reference methods in chain directly

chain.py

@@ -149,11 +149,11 @@ class ChainBuilder:
         return get_and_cache_model(endpoint, extra_params)
 
     def build_chain(self):
-        model = self.get_model()
-        prompt = self.get_prompt()
-        format_context = self.format_context()
-        vector_search_as_retriever = self.get_retriever()
-        query_rewrite_prompt = self.get_query_rewrite_prompt()
+        # model = self.get_model()
+        # prompt = self.get_prompt()
+        # format_context = self.format_context()
+        # vector_search_as_retriever = self.get_retriever()
+        # query_rewrite_prompt = self.get_query_rewrite_prompt()
 
         # RAG Chain
         chain = (
@@ -169,17 +169,17 @@ class ChainBuilder:
                 "context": RunnableBranch(  # Only re-write the question if there is a chat history - RunnableBranch() is essentially a LCEL if statement
                     (
                         lambda x: len(x["chat_history"]) > 0, #https://python.langchain.com/api_reference/core/runnables/langchain_core.runnables.branch.RunnableBranch.html
-                        query_rewrite_prompt | model | StrOutputParser(), # rewrite question with context
+                        self.get_query_rewrite_prompt() | self.get_model() | StrOutputParser(), # rewrite question with context
                     ),
                     itemgetter("question"), # else, just ask the question
                 )
-                | vector_search_as_retriever # set 'context' to the result of passing either the base question, or the reformatted question to the retriever for semantic search
-                | RunnableLambda(format_context), 
+                | self.get_retriever() # set 'context' to the result of passing either the base question, or the reformatted question to the retriever for semantic search
+                | RunnableLambda(self.format_context), 
                 "formatted_chat_history": itemgetter("formatted_chat_history"),
                 "question": itemgetter("question"),
             }
-            | prompt # 'context', 'formatted_chat_history', and 'question' passed to prompt
-            | model # prompt passed to model
+            | self.get_prompt() # 'context', 'formatted_chat_history', and 'question' passed to prompt
+            | self.get_model() # prompt passed to model
             | StrOutputParser()
         )
         return chain