Update ask.py

ask.py

@@ -91,14 +91,14 @@ def askQuestion(brain, question, prompt, temperature, maxTokens, combined):
 
     query = question
 
-    # new = LLMPredictor(llm=ChatOpenAI(temperature=temperature,
-    #                    model_name="gpt-3.5-turbo-16k", max_tokens=maxTokens))
-    new = LLMPredictor(llm=OpenAI(
-        temperature=temperature, model_name="text-davinci-003", max_tokens=maxTokens))
+    new = LLMPredictor(llm=ChatOpenAI(temperature=temperature,
+                       model_name="gpt-3.5-turbo-16k", max_tokens=maxTokens))
+    # new = LLMPredictor(llm=OpenAI(
+    #     temperature=temperature, model_name="text-davinci-003", max_tokens=maxTokens))
 
     service_context = ServiceContext.from_defaults(llm_predictor=new)
     query_engine = index.as_query_engine(
-        similarity_top_k=4, response_mode="compact", service_context=service_context)
+        similarity_top_k=7, response_mode="compact", service_context=service_context,text_qa_template=QA_PROMPT)
 
     query = question
     if combined == False:
@@ -108,7 +108,7 @@ def askQuestion(brain, question, prompt, temperature, maxTokens, combined):
             response.response = defaultResponse
         res = {"data": [response.response, {
             "memory": [question, response.response]}]}
-        print(response.source_nodes)
+        
         return json.dumps(res)
     else:
         response = query_engine.query(finalQuestion)
@@ -121,7 +121,8 @@ def askQuestion(brain, question, prompt, temperature, maxTokens, combined):
 def followUp(brainName, question, memory, temperature, maxTokens):
     if (memory == False):
         return "Please Initiate the Chat first.."
-
+    if(int(maxTokens)>2000):
+        maxTokens=2000
     newMemory = ConversationBufferMemory(memory_key="chat_history")
     for index in range(len(memory)):
         if index % 2 == 0: