Update pages/model.py

pages/model.py

@@ -1,80 +1,80 @@
-
 import streamlit as st
 import pickle
 import re
 import numpy as np
 import pandas as pd
-from io import StringIO
-from streamlit_lottie import st_lottie
-import json
 
-st.set_page_config(page_title="🔖 TagGPT - Auto Tag Your Questions", layout="centered")
+st.set_page_config(page_title="🔖 TagGPT - Stack Overflow Tag Generator", layout="centered")
 
-def load_lottie(filepath: str):
-    with open(filepath, "r") as f:
-        return json.load(f)
+def clean_text(text):
+    text = re.sub(r"<.*?>", " ", text)
+    text = re.sub(r"[^a-zA-Z0-9\s]", " ", text)
+    text = re.sub(r"\s+", " ", text.lower()).strip()
+    return text
 
 @st.cache_resource
-
-def load_assets():
-    with open("model.pkl", "rb") as m:
-        tag_model = pickle.load(m)
-    with open("tfidf.pkl", "rb") as v:
-        text_vectorizer = pickle.load(v)
-    with open("mlb.pkl", "rb") as e:
-        tag_encoder = pickle.load(e)
-    return tag_model, text_vectorizer, tag_encoder
-
-tag_model, text_vectorizer, tag_encoder = load_assets()
-
-def clean_text(raw_text):
-    raw_text = re.sub(r"<.*?>", " ", raw_text)  # Remove HTML tags
-    raw_text = re.sub(r"[^a-zA-Z0-9\s]", " ", raw_text)  # Remove special characters
-    raw_text = re.sub(r"\s+", " ", raw_text.lower()).strip()
-    return raw_text
-
-st_lottie(load_lottie("tag_animation.json"), height=150, key="intro")  # Lottie animation (optional)
-
-st.title("🏷️ TagGPT - Smart Stack Overflow Tag Suggester")
-st.markdown("**🚀 Instantly generate relevant tags for your coding questions using AI.**")
-
-q_title = st.text_input("🧠 Enter your question title")
-q_desc = st.text_area("📝 Describe your problem in detail", height=200)
-
-if st.button("🔍 Predict Tags"):
-    if not q_title.strip() or not q_desc.strip():
-        st.warning("⚠️ Please provide both title and description.")
+def load_model_assets():
+    try:
+        with open("model.pkl", "rb") as m:
+            model = pickle.load(m)
+        with open("tfidf.pkl", "rb") as v:
+            vectorizer = pickle.load(v)
+        with open("mlb.pkl", "rb") as e:
+            encoder = pickle.load(e)
+        return model, vectorizer, encoder
+    except Exception as err:
+        st.error(f"❌ Failed to load model assets: {err}")
+        st.stop()
+
+model, vectorizer, encoder = load_model_assets()
+
+st.title("💡 TagGPT")
+st.markdown("✨ _Auto-tag your Stack Overflow questions using AI_")
+
+q_title = st.text_input("📝 **Question Title**")
+q_body = st.text_area("📄 **Detailed Description**", height=200)
+
+if st.button("🔮 Generate Tags"):
+    if not q_title.strip() or not q_body.strip():
+        st.warning("⚠️ Please enter both title and description.")
     else:
-        combined_input = clean_text(q_title + " " + q_desc)
-        transformed_input = text_vectorizer.transform([combined_input])
+        user_input = clean_text(q_title + " " + q_body)
+        transformed = vectorizer.transform([user_input])
 
         try:
-            if hasattr(tag_model, "predict_proba"):
-                probs = tag_model.predict_proba(transformed_input)[0]
-                pred_tags = (probs >= 0.3).astype(int)  # Fixed threshold
-            else:
-                pred_tags = tag_model.predict(transformed_input)[0]
-                probs = pred_tags
-
-            tag_names = tag_encoder.classes_
-            selected_tags = [(tag, round(probs[i]*100, 2)) for i, tag in enumerate(tag_names) if pred_tags[i] == 1]
-
-            if selected_tags:
-                st.success("🏷️ **Predicted Tags with Confidence:**")
-                for tag, score in selected_tags:
-                    st.markdown(f"**{tag}** — `{score}%`")
+            if hasattr(model, "predict_proba"):
+                probs = model.predict_proba(transformed)[0]
+                tag_names = encoder.classes_
+                tag_conf_df = pd.DataFrame({
+                    "Tag": tag_names,
+                    "Confidence": probs
+                }).sort_values("Confidence", ascending=False)
+
+                top_tags = tag_conf_df[tag_conf_df["Confidence"] > 0.3]  # adjustable threshold
+
+                if not top_tags.empty:
+                    st.success("🏷️ **Top Suggested Tags with Confidence:**")
+                    st.dataframe(top_tags, use_container_width=True)
+
+                    # Download Button
+                    csv = top_tags.to_csv(index=False).encode('utf-8')
+                    st.download_button("⬇️ Download Tags as CSV", csv, "predicted_tags.csv", "text/csv")
+
+                else:
+                    st.info("🤖 No high-confidence tags predicted. Try improving the input.")
             else:
-                st.info("🤖 No tags predicted. Try refining your input.")
-
-            # --- Download Option
-            tag_list = [tag for tag, _ in selected_tags]
-            df = pd.DataFrame({
-                "Title": [q_title],
-                "Description": [q_desc],
-                "Predicted Tags": [", ".join(tag_list)]
-            })
-            csv = df.to_csv(index=False).encode('utf-8')
-            st.download_button("📥 Download Tags as CSV", data=csv, file_name="tag_predictions.csv", mime='text/csv')
-
-        except Exception as error:
-            st.error(f"🚫 Prediction failed: {error}")
+                preds = model.predict(transformed)
+                tags = encoder.inverse_transform(preds)
+
+                if tags and tags[0]:
+                    st.success("🏷️ **Predicted Tags:**")
+                    tag_list = ", ".join(tags[0])
+                    st.markdown("🔸 " + tag_list)
+
+                    # Download plain text
+                    txt = tag_list.encode('utf-8')
+                    st.download_button("⬇️ Download Tags as TXT", txt, "predicted_tags.txt", "text/plain")
+                else:
+                    st.info("🤖 No tags predicted.")
+        except Exception as e:
+            st.error(f"🚫 Prediction failed: {e}")