Training the model

Let us define the training function.

%%writefile train-model.py


import pandas as pd
from sklearn.datasets import load_breast_cancer
from sklearn.model_selection import train_test_split

from digitalhub_runtime_python import handler
from sklearn.svm import SVC
from pickle import dump
import sklearn.metrics
import os

@handler(outputs=["dataset"])
def train(project, di):

    df_cancer = di.as_df()
    X = df_cancer.drop(['target'],axis=1)
    y = df_cancer['target']
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.20, random_state=5)
    svc_model = SVC()
    svc_model.fit(X_train, y_train)
    y_predict = svc_model.predict(X_test)

    if not os.path.exists("model"):
        os.makedirs("model")

    with open("model/cancer_classifier.pkl", "wb") as f:
        dump(svc_model, f, protocol=5)

    metrics = {
        "f1_score": sklearn.metrics.f1_score(y_test, y_predict),
        "accuracy": sklearn.metrics.accuracy_score(y_test, y_predict),
        "precision": sklearn.metrics.precision_score(y_test, y_predict),
        "recall": sklearn.metrics.recall_score(y_test, y_predict),
    }
    project.log_model(
            name="cancer_classifier",
            kind="sklearn",
            source="./model/",
            metrics=metrics
    )

The function takes the analysis dataset as input, creates an SVC model with the scikit-learn framework and logs the model with its metrics.

Let us register it:

train_fn = project.new_function(name="train",
                                kind="python",
                                python_version="PYTHON3_10",
                                code_src="train-model.py",
                                handler="train",
                                requirements=["scikit-learn==1.2.2"])

and run it locally:

train_run = train_fn.run(action="job", inputs={"di": gen_data_run.output("dataset").key})

As a result, a new model is registered in the Core and may be used by different inference operations:

model = project.get_model("cancer_classifier")
model.spec.path

Lastly, we'll deploy and test the model.