sklearn-model-to-pickle.py

from sklearn.feature_extraction.text import TfidfVectorizer
import pandas as pd
from sklearn import svm
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import pickle
from sklearn import linear_model
Project_path = "<path to the project folder>"


data = pd.read_csv('https://storage.googleapis.com/dataset-uploader/bbc/bbc-text.csv')
print (data.category.unique())
vectorizer = TfidfVectorizer(sublinear_tf=True, encoding='utf-8',
                             decode_error='ignore')


def train_bpsd(df, vectorizer):
    tfidf = vectorizer.fit(df["text"].values.astype('U'))

    X = vectorizer.fit_transform(df["text"].values.astype('U'))
    y = df['category']
    train_test(X, y)
    model = svm.LinearSVC()

    print("==fitting the model===")
    model.fit(X.A, y)

    print("==fit done\=")
    return model, tfidf



model_path = Project_path + "/08. Multi-class_text_classification/models/model.pickle"
vectorizer_path = Project_path + "/08. Multi-class_text_classification/models/vectorizer.pickle"
model, vectorizer_model = train_bpsd(data, vectorizer)
pickle.dump(model, open(model_path, 'wb'))
pickle.dump(vectorizer_model, open(vectorizer_path, "wb"))