app.py

import os
import streamlit as st
from langchain_community.vectorstores import FAISS
from langchain_community.embeddings import HuggingFaceEmbeddings

from langchain_huggingface import HuggingFaceEndpoint

from langchain.prompts import PromptTemplate
from langchain.schema.runnable import RunnablePassthrough
from langchain.chains import LLMChain
# Load Google service account credentials from Hugging Face secrets
GOOGLE_SERVICE_ACCOUNT_JSON = st.secrets["GOOGLE_SERVICE_ACCOUNT_JSON"]

# Google Sheets API v4 setup
scope = ["https://www.googleapis.com/auth/spreadsheets", "https://www.googleapis.com/auth/drive"]
service_account_info = json.loads(GOOGLE_SERVICE_ACCOUNT_JSON)
creds = ServiceAccountCredentials.from_json_keyfile_dict(service_account_info, scope)
client = gspread.authorize(creds)
spreadsheet_id = '1Jf1k7Q71ihsxBf-XQYyucamMy14q7IjhUDlU8ZzR_Nc'  # Replace with your actual spreadsheet ID
sheet = client.open_by_key(spreadsheet_id).sheet1

# Function to save user feedback to Google Sheets
def save_feedback(user_input, bot_response, rating, comment):
    feedback = [user_input, bot_response, rating, comment]
    sheet.append_row(feedback)

from huggingface_hub import login
login(token=st.secrets["HF_TOKEN"])

from langchain_community.document_loaders import TextLoader
from langchain_text_splitters import CharacterTextSplitter
from langchain_community.document_loaders import PyPDFLoader
from langchain.chains import RetrievalQA
from langchain.prompts import PromptTemplate
from langchain.embeddings.huggingface import HuggingFaceEmbeddings

db = FAISS.load_local("faiss_index", HuggingFaceEmbeddings(model_name='sentence-transformers/all-MiniLM-L12-v2'),allow_dangerous_deserialization=True)



retriever = db.as_retriever(
    search_type="similarity",
    search_kwargs={'k': 2}
)


prompt_template = """
### [INST]
Instruction: You are a Q&A assistant. Your goal is to answer questions as accurately as possible based on the instructions and context provided without using prior knowledge.You answer in FRENCH
        Analyse carefully the context and provide a direct answer based on the context. If the user said Bonjour or Hello  your only answer will be  Hi! comment puis-je vous aider?
Answer in french only
        
{context}
Vous devez répondre aux questions en français.

### QUESTION:
{question}
[/INST]
Answer in french only
 Vous devez répondre aux questions en français.

 """

repo_id = "mistralai/Mistral-7B-Instruct-v0.3"

mistral_llm = HuggingFaceEndpoint(
    repo_id=repo_id, max_length=2048, temperature=0.02, huggingfacehub_api_token=st.secrets["HF_TOKEN"]
)

# Create prompt from prompt template
prompt = PromptTemplate(
    input_variables=["question"],
    template=prompt_template,
)

# Create llm chain
llm_chain = LLMChain(llm=mistral_llm, prompt=prompt)


retriever.search_kwargs = {'k':4}
qa = RetrievalQA.from_chain_type(
    llm=mistral_llm,
    chain_type="stuff",
    retriever=retriever,
    chain_type_kwargs={"prompt": prompt},
)

import streamlit as st

# Streamlit interface with improved aesthetics
st.set_page_config(page_title="Alter-IA Chat", page_icon="🤖")

# Define function to handle user input and display chatbot response
def chatbot_response(user_input):
    response = qa.run(user_input)
    return response

# Session state to hold user input and chatbot response
if 'user_input' not in st.session_state:
    st.session_state.user_input = ""
if 'bot_response' not in st.session_state:
    st.session_state.bot_response = ""

# Create columns for logos
col1, col2, col3 = st.columns([2, 3, 2])

with col1:
    st.image("Design 3_22.png", width=150, use_column_width=True)  # Adjust image path and size as needed

with col3:
    st.image("Altereo logo 2023 original - eau et territoires durables.png", width=150, use_column_width=True)  # Adjust image path and size as needed

# Streamlit components
st.markdown("""
    <style>
    .centered-text {
        text-align: center;
    }
    .centered-orange-text {
        text-align: center;
        color: darkorange;
    }
    </style>
    """, unsafe_allow_html=True)

# Use CSS classes to style the text
st.markdown('<h3 class="centered-text">🤖 AlteriaChat 🤖 </h3>', unsafe_allow_html=True)
st.markdown('<p class="centered-orange-text">"Votre Réponse à Chaque Défi Méthodologique "</p>', unsafe_allow_html=True)

# Input form for user interaction
with st.form(key='interaction_form'):
    st.session_state.user_input = st.text_input("You:", key="user_input_input")
    ask_button = st.form_submit_button("Ask 📨")  # Button to submit the question

    if ask_button and st.session_state.user_input.strip():
        st.session_state.bot_response = chatbot_response(st.session_state.user_input)

# Display the bot response if available
if st.session_state.bot_response:
    st.markdown("### Bot:")
    st.text_area("", value=st.session_state.bot_response, height=600)

    # Separate form for feedback submission
    with st.form(key='feedback_form'):
        st.markdown("### Rate the response:")
        rating = st.slider("Select a rating:", min_value=1, max_value=5, value=1, key="rating")

        st.markdown("### Leave a comment:")
        comment = st.text_area("", key="comment")

        # Separate submit button for feedback
        feedback_submit_button = st.form_submit_button("Submit Feedback")
        
        if feedback_submit_button:
            if comment.strip():
                save_feedback(st.session_state.user_input, st.session_state.bot_response, rating, comment)
                st.success("Thank you for your feedback!")
                # Clear the session state after submission
                st.session_state.user_input = ""
                st.session_state.bot_response = ""
            else:
                st.warning("Please provide a comment before submitting feedback.")

st.markdown("---")
st.markdown("Collaboration is the key to success. Each question finds its answer, each challenge becomes an opportunity.")