App_Update

app.py

@@ -1,31 +1,55 @@
-pip install transformers
 import streamlit as st
-from transformers import AutoTokenizer, AutoModelForSequenceClassification
+import pickle
+import nltk
+from nltk.stem import PorterStemmer
+from nltk.corpus import stopwords
+import string
+ps = PorterStemmer()
 
-# Replace "model_name" with the chosen pre-trained model name
-model_name = "model.pkl"
+#lets load the saved vectorizer and native model
+tfidf = pickle.load(open('vectorizer.pkl','rb'))
+model = pickle.load(open('model.pkl','rb'))
 
-# Load tokenizer and model from Hugging Face
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForSequenceClassification.from_pretrained(model_name)
-
-# Streamlit App Interface (similar to your previous code)
-st.title("Email Spam Classifier")
+#saving streamlit code
+st.title("Email spam Classifier")
 input_sms = st.text_area("Enter message")
+def transform_text(text):
+        text = text.lower()
+        text = nltk.word_tokenize(text)
+        ###############################################
+        # Check for alphabetics
+        y = []
+        for i in text:
+            if i.isalnum():
+                y.append(i)
+        text = y[:]
+        y.clear()
+        ######################################################
+        # Remove punctuations
+        for i in text:
+            if i not in stopwords.words('english') and i not in string.punctuation:
+                y.append(i)
+        text = y[:]  # copy content
+        y.clear()  # Clear original list
+        #######################################################
+        # append steaming method(Root words)
+        for i in text:
+            y.append(ps.stem(i))
+        return " ".join(y)
 
-if st.button('predict'):
-    # Preprocess (optional): You can still define a function here for additional cleaning
-    transformed_sms = input_sms  # Replace with your preprocessing logic if needed
-
-    # Tokenize the text
-    inputs = tokenizer(transformed_sms, return_tensors="pt")
+    # transform_text("I'm gonna be home soon and i don't want to talk about this stuff anymore tonight, k? I've cried enough today.")
+        return transform_text
 
-    # Make prediction
-    outputs = model(**inputs)
-    predictions = np.argmax(outputs.logits.detach().cpu().numpy(), axis=1)
 
-    # Display result (assuming binary classification)
-    if predictions[0] == 1:
+if st.button('predict'):
+    #preprocess
+    transformed_sms = transform_text(input_sms)
+    #vectorize
+    vector_input = tfidf.transform([transformed_sms])
+    #predict
+    result = model.predict(vector_input)[0]
+    #display
+    if result ==1:
         st.header("Spam")
     else:
-        st.header("Not Spam")
+        st.header("Not Spam")