Update app.py

app.py

@@ -77,10 +77,11 @@ def make_gradcam(img_path, model, last_conv_layer_name="conv5_block32_concat", c
 
     with tf.GradientTape() as tape:
         conv_outputs, predictions = grad_model(input_array)
+        predictions = tf.convert_to_tensor(predictions)   # 👈 correction
         loss = predictions[:, class_index]
 
-    grads = tape.gradient(loss, conv_outputs)[0]   # shape: (H, W, C)
-    pooled_grads = tf.reduce_mean(grads, axis=(0, 1))  # shape: (C,)
+    grads = tape.gradient(loss, conv_outputs)[0]   # (H, W, C)
+    pooled_grads = tf.reduce_mean(grads, axis=(0, 1))  # (C,)
 
     conv_outputs = conv_outputs[0]  # (H, W, C)
     heatmap = tf.reduce_sum(conv_outputs * pooled_grads, axis=-1)  # (H, W)
@@ -91,11 +92,12 @@ def make_gradcam(img_path, model, last_conv_layer_name="conv5_block32_concat", c
     heatmap = cv2.applyColorMap(heatmap, cv2.COLORMAP_JET)
 
     superimposed_img = cv2.addWeighted(
-        cv2.cvtColor(img_array.astype("uint8"), cv2.COLOR_RGB2BGR), 
+        cv2.cvtColor(img_array.astype("uint8"), cv2.COLOR_RGB2BGR),
         0.6, heatmap, 0.4, 0
     )
     return cv2.cvtColor(superimposed_img, cv2.COLOR_BGR2RGB)
 
+
 # ---- Fonction Gradio ----
 def gradio_predict(image_file):
     probs = predict_single(image_file)