Update app.py

app.py

@@ -1,108 +1,73 @@
-# app.py
-# Very simple 3-language emotion detector
-# Beginner version - no cache, no complicated things
-
-import gradio as gr
+from flask import Flask, render_template, request
 import joblib
 from huggingface_hub import hf_hub_download
 from transformers import pipeline
 
-# === Your model links ===
-ENGLISH_MODEL = "E-motionAssistant/Englsih_Trained_Model_LR"
-SINHALA_MODEL = "E-motionAssistant/SInhala_Text_Emotion_Recognition_Model"
-TAMIL_MODEL   = "E-motionAssistant/Tamil_Emotion_Recognition_Model"
+app = Flask(__name__)
+
+# ─── Load models once when the app starts ────────────────────────────────
+
+# English (joblib model)
+vectorizer = joblib.load(hf_hub_download(
+    "E-motionAssistant/Englsih_Trained_Model_LR",
+    "tfidf_vectorizer.joblib"
+))
+classifier = joblib.load(hf_hub_download(
+    "E-motionAssistant/Englsih_Trained_Model_LR",
+    "logreg_model.joblib"
+))
+label_encoder = joblib.load(hf_hub_download(
+    "E-motionAssistant/Englsih_Trained_Model_LR",
+    "label_encoder.joblib"
+))
+
+# Sinhala & Tamil (transformers)
+sinhala_pipe = pipeline("text-classification", model="E-motionAssistant/SInhala_Text_Emotion_Recognition_Model")
+tamil_pipe   = pipeline("text-classification", model="E-motionAssistant/Tamil_Emotion_Recognition_Model")
+
+# ─── Routes ───────────────────────────────────────────────────────────────
 
-# === Load English model (TF-IDF + Logistic Regression) ===
-vectorizer = joblib.load(hf_hub_download(ENGLISH_MODEL, "tfidf_vectorizer.joblib"))
-classifier = joblib.load(hf_hub_download(ENGLISH_MODEL, "logreg_model.joblib"))
-label_encoder = joblib.load(hf_hub_download(ENGLISH_MODEL, "label_encoder.joblib"))
+@app.route("/", methods=["GET", "POST"])
+def home():
+    english_result = ""
+    sinhala_result = ""
+    tamil_result   = ""
 
-# === Load Sinhala and Tamil models (transformers) ===
-sinhala_classifier = pipeline("text-classification", model=SINHALA_MODEL)
-tamil_classifier   = pipeline("text-classification", model=TAMIL_MODEL)
+    if request.method == "POST":
+        action = request.form.get("action")
 
-# === English prediction function ===
-def predict_english(text):
-    if text.strip() == "":
-        return "Please write something"
-    
-    # transform text to numbers
-    numbers = vectorizer.transform([text])
-    
-    # make prediction
-    prediction_number = classifier.predict(numbers)[0]
-    
-    # convert number back to emotion name
-    emotion = label_encoder.inverse_transform([prediction_number])[0]
-    
-    return f"English → **{emotion}**"
+        if action == "predict_english":
+            text = request.form.get("english_text", "").strip()
+            if text:
+                X = vectorizer.transform([text])
+                pred = classifier.predict(X)[0]
+                emotion = label_encoder.inverse_transform([pred])[0]
+                english_result = f"Emotion: {emotion}"
+            else:
+                english_result = "Please write something"
 
-# === Sinhala prediction function ===
-def predict_sinhala(text):
-    if text.strip() == "":
-        return "කරුණාකර යමක් ලියන්න"
-    
-    result = sinhala_classifier(text)[0]
-    label = result["label"]
-    score = result["score"]
-    
-    return f"සිංහල → **{label}** ({score:.2f})"
+        elif action == "predict_sinhala":
+            text = request.form.get("sinhala_text", "").strip()
+            if text:
+                res = sinhala_pipe(text)[0]
+                sinhala_result = f"හැඟීම: {res['label']} ({res['score']:.3f})"
+            else:
+                sinhala_result = "කරුණාකර යමක් ලියන්න"
 
-# === Tamil prediction function ===
-def predict_tamil(text):
-    if text.strip() == "":
-        return "கருணையுடன் உரையை உள்ளிடவும்"
-    
-    result = tamil_classifier(text)[0]
-    label = result["label"]
-    score = result["score"]
-    
-    return f"தமிழ் → **{label}** ({score:.2f})"
+        elif action == "predict_tamil":
+            text = request.form.get("tamil_text", "").strip()
+            if text:
+                res = tamil_pipe(text)[0]
+                tamil_result = f"உணர்வு: {res['label']} ({res['score']:.3f})"
+            else:
+                tamil_result = "கருணையுடன் உரையை உள்ளிடவும்"
 
-# === Create the interface ===
-demo = gr.TabbedInterface(
-    [
-        # Tab 1 - English
-        gr.Interface(
-            fn=predict_english,
-            inputs=gr.Textbox(lines=3, placeholder="Type English text here..."),
-            outputs="text",
-            title="English Emotion",
-            examples=[
-                "I am very happy today",
-                "This is so boring",
-                "My head hurts"
-            ]
-        ),
-        
-        # Tab 2 - Sinhala
-        gr.Interface(
-            fn=predict_sinhala,
-            inputs=gr.Textbox(lines=3, placeholder="සිංහලෙන් ලියන්න..."),
-            outputs="text",
-            title="සිංහල හැඟීම්",
-            examples=[
-                "මම ගොඩක් සතුටින්",
-                "මට බයයි",
-                "හිසරදයි බොහොමයි"
-            ]
-        ),
-        
-        # Tab 3 - Tamil
-        gr.Interface(
-            fn=predict_tamil,
-            inputs=gr.Textbox(lines=3, placeholder="தமிழில் எழுதுங்கள்..."),
-            outputs="text",
-            title="தமிழ் உணர்வு",
-            examples=[
-                "இன்று மிகவும் சந்தோஷமாக இருக்கிறேன்",
-                "எனக்கு ரொம்ப கோபமா இருக்கு",
-                "தலைவலி கொடுமை"
-            ]
-        )
-    ],
-    tab_names=["English", "සිංහල", "தமிழ்"]
-)
+    return render_template(
+        "index.html",
+        english_result=english_result,
+        sinhala_result=sinhala_result,
+        tamil_result=tamil_result
+    )
 
-# Start the app
-demo.launch(server_name="0.0.0.0", server_port=7860)
+if __name__ == "__main__":
+    app.run(host="0.0.0.0", port=7860, debug=False)