Update model/multimodal_encoder/builder.py

model/multimodal_encoder/builder.py

@@ -3,7 +3,7 @@ from .clip_encoder import CLIPVisionTower
 
 
 def build_vision_tower(vision_tower_cfg, **kwargs):
-    vision_tower = getattr(vision_tower_cfg, 'mm_vision_tower', getattr(vision_tower_cfg, 'vision_tower', None))
+    vision_tower = 'google/siglip-so400m-patch14-384'
     is_absolute_path_exists = os.path.exists(vision_tower)
     if is_absolute_path_exists or vision_tower.startswith("openai") or vision_tower.startswith("laion") or vision_tower.startswith("google"):
         return CLIPVisionTower(vision_tower, args=vision_tower_cfg, **kwargs)