)])
preprocessor_pipeline.fit(X_train)
X_train_transformed = preprocessor_pipeline.transform(X_train)
X_test_transformed = preprocessor_pipeline.transform(X_test)
# XGBoost Scikit-Learn API
model = xgb.XGBClassifier(
n_estimators=100, learning_rate=0.01, random_state=42,
use_label_encoder=False
)
model.fit(X_train_transformed, y_train_transformed)
# Deploy the model, test it and clean up.
artifact_dir = tempfile.mkdtemp()
xgboost_model = XGBoostModel(estimator=model, artifact_dir=artifact_dir)
xgboost_model.prepare(
inference_conda_env="generalml_p37_cpu_v1",
training_conda_evn="generalml_p37_cpu_v1",
use_case_type=UseCaseType.BINARY_CLASSIFICATION,
X_sample=X_test_transformed,
y_sample=y_test_transformed,
)
xgboost_model.verify(X_test_transformed[:10])['prediction']
model_id = xgboost_model.save()
xgboost_model.deploy()
xgboost_model.predict(X_test_transformed[:10])['prediction']
xgboost_model.delete_deployment(wait_for_completion=True)
ModelCatalog(compartment_id=os.environ['NB_SESSION_COMPARTMENT_OCID']).delete_model(model_id)