Update app.py

app.py

@@ -60,15 +60,6 @@ def run_llm(input_text,history):
     qur= hf.embed_query(input_text)
     docs = db.similarity_search_by_vector(qur, k=3)
     
-    '''if len(docs) >2:
-
-    doc_list = str(docs).split(" ")
-    if len(doc_list) > MAX_TOKENS:
-        doc_cnt = int(len(doc_list) / MAX_TOKENS)
-        print(doc_cnt)
-        for ea in doc_cnt:'''
-
-
     print(docs)
 
     callbacks = [StreamingStdOutCallbackHandler()]  
@@ -85,30 +76,12 @@ def run_llm(input_text,history):
     streaming=True,  
     huggingfacehub_api_token=token,
     )  
-
-
-    '''llm=HuggingFaceEndpoint(
-        endpoint_url=repo_id,
-        streaming=True,
-        max_new_tokens=2400,
-        huggingfacehub_api_token=token)'''
-    print(input_text)
-    print(history)
     out=""
     #prompt = ChatPromptTemplate.from_messages(
-    sys_prompt = f"Use this data to help answer users questions: {str(docs)}"
-    user_prompt = f"{input_text}"
     prompt=[
         {"role": "system", "content": f"[INST] Use this data to help answer users questions: {str(docs)} [/INST]"},
         {"role": "user", "content": f"[INST]{input_text}[/INST]"},
     ]
-    #chat = ChatHuggingFace(llm=llm, verbose=True)
-    messages = [  
-    ("system", f"[INST] Use this data to help answer users questions: {str(docs)} [/INST]"),  
-    ("user", f"[INST]{input_text}[/INST]"),  
-    ]  
-    
-    #yield(llm.invoke(prompt))
     
     t=llm.invoke(prompt)
     for chunk in t: