use pickkles

app.py

@@ -11,6 +11,7 @@ from nltk.stem import WordNetLemmatizer
 from nltk.corpus import wordnet
 import nltk
 import os
+import pickle
 
 
 def lowercasing(lda_samples):
@@ -91,16 +92,8 @@ nltk.download('averaged_perceptron_tagger')
 nltk.download('wordnet')
 nltk.download('omw-1.4')
 
-preprocessing_samples = list(filter(lambda x: isinstance(x, str), df['reviewText']))
-preprocessing_samples = lowercasing(preprocessing_samples)
-preprocessing_samples = punctuation_removal(preprocessing_samples)
-preprocessing_samples = lemmatize(preprocessing_samples)
-
-tf_vectorizer = CountVectorizer(
-    max_df=0.01, min_df=2, max_features=n_features, stop_words="english"
-)
-tf = tf_vectorizer.fit_transform(preprocessing_samples)
-
+with open("vectorizer.pkl", "rb") as f:
+    tf_vectorizer = pickle.load(f)
 
 product_id = 'B009MA34NY'
 lda_samples = list(filter(lambda x: isinstance(x, str), df[df['asin'] == product_id]['reviewText']))