File size: 5,137 Bytes
364893a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
import streamlit as st
import os
from typing import List, Tuple, Optional
from pinecone import Pinecone
from langchain_pinecone import PineconeVectorStore
from langchain_huggingface import HuggingFaceEmbeddings
from langchain_openai import ChatOpenAI
from langchain_core.prompts import PromptTemplate
from dotenv import load_dotenv
from RAG import RAG
import logging

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Page configuration
st.set_page_config(
    page_title="RAG Chatbot",
    page_icon="🤖",
    layout="wide"
)

def initialize_models() -> Tuple[Optional[ChatOpenAI], HuggingFaceEmbeddings]:
    """Initialize the language model and embeddings."""
    try:
        load_dotenv()
        
        # Initialize OpenAI model
        llm = ChatOpenAI(
            model="gpt-4",  # Changed from gpt-4o-mini which appears to be a typo
            temperature=0,
            timeout=60,  # Added reasonable timeout
            max_retries=2
        )
        
        # Initialize embeddings
        embeddings = HuggingFaceEmbeddings(
            model_name="sentence-transformers/all-MiniLM-L6-v2"
        )
        
        return llm, embeddings
        
    except Exception as e:
        logger.error(f"Error initializing models: {str(e)}")
        st.error(f"Failed to initialize models: {str(e)}")
        return None, None

def process_message(

    query: str,

    llm: ChatOpenAI,

    index_name: str,

    embeddings: HuggingFaceEmbeddings

) -> Tuple[str, List]:
    """Process the user message using the RAG system."""
    try:
        response, sources = RAG(
            query=query,
            llm=llm,
            index_name=index_name,
            embeddings=embeddings
        )
        return response, sources
    except Exception as e:
        logger.error(f"Error in process_message: {str(e)}")
        return f"Error processing message: {str(e)}", []

def display_sources(sources: List) -> None:
    """Display sources in expandable sections with proper formatting and custom URLs."""
    if not sources:
        st.info("No sources available for this response.")
        return

    st.subheader("Sources")
    for i, doc in enumerate(sources, 1):
        try:
            with st.expander(f"Source {i}"):
                if hasattr(doc, 'page_content'):
                    st.markdown(f"**Content:** {doc.page_content}")
                    if hasattr(doc, 'metadata'):
                        # Construct URL from source metadata
                        
                        # Display other metadata
                        for key, value in doc.metadata.items():
                            if key != 'source':  # Skip source since we already used it for URL
                                st.markdown(f"**{key.title()}:** {value}")
                else:
                    st.markdown(f"**Content:** {str(doc)}")
        except Exception as e:
            logger.error(f"Error displaying source {i}: {str(e)}")
            st.error(f"Error displaying source {i}")

def main():
    st.title("RAG Chatbot")
    
    # Initialize session state
    if "messages" not in st.session_state:
        st.session_state.messages = []
    
    # Initialize models
    llm, embeddings = initialize_models()
    if not llm or not embeddings:
        st.error("Failed to initialize the application. Please check the logs.")
        return
    
    # Constants
    INDEX_NAME = 'bpl-rag'
    
    # Display chat history
    for message in st.session_state.messages:
        with st.chat_message(message["role"]):
            st.markdown(message["content"])
    
    # Chat input
    user_input = st.chat_input("Type your message here...")
    if user_input:
        # Display user message
        with st.chat_message("user"):
            st.markdown(user_input)
        st.session_state.messages.append({"role": "user", "content": user_input})
        
        # Process and display assistant response
        with st.chat_message("assistant"):
            with st.spinner("Thinking..."):
                response, sources = process_message(
                    query=user_input,
                    llm=llm,
                    index_name=INDEX_NAME,
                    embeddings=embeddings
                )
                
                if isinstance(response, str):
                    st.markdown(response)
                    st.session_state.messages.append({
                        "role": "assistant",
                        "content": response
                    })
                    
                    # Display sources
                    display_sources(sources)
                else:
                    st.error("Received an invalid response format")
    
    # Footer
    st.markdown("---")
    st.markdown(
        "Built with ❤️ using Streamlit + LangChain + OpenAI",
        help="An AI-powered chatbot with RAG capabilities"
    )

if __name__ == "__main__":
    main()