commit

.gitignore

@@ -0,0 +1 @@
+.env

app.py

@@ -0,0 +1,45 @@
+import streamlit as st
+from langchain_core.messages import AIMessage, HumanMessage
+from src import get_vectorstore_from_url,get_response
+
+
+# app config
+st.set_page_config(page_title="Chat with websites", page_icon="asset/alien_008.jpg")
+st.title("Chat with websites")
+
+# sidebar
+
+with st.sidebar:
+    st.header("Settings")
+    website_url = st.text_input("Website URL",
+                                placeholder='https://en.wikipedia.org/wiki/Artificial_intelligence') 
+
+if website_url is None or website_url == "":
+    st.info("Please enter a website URL in the sidebar")
+
+else:
+    # session state
+    if "chat_history" not in st.session_state:
+        st.session_state.chat_history = [
+            AIMessage(content="Hello. How can I help you?"),
+        ]
+    if "vector_store" not in st.session_state:
+        st.session_state.vector_store = get_vectorstore_from_url(website_url) 
+
+    user_query = st.chat_input("Type your message here...")
+    
+    if user_query is not None and user_query != "":
+        response = get_response(user_query)
+        # input chat
+        st.session_state.chat_history.append(HumanMessage(content=user_query))
+        # output chat
+        st.session_state.chat_history.append(AIMessage(content=response))
+
+    # conversation
+    for message in st.session_state.chat_history:
+        if isinstance(message, AIMessage):
+            with st.chat_message("AI",avatar='asset/alien_008.jpg'):
+                st.write(message.content)
+        elif isinstance(message, HumanMessage):
+            with st.chat_message("Human",avatar='asset/human.jpg'):
+                st.write(message.content)

requirements.txt

@@ -0,0 +1,8 @@
+openai==1.12.0
+langchain==0.1.9
+langchain_openai
+streamlit
+python-dotenv
+openpyxl==3.1.2
+chromadb==0.4.24
+bs4==4.12.3

src/__init__.py

@@ -0,0 +1 @@
+from .get_vector import *

src/get_vector.py

@@ -0,0 +1,74 @@
+import streamlit as st
+from langchain_community.document_loaders import WebBaseLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import Chroma
+from langchain_openai import OpenAIEmbeddings, ChatOpenAI
+from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
+from langchain.chains import create_history_aware_retriever,create_retrieval_chain
+from langchain.chains.combine_documents import create_stuff_documents_chain
+from dotenv import load_dotenv
+import os
+
+load_dotenv()
+KEY=os.getenv("MY_KEY")
+
+
+def get_vectorstore_from_url(url):
+    # get the text in document form
+    loader = WebBaseLoader(url)
+    document = loader.load()
+    
+    # split the document into chunks
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=200,
+                                                chunk_overlap=10,
+                                                length_function=len)
+    document_chunks = text_splitter.split_documents(document)
+    
+    # create a vectorstore from the chunks
+    vector_store = Chroma.from_documents(document_chunks,
+                                          OpenAIEmbeddings(openai_api_key=KEY))
+
+    return vector_store
+
+def get_response(user_input):
+    retriever_chain = get_context_retriever_chain(st.session_state.vector_store)
+    conversation_rag_chain = get_conversational_rag_chain(retriever_chain)
+    
+    response = conversation_rag_chain.invoke({
+        "chat_history": st.session_state.chat_history,
+        "input": user_input
+    })
+    
+    return response['answer']
+
+
+def get_context_retriever_chain(vector_store):
+    llm = ChatOpenAI(openai_api_key=KEY)
+    
+    retriever = vector_store.as_retriever()
+    
+    prompt = ChatPromptTemplate.from_messages([
+      MessagesPlaceholder(variable_name="chat_history"),
+      ("user", "{input}"),
+      ("user", "Given the above conversation, generate a search query to look up in order to get information relevant to the conversation")
+    ])
+    
+    retriever_chain = create_history_aware_retriever(llm, retriever, prompt)
+    
+    return retriever_chain
+
+def get_conversational_rag_chain(retriever_chain): 
+    
+    llm = ChatOpenAI(openai_api_key=KEY)
+    
+    prompt = ChatPromptTemplate.from_messages([
+      ("system", "Answer the user's questions based on the below context:\n\n{context}"),
+      MessagesPlaceholder(variable_name="chat_history"),
+      ("user", "{input}"),
+    ])
+    
+    stuff_documents_chain = create_stuff_documents_chain(llm,prompt)
+    
+    return create_retrieval_chain(retriever_chain, stuff_documents_chain)
+
+