Update app.py

app.py

@@ -6,7 +6,7 @@ import os
 # from langchain_experimental.llms import ChatLlamaAPI
 from langchain.embeddings import HuggingFaceEmbeddings
 from pinecone import Pinecone
-from langchain.vectorstores import Pinecone
+
 from langchain.prompts import PromptTemplate
 from langchain.chains import RetrievalQA
 import streamlit.components.v1 as components
@@ -14,9 +14,12 @@ from langchain_groq import ChatGroq
 from langchain.chains import ConversationalRetrievalChain
 from langchain.memory import ChatMessageHistory, ConversationBufferMemory
 import time
+import os
+os.environ['PINECONE_API_KEY'] = "fa944c7c-5775-4a96-8704-e04f7a86614e"
+pc = Pinecone()
 
 HUGGINGFACEHUB_API_TOKEN = st.secrets['HUGGINGFACEHUB_API_TOKEN']
-
+from langchain.vectorstores import Pinecone
 @dataclass
 class Message:
  """Class for keeping track of a chat message."""
@@ -39,17 +42,22 @@ def initialize_session_state():
 
  embeddings = download_hugging_face_embeddings()
 
- import os
- os.environ['PINECONE_API_KEY'] = st.secrets["PINECONE_API_KEY"]
- pc = Pinecone()
- index_name = "medical-advisor" # put in the name of your pinecone index here
+ 
+ index_name = "medical-advisor"
+
+ if index_name in pc.list_indexes().names():
+ print("index already exists" , index_name)
+ index= pc.Index(index_name) #your index which is already existing and is ready to use
+ print(index.describe_index_stats())
+ # put in the name of your pinecone index here
 
  docsearch = Pinecone.from_existing_index(index_name, embeddings)
 
  prompt_template = """
  You are a trained bot to guide people about their medical concerns. You will answer user's query with your knowledge and the context provided. 
+ If question given by user ask about personal details about them refer to last user inputs.
  If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
- Do not say thank you and tell you are an AI Assistant and be open about everything.
+ You have to act as conversational bot. Paying attention to the personal details given in the context and respond helpfully. Provide personal detail if asked
  Use the following pieces of context to answer the users question.
  Context: {context}
  Question: {question}
@@ -70,7 +78,7 @@ def initialize_session_state():
  retrieval_chain = ConversationalRetrievalChain.from_llm(llm=chat,
  chain_type="stuff",
  retriever=docsearch.as_retriever(
- search_kwargs={'k': 2}),
+ search_kwargs={'k': 3}),
  return_source_documents=True,
  combine_docs_chain_kwargs={"prompt": PROMPT},
  memory= memory
@@ -96,7 +104,7 @@ def on_click_callback():
 
 initialize_session_state()
 
-st.title("LegalEase Advisor Chatbot 🇮🇳")
+st.title("Medical Advisor Chatbot 🇮🇳")
 
 # st.markdown(
 # """