Update the timeout; update the params in the db method;

Author: ranjan-stha

chatbotcore/custom_embeddings.py

@@ -27,7 +27,7 @@ def __post_init__(self):
             if not (self.url and self.model_name and self.base_url):
                 raise Exception("Url or base_url or both are not provided.")
 
-    def embed_query(self, text: str, timeout: int = 240) -> List[float]:
+    def embed_query(self, text: str, timeout: int = 45) -> List[float]:
         """
         Sends the request to Embedding module to
         embed the query to the vector representation
@@ -39,7 +39,7 @@ def embed_query(self, text: str, timeout: int = 240) -> List[float]:
             raise Exception(e)
         return response.json()
 
-    def embed_documents(self, texts: List[str], timeout: int = 30) -> List[List[float]]:
+    def embed_documents(self, texts: List[str], timeout: int = 45) -> List[List[float]]:
         """
         Sends the request to Embedding module to
         embed multiple queries to the vector representation

chatbotcore/database.py

@@ -98,13 +98,13 @@ def data_search(
         # Note the results shall contain score key; sort the results using score key and get top 5 among them.
         return results
 
-    def delete_data_by_src_uuid(self, collection_name: str, key: str, value: Any) -> bool:
+    def delete_data_by_src_uuid(self, key: str, value: Any) -> bool:
         """
         Delete data by source uuid
         Note that the document source key should be doc_uuid
         """
         points_selector = FilterSelector(filter=Filter(must=[FieldCondition(key=key, match=MatchValue(value=value))]))
-        result = self.db_client.delete(collection_name=collection_name, points_selector=points_selector)
+        result = self.db_client.delete(collection_name=self.collection_name, points_selector=points_selector)
 
         return result.status == q_models.UpdateStatus.COMPLETED
 

chatbotcore/doc_loaders.py

@@ -78,6 +78,7 @@ def create_document_chunks(self):
         Creates multiple documents from the input texts
         """
         documents = [Document(page_content=self.text)]
+        # Note that the token based splitting can be used as shown below.
         # doc_chunks = self._get_split_documents_using_token_based(documents=documents)
         doc_chunks = self._get_split_documents_with_recursive_char(documents=documents)
         contextualized_chunks = self.context_retrieval.generate_contextualized_chunks(document=self.text, chunks=doc_chunks)