Update app.py

app.py

@@ -5,7 +5,7 @@ import logging
 from langchain.document_loaders import PDFPlumberLoader
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.prompts import ChatPromptTemplate
-from langchain.llms import HuggingFacePipeline
+from langchain.llms import HuggingFaceLLM
 from transformers import pipeline
 
 # Configure logging
@@ -13,7 +13,7 @@ logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
 # Page configuration
-st.set_page_config(page_title="DeepSeek Chatbot - ruslanmv.com", page_icon="🤖", layout="centered")
+st.set_page_config(page_title="DeepSeek Chatbot RAG", page_icon="🤖", layout="centered")
 
 # Initialize session state for chat history
 if "messages" not in st.session_state:
@@ -70,10 +70,13 @@ def generate_response_with_langchain(question, context):
     """
 
     prompt = ChatPromptTemplate.from_template(prompt_template)
-    model = HuggingFacePipeline(pipeline("text-generation", model=selected_model))
+
+    # Initialize HuggingFace model with LangChain's HuggingFaceLLM
+    hf_pipeline = pipeline("text-generation", model=selected_model)
+    llm = HuggingFaceLLM(pipeline=hf_pipeline)
 
     # Use LangChain to generate an answer
-    chain = prompt | model
+    chain = prompt | llm
     response = chain.invoke({"question": question, "context": context})
     return response
 
@@ -94,10 +97,10 @@ if uploaded_file:
     documents = process_pdf(uploaded_file)
     context = "\n\n".join([doc.page_content for doc in documents])
 
-    # Combine system message and user input into a single prompt
+    # Show the PDF-based question input if the PDF is uploaded
     prompt_input = "Ask a question about the PDF content"
 
-    # Show the PDF-based question input if the PDF is uploaded
+    # Show the chat input if PDF is uploaded
     prompt = st.chat_input(prompt_input) if documents else None
 
     if prompt: