Update app.py

app.py

@@ -4,62 +4,45 @@ import numpy as np
 from PIL import Image
 import os
 import time
-import joblib
+import cloudpickle
 
-# Config
 IMAGE_SIZE = (64, 64)
 LEADERBOARD_PATH = "leaderboard.csv"
 LABEL_FILE = "labels.csv"
 TEST_DIR = "test_images"
 
-# ====================================
-# Image preprocessing
-# ====================================
 def image_to_features(image: Image.Image) -> np.ndarray:
     image = image.resize(IMAGE_SIZE)
     return np.array(image.convert("L")).flatten()
 
-# ====================================
-# Load test images + labels
-# ====================================
 def load_test_data():
     if not os.path.exists(LABEL_FILE) or not os.path.exists(TEST_DIR):
-        raise FileNotFoundError("Missing test data or label file.")
-    
-    df = pd.read_csv(LABEL_FILE)
-    X_test = []
-    y_test = []
+        raise FileNotFoundError("Missing labels.csv or test_images/ folder.")
 
+    df = pd.read_csv(LABEL_FILE)
+    X_test, y_test = [], []
     for _, row in df.iterrows():
         img_path = os.path.join(TEST_DIR, row["filename"])
-        label = row["label"]
         try:
             img = Image.open(img_path)
-            features = image_to_features(img)
-            X_test.append(features)
-            y_test.append(label)
+            X_test.append(image_to_features(img))
+            y_test.append(row["label"])
         except Exception as e:
-            print(f"Error reading {img_path}: {e}")
-    
+            print(f"❌ Error loading {img_path}: {e}")
     return np.array(X_test), np.array(y_test)
 
-# ====================================
-# Initialize leaderboard
-# ====================================
 if not os.path.exists(LEADERBOARD_PATH):
     pd.DataFrame(columns=["Name", "Accuracy", "Avg Time (ms)"]).to_csv(LEADERBOARD_PATH, index=False)
 
-# ====================================
-# Evaluate uploaded model
-# ====================================
 def evaluate_model(file, name):
     try:
-        model = joblib.load(file.name)
-        X_test, y_test = load_test_data()
+        with open(file.name, "rb") as f:
+            model = cloudpickle.load(f)
 
+        X_test, y_test = load_test_data()
         start = time.time()
         y_pred = model.predict(X_test)
-        elapsed = (time.time() - start) * 1000  # ms
+        elapsed = (time.time() - start) * 1000
 
         if len(y_pred) != len(y_test):
             return "❌ Model output length does not match test set."
@@ -68,29 +51,26 @@ def evaluate_model(file, name):
         avg_time = elapsed / len(X_test)
 
         leaderboard = pd.read_csv(LEADERBOARD_PATH)
-        new_entry = pd.DataFrame([{
+        leaderboard = pd.concat([leaderboard, pd.DataFrame([{
             "Name": name,
             "Accuracy": round(accuracy, 2),
             "Avg Time (ms)": round(avg_time, 2)
-        }])
-        leaderboard = pd.concat([leaderboard, new_entry], ignore_index=True)
+        }])])
         leaderboard.to_csv(LEADERBOARD_PATH, index=False)
 
         return leaderboard
 
     except Exception as e:
-        return f"❌ Error during evaluation: {e}"
+        import traceback
+        return f"❌ Error:\n{traceback.format_exc()}"
 
-# ====================================
-# Launch Gradio app
-# ====================================
 gr.Interface(
     fn=evaluate_model,
     inputs=[
-        gr.File(label="Upload your `.joblib` model"),
+        gr.File(label="Upload your `classifier-joblib.joblib` file"),
         gr.Text(label="Your name or team")
     ],
     outputs=gr.Dataframe(label="Leaderboard"),
-    title="🧠 Joblib Classifier Leaderboard",
-    description="Upload a `.joblib` model trained on 64×64 grayscale images (flattened). We'll test it and update the leaderboard."
+    title="🧠 Olive Fly Classifier Leaderboard",
+    description="Upload your `classifier-joblib.joblib` model trained on 64×64 grayscale images. We'll evaluate it and update the leaderboard."
 ).launch()