Update app.py

app.py

@@ -2,6 +2,7 @@
 # ENVIRONMENT (MUST BE FIRST)
 # ================================
 import os
+os.environ["KERAS_BACKEND"] = "tensorflow"
 os.environ["TF_ENABLE_ONEDNN_OPTS"] = "0"
 
 # ================================
@@ -12,33 +13,34 @@ import tensorflow as tf
 import numpy as np
 import json
 from PIL import Image
-from tensorflow.keras.preprocessing.text import tokenizer_from_json
-from tensorflow.keras.preprocessing.sequence import pad_sequences
+from keras.preprocessing.text import tokenizer_from_json
 
 # ================================
-# CONSTANTS (MUST MATCH TRAINING)
+# LOAD MODEL (KERAS 3 SAFE)
 # ================================
-IMG_SIZE = (128, 128)
-MAX_LEN = 50
-LABELS = ["Critical", "High", "Medium", "Low"]
+MODEL_PATH = "fusion_ticket_model_final.keras"
+
+model = tf.keras.models.load_model(
+    MODEL_PATH,
+    compile=False
+)
+
+print("✅ Model loaded")
 
 # ================================
-# LOAD TOKENIZER (JSON – SAFE)
+# LOAD TOKENIZER (JSON ONLY)
 # ================================
-with open("tokenizer.json", "r") as f:
+with open("tokenizer.json", "r", encoding="utf-8") as f:
     tokenizer = tokenizer_from_json(json.load(f))
 
 print("✅ Tokenizer loaded")
 
 # ================================
-# LOAD FUSION MODEL (.keras)
+# CONSTANTS (MUST MATCH TRAINING)
 # ================================
-fusion_model = tf.keras.models.load_model(
-    "fusion_ticket_model_final.keras",
-    compile=False
-)
-
-print("✅ Fusion model loaded")
+IMG_SIZE = (128, 128)
+MAX_LEN = 50
+LABELS = ["Critical", "High", "Medium", "Low"]
 
 # ================================
 # IMAGE PREPROCESSING
@@ -47,21 +49,21 @@ def preprocess_image(image: Image.Image):
     image = image.convert("RGB")
     image = image.resize(IMG_SIZE)
     img = np.array(image, dtype=np.float32) / 255.0
-    img = np.expand_dims(img, axis=0)
-    return img
+    return np.expand_dims(img, axis=0)
 
 # ================================
 # TEXT PREPROCESSING
 # ================================
-def preprocess_text(text: str):
-    if text is None:
+def preprocess_text(text):
+    if not text:
         text = ""
     seq = tokenizer.texts_to_sequences([text])
-    padded = pad_sequences(seq, maxlen=MAX_LEN)
-    return padded
+    return tf.keras.preprocessing.sequence.pad_sequences(
+        seq, maxlen=MAX_LEN
+    )
 
 # ================================
-# PREDICTION FUNCTION (PURE MODEL)
+# PREDICTION FUNCTION
 # ================================
 def predict_ticket(image, text):
     if image is None:
@@ -69,44 +71,40 @@ def predict_ticket(image, text):
             "Critical": 0.0,
             "High": 0.0,
             "Medium": 0.0,
-            "Low": 0.0
+            "Low": 0.0,
         }
 
     img = preprocess_image(image)
     txt = preprocess_text(text)
 
-    probs = fusion_model.predict([img, txt], verbose=0)[0]
+    preds = model.predict([img, txt], verbose=0)[0]
 
     return {
-        "Critical": float(probs[0]),
-        "High": float(probs[1]),
-        "Medium": float(probs[2]),
-        "Low": float(probs[3]),
+        "Critical": float(preds[0]),
+        "High": float(preds[1]),
+        "Medium": float(preds[2]),
+        "Low": float(preds[3]),
     }
 
 # ================================
-# GRADIO UI
+# GRADIO APP
 # ================================
-interface = gr.Interface(
+app = gr.Interface(
     fn=predict_ticket,
     inputs=[
         gr.Image(type="pil", label="📤 Upload Ticket Screenshot"),
         gr.Textbox(
             lines=4,
             placeholder="Describe the issue (recommended)",
-            label="✍️ Ticket Description"
-        )
+            label="✍️ Ticket Description",
+        ),
     ],
-    outputs=gr.Label(
-        num_top_classes=4,
-        label="🚨 Predicted Severity"
-    ),
+    outputs=gr.Label(num_top_classes=4, label="🚨 Predicted Severity"),
     title="🎫 Ticket Severity Classification System",
     description=(
-        "This application uses a **CNN + NLP Fusion Model** trained on "
-        "ticket screenshots and descriptions.\n\n"
+        "CNN + NLP Fusion Model to predict ticket urgency.\n\n"
         "**Classes:** Critical | High | Medium | Low"
     ),
 )
 
-interface.launch(server_name="0.0.0.0", server_port=7860)
+app.launch(server_name="0.0.0.0", server_port=7860)