diff --git a/api/core/index/vector_index/weaviate_vector_index.py b/api/core/index/vector_index/weaviate_vector_index.py
index 1305e576c3..23fa4ec8f5 100644
--- a/api/core/index/vector_index/weaviate_vector_index.py
+++ b/api/core/index/vector_index/weaviate_vector_index.py
@@ -141,6 +141,16 @@ class WeaviateVectorIndex(BaseVectorIndex):
             "valueText": document_id
         })
 
+    def delete_by_group_id(self, group_id: str):
+        if self._is_origin():
+            self.recreate_dataset(self.dataset)
+            return
+
+        vector_store = self._get_vector_store()
+        vector_store = cast(self._get_vector_store_class(), vector_store)
+
+        vector_store.delete()
+
     def _is_origin(self):
         if self.dataset.index_struct_dict:
             class_prefix: str = self.dataset.index_struct_dict['vector_store']['class_prefix']
diff --git a/api/core/tool/dataset_multi_retriever_tool.py b/api/core/tool/dataset_multi_retriever_tool.py
index 42dab1ca98..88eabe044d 100644
--- a/api/core/tool/dataset_multi_retriever_tool.py
+++ b/api/core/tool/dataset_multi_retriever_tool.py
@@ -82,7 +82,8 @@ class DatasetMultiRetrieverTool(BaseTool):
         hit_callback.on_tool_end(all_documents)
         document_score_list = {}
         for item in all_documents:
-            document_score_list[item.metadata['doc_id']] = item.metadata['score']
+            if 'score' in item.metadata and item.metadata['score']:
+                document_score_list[item.metadata['doc_id']] = item.metadata['score']
 
         document_context_list = []
         index_node_ids = [document.metadata['doc_id'] for document in all_documents]
diff --git a/api/core/tool/dataset_retriever_tool.py b/api/core/tool/dataset_retriever_tool.py
index f76a8904b9..b38fd82493 100644
--- a/api/core/tool/dataset_retriever_tool.py
+++ b/api/core/tool/dataset_retriever_tool.py
@@ -158,7 +158,8 @@ class DatasetRetrieverTool(BaseTool):
             document_score_list = {}
             if dataset.indexing_technique != "economy":
                 for item in documents:
-                    document_score_list[item.metadata['doc_id']] = item.metadata['score']
+                    if 'score' in item.metadata and item.metadata['score']:
+                        document_score_list[item.metadata['doc_id']] = item.metadata['score']
             document_context_list = []
             index_node_ids = [document.metadata['doc_id'] for document in documents]
             segments = DocumentSegment.query.filter(DocumentSegment.dataset_id == self.dataset_id,