Created new prediction function.

app.py

@@ -11,10 +11,10 @@ from sklearn.feature_extraction.text import CountVectorizer
 from sklearn.feature_extraction.text import TfidfVectorizer
 
 # file name
-lr_filename = 'lr_021223.pkl'
+#lr_filename = 'lr_021223.pkl'
 
 # Load model from pickle file
-model = pickle.load(open(lr_filename, 'rb'))
+# model = pickle.load(open(lr_filename, 'rb'))
 
 
 # Process input text, including removing stopwords, converting to lowercase, and removing punctuation
@@ -29,21 +29,31 @@ def process_text(text):
     return text
 
 # Vectorize input text
-vectorizer = CountVectorizer()
+vec = CountVectorizer()
+'''
 def vectorize_text(text):
-    text = process_text(text)
-    text = vectorizer.fit_transform([text])
-    return text
+    #text = process_text(text)
+    #text = vectorizer.fit_transform([text])
+    #return text
+'''
 
 # Valid input for the model so number of features match
-# Code will go here
-
-# Prediction function
 def predict(text):
-    text = vectorize_text(text)
-    prediction = model.predict(text)
+    # Load the pickled model
+    filename = 'lr_021223.pkl'
+    loaded_model = pickle.load(open(filename, 'rb'))
+    text = process_text(text)
+    text = vec.transform([text])
+    prediction = loaded_model.predict(text)
     return prediction
 
+'''
+Prediction function
+#def predict(text):
+    #text = vectorize_text(text)
+    #prediction = model.predict(text)
+    #return prediction
+'''
 
 # Define interface
 demo = gr.Interface(fn=predict,