fixed resolution

app.py

@@ -266,14 +266,12 @@ def generate():
         is_text_to_video = 'text-to-video' in model_endpoint
         is_image_to_video = 'image-to-video' in model_endpoint
 
-        if 'image_size' in data:
-            fal_arguments['image_size'] = data['image_size']
-        if 'num_images' in data:
-            fal_arguments['num_images'] = data['num_images']
-
         if is_text_to_image:
             # Image generation (text-to-image)
-            pass
+            if 'image_size' in data:
+                fal_arguments['image_size'] = data['image_size']
+            if 'num_images' in data:
+                fal_arguments['num_images'] = data['num_images']
         elif is_text_to_video:
             # Video generation (text-to-video)
             for k in ['aspect_ratio', 'resolution', 'duration', 'camera_fixed']:
@@ -294,6 +292,10 @@ def generate():
                     fal_arguments[k] = data[k]
         else:
             # Image edit mode (default)
+            if 'image_size' in data:
+                fal_arguments['image_size'] = data['image_size']
+            if 'num_images' in data:
+                fal_arguments['num_images'] = data['num_images']
             processed_image_urls = []
             for url in data.get('image_urls', []):
                 if url.startswith('data:'):